text
stringlengths 87
880k
| pmid
stringlengths 1
8
| accession_id
stringlengths 9
10
| license
stringclasses 2
values | last_updated
stringlengths 19
19
| retracted
stringclasses 2
values | citation
stringlengths 22
94
| decoded_as
stringclasses 2
values | journal
stringlengths 3
48
| year
int32 1.95k
2.02k
| doi
stringlengths 3
61
| oa_subset
stringclasses 1
value |
|---|---|---|---|---|---|---|---|---|---|---|---|
==== Front
spbir
BIR
Business Information Review
0266-3821
1741-6450
SAGE Publications Sage UK: London, England
10.1177_02663821221109928
10.1177/02663821221109928
Research Articles
Sustainable knowledge management during crisis: Focus on Covid-19 pandemic
Nyoni Austin M
193167 Sharda University , Greater Noida, India
https://orcid.org/0000-0001-5938-166X
Kaushal Sanjay
193167 Sharda University , Greater Noida, India
Sanjay Kaushal, Sharda University, Greater Noida 201310, India. Email: [email protected]
12 2022
12 2022
12 2022
39 4 136146
© The Author(s) 2022
2022
SAGE Publications
This article is made available via the PMC Open Access Subset for unrestricted re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the COVID-19 pandemic or until permissions are revoked in writing. Upon expiration of these permissions, PMC is granted a perpetual license to make this article available via PMC and Europe PMC, consistent with existing copyright protections.
This paper consolidates literature that justifies effective knowledge management as a precursor for mitigating the effects of a crisis, Covid-19 pandemic in particular, through key antecedents of leadership, culture, and information and communication technology (ICT). A thorough review of retrieved literature relevant to the topic was conducted. The study materials were rigorously screened to limit any potential biases regarding their selection. Through the study, the paper concludes that the fight against Covid-19 crisis indeed requires knowledge to, among other things, find a lasting solution, mitigate the impacts, limit misinformation, revert to normalcy, and plan for similar crises in future. Further, the paper concludes that sustainable knowledge management during the Covid-19 crisis largely depends on a decisive leadership style that puts employees at the centre; a culture that embraces knowledge as a core asset, and supportive ICT infrastructure. Furthermore, the study reveals that relevance of ICT in the process of managing knowledge, largerly depends on a culture that accepts knowledge as a critical resource. The study establishes some challenges associated with ICT where a way forward for migrating from knowledge capture to knowledge creation and sharing has been re-affirmed. The present paper has led to the development of a model that further explains the relationships between the determinants of leadership, culture and ICT against effective management of a crisis using knowledge as a strategic resource. Further, six propositions have been put forward to provide clarity on the relationships.
Knowledge management
knowledge sharing
crisis management
Covid-19
information and communication technology
typesetterts10
==== Body
pmcIntroduction
Managing a crisis is a daunting task due to, among other things, its complexity as it comes unexpectedly and catches everyone off guard (Ammirato et al., 2021). According to Wang and Belardo (2005, p. 2), terms like “disaster” and “emergency” are synonymous with crisis. Moe et al. (2007, p. 787) posited that a disaster is “a condition that beats local capacity thereby necessitating a call for joint efforts both at national and international level.” Pearson and Clair (1998, p. 3) explained an organisational crisis as “a low likelihood, high impact occurrence that poses devastating threats to the survival of an organisation and is characterised by ambiguity of cause, effect, and means of resolution, and calls for timely decisions.”Jasko et al. (2012) alluded that despite the low likelihood of crisis occurrences, there is a greater need for organisations, as long as they operate, to be always on alert for a possibility of resolving a certain kind of crisis.
Moe et al. (2007, p. 787) classified disasters into two categories, i.e. technological and natural. According to the authors, technological disasters are like “transport accidents,” and “industrial accidents.” On the other hand, natural disasters are categorised into three categories:the first category includes “hydro-meteorological disasters” like “flush floods,” and “hail storms”; the second category refers to “geographical disasters” like “earthquakes”; and the third category refers to “biological disasters” like “epidemics” (Moe et al., 2007, p. 787). According to Ammirato et al. (2021), influenza pandemics fall under the category of biological disasters which occur naturally. According to the authors, the term pandemic originates from the Greek word “pandemos” which simply means “common to everyone” (p.2). Ammirato et al. (2021, p. 2) described pandemic as “a spread of a new disease at global level.” Clark (2016) cited in Ammirato et al. (2021, p. 2) captured some notable previous pandemics such as “black death which occurred between 1347 and 1351, Spanish flu which occurred between1918 and 1919, and then Covid-19,” among others.
Despite the difficulties associated with managing pandemic crises, managers in various organisations still have an obligation to cushion its implications through decision making (Ammirato et al., 2021). However, according to the authors, for the managers to make an informed decision that best fits the crises’ demands, the need for knowledge cannot be overemphasised. Despite this submission, there has been no study, to our knowledge, that has focused on sustaining knowledge management as a strategic resource for managing Covid-19 crisis with a focus on antecedents of leadership, culture, and ICT and then coming up with a framework that demonstrates the relationships.
This paper, therefore,explores relevant literature on the subject area of sustaining knowledge management amidst the Covid-19 crisis with a focus on effective leadership,culture, and ICT. Therefore, the paper is structured as follows: Research methodology, literature review followed by propositions, discussion, study implications, study limitations and conclusion, and references.
Methodology
The present study has adopted the literature review method where relevant secondary data was critically and thoroughly scrutinised and appraised to bring new knowledge that can further explain the variables through a model as posited by Snyder (2019) and Torraco (2005). To widen the scope for acquiring the study materials and limiting biases associated with the identification of the same, the search for relevant literature was guided by the three-phased approach as guided by Rosenbush et al. (2013). At the onset, keywords such as knowledge management, knowledge sharing, crisis management, Covid-19, and determinantswere identified based on the topic under study. These keywords were used as input for the computerised search process. To further widen the search process, the term “determinants” was being substituted by other terms like “antecedents” and “factors.”The actual retrieval was preceded by a careful perusal of the abstracts to get an overall picture of the content. This aimed to ensure retrieval of onlyrelevant materialsfor the topic under investigation.
The search initially targeted credible databases like Scopus and Web of Science to look for relevant published studies. Further, relevant journals like the Journal of knowledge management, international journal of knowledge management, journal of knowledge economy, and journal of information, knowledge management, and VINE Journal of Information and Knowledge Management Systems were manually browsed. This process was further facilitated by the use of the ABDC journal list, followed by a thorough check of the study materials highlighted in the reference list of each retrieved study material. After manual screening by the authors of this paper, the retrieved study materials were subjected to yet another screening process by knowledge management experts. Again, this was to ensure that the present study has used only the relevant study materials. Since the topic under investigation is relatively new and emerging, we felt that the critical literature review would be the most suitable approach, as posited by Snyder (2019).
Literature review
Focus on Covid-19 pandemic as a crisis
Covid-19 was first detected in December, 2019 as an outbreak in a Chinese city, Wuhan (Bratianu, 2020; Tomé et al., 2022). Following its emergency and the subsequent rapid spread and effects at the global level, the World Health Organization declared it a pandemic on 11th March 2020 (Bratianu, 2020). According to Baldwin and Weder di Mauro (2020) cited in Bratianu (2020), the disease is highly communicable, and since its advent, it has razed havoc in all sectors worldwide. Apart from claiming the lives of many people and affecting others in one way or another, the pandemic has also not spared business operations (Wang et al., 2020; Solnit, 2020; Peeri et al., 2020; Bratianu, 2020). Against this background, the pandemic qualifies as a global crisis as it has been characterised by devastating effects to the extent that some researchers equate it with “world war” (Tomé et al., 2022).
To manage the pandemic, many countries have been devising and implementing preventive measures such as masking up, maintaining social distance, washing hands, sanitising, and lockdowns, all aimed at containing the further spread of the pandemic (Bratianu, 2020). According to the author, despite the positive impact of the curfews in curbing further spread of the pandemic, the same has affected businesses’ normal operations, leading to other socio-economic implications that further justify the pandemic’s complexity as a crisis.
Literature has revealed that the pandemic has heavily paralysed normal operations of various businesses. In some cases, businesses have ceased to operate, scaled down their operations, or adopted some new techniques of discharging their operations like working from home, working in shifts, and virtual assemblies (Kniffin et al., 2021; Kaushik and Guleria, 2020; Wang et al., 2020). Some of the sectors heavily affected by the pandemic are “transportation, hospitality, manufacturing and education” (Kaushik and Guleria, 2020, p. 9). Despite having singled out these sectors, it should be noted here that Covid-19’s effects have not spared any business sector. This has therefore led to the following proposition (P):
P1 Impact of Covid-19 crisis has heavily affected normal business operations
Managing Covid-19 crisis
Despite the complexity associated with managing a crisis, a pandemic,particularly its implications, can be alleviated through informed decision-making in an organisation (Ammirato et al., 2021, p.1). Mittroff et al. (1987) conceded the complexity associated with any crisis. However, the authors posited that there are still possibilities of managing organisations in the midst of a crisis within that complexity. Pearson and Clair (1998, p.3) define organisational crisis management as “a systematic way whereby organisational members together with external key stakeholders joint efforts in mitigating crises or to effectively manage those that occur.” The authors highlighted the difference between managing an organisation under normal circumstances against managing the same organisation during a crisis which calls for “detecting early signals of potential crisis, make efforts to avert the crisis, and mitigating it in an event that it has occurred.”
Based on the previous crises and their subsequent implications towards various economies, the concept of crisis management has become an area of interest to various research scholars as they seek to find better ways of hedging against, responding to, and recovering from the effects of a crisis (Jasko et al., 2012; Wang and Belardo, 2005). To this effect, literature has it on record that several researchers (e.g. Roberts, 1990; Schwartz, 1987; Shrivastava et al., 1988; Weick, 1988) made an effort to investigate how different organisations manage their operations during crises. Mittroff et al. (1987) posited that managing a crisis involves going through four phases of “detection, crisis, repair, and assessment” as indicated in the model below.
According to Mittroff et al. (1987), the initial stage of detection demands environmental alertness through continuous assessments. At this stage, the author posits thatefforts are mainly put on preparation and prevention which are proactive in nature. However, when the initial phase fails due to other uncontrollable factors and a crisis ensues, efforts are put into coping, like what is happening with Covid-19 where people are learning to live with it, which is reactive in nature (Mittroff et al., 1987). Upon subsiding the crisis, the author says that there takes place repair and assessment stages that are reactive in nature where the focus is mostly put on fixing the damages caused by the crisis. Based on the assessments made, some lessons are learnt that can help in preparing for future crises in a proactive manner. Based on this model, it can be established that managing a crisis in an organisational set-up is a continuous process, and any setback in the process can lead to far-fetching consequences.
Nonetheless, alertness for any eventuality is of paramount importance in managing a crisis (Mittroff et al., 1987). Based on how fatal the Covid-19 has so far been as posited by Bratianu (2020), and taking into cognisance how various sectors have so far been affected and how the healthcare systems have so far been overstretched, the degree of preparedness for a potential pandemic like Covid-19 at global level having learnt lessons from previous crises, requires substantiation. In their research findings inthe research that was done in China, Peeri et al. (2020) concluded that as a country, China was not prepared enough to contain the pandemic despite previous lessons from similar crises. This has led to the following proposition:
P2 The degree of preparedness at global level in anticipating a potential pandemic has negatively affected effective management of Covid-19 crisis
Managing Covid-19 crisis through knowledge management
According to Clark (2016) cited in Ammirato et al. (2021), a pandemic is characterisedby, among other things, being “fatal, instilling a sense of fear and restless, faster spreading more especially when there is no vaccination to contain it”. In addition to these characteristics, Covid-19 implications have further worsened due to its continuous emergency in different forms, increasing the uncertainty of its lead period (Bratianu, 2020). According to Ammirato et al. (2021), the objective behind managing a disaster is to “ensure reduction of potential losses, ensure timely and suitable interventions to victims of the disaster in an event that the disaster has hit, and achieve timely and effective recovery after the crisis.” These interventions, according to the authors, require decision making.
Research has revealed that what makes organisations differ in how they manage crises is decision-making, which, in most cases, depends upon how knowledge isutilised as a strategic resource (Ammirato et al., 2021). According to the authors, decision-making during a crisis like the Covid-19 pandemic is supposed to be timely and with utmost accuracy so as to limit further spread of the virus. In other words, it simply means that if decisions are not being made during a crisis, or they are being made but not on time and with little or no accuracy, the crises cannot be mitigated, and the implications can be dire. However, the authors further indicate that the quality of a decision made during a pandemic largely depends on how knowledge is managed as a strategic resource. This submission is complemented by Tomé et al. (2022), who posited that Covid-19 crisis is knowledge bound, and its solutions also require the same knowledge.
Various researchers have proved the essence of knowledge as a key resource in today’s business environment which is being termed as a knowledge based economy (Linderman et al., 2004; Schiuma et al., 2012; Wee and Chua, 2013; Mahdi et al., 2019). Knowledge, which combines such attributes as experiences acquired over a period of time helps solve various challenges facing an organisation by, among other things, bringing in new ideas (Wu and Lee, 2007; Shahzad et al., 2016).
Ammirato et al. (2021, p.3) allude that “knowledge management in pandemics has a higher strategic objective as compared to other disasters, and the ultimate goal is to save life.” According to Seneviratne et al. (2010)cited in Ammirato et al. (2021, p. 2), knowledge management in a pandemic like Covid-19 has the potential of “enhancing the process of disaster management by ensuring availability and accessibility of accurate and reliable disaster risk information as and when required through effective lesson learning.” Overall,Ammirato et al. (2021, p. 2) indicate that in a crisis like Covid-19, knowledge management is of paramount importance during a pandemic as opposed to any other crisis since the ultimate goal is to limit morbidity and mortality rates. Literature has revealed that the Covid-19 crisis has re-affirmed the value of knowledge as a strategic resource as it has enhanced the need for “creation of knowledge to fight against the pandemic, knowledge sharing through appropriate channels, and decision making by stakeholders,” among other things (Ammirato et al., 2021, p. 3). Jennex and Raman (2009, p.75) posited that “the potential of an enterprise to survive during a crisis hinges on its ability to quickly respond to the phenomena which can be possible through knowledge management,” among other things.
Literature has further revealed the significance of knowledge management in managing a crisis among the crisis management stages of “repair, assessment, and detection” (Jennex and Raman, 2009). In complimenting the submission made by the latter author, Bratianu (2020) emphasisesthe uniqueness of each single crisis compared to those that previously had hit the masses. The author indicates that despite some similarities between similar crises, the distinctive features are inevitable, and knowledge gaps arise through such differences. The author further indicates that in an effort to bridge the gaps, new knowledge is created, which should also be shared and stored in readiness for future crises. Jennex and Raman (2009) posit that new knowledge acquired in managing the present crisis can help make decisions regarding the preparation and prevention of potential future crises. The authors indicate that managing a crisis usually calls for proficiency, experiences, competency, and resources from various stakeholders hence the need for knowledge management to successfullyrestrain the vice.
In a nutshell, Jennex and Raman (2009) posit that knowledge management is essential in all stages of managing a crisis, as indicated in Figure 1. Before crisis, the authors indicate that there is need for knowledge on how the previous crises which are similar to the present one were managed, during the crisis, there is need for knowledge on how to deal with it bearing in mind the gaps that have arisen, and when the crisis is over, there is need for knowledge on how to recover to normalcy as well as planning for any similar eventualities in future. Having proven through literature that Covid-19 is indeed a crisis and knowledge management is crucial in managing a crisis, the relationship between knowledge management and management of Covid-19 crisis can therefore not be overemphasised. This has led to the following proposition:Figure 1. Crisis management model (Mittroff et al., 1987).
P3 Knowledge management is positively related to effective management of Covid-19 crisis
Sustaining knowledge management during crisis: focus on key antecedents of leadership, culture, and ICT
Leadership
Various research scholars (e.g. Yin et al., 2019; Sayyadi, 2019) have proven the essence of leadership towards effective management of knowledge in an organisation. In particular, leadership styles like the transformational stylehave emerged to significantly enhance knowledge sharing behaviours among employees (Yin et al., 2019). In the same vein, the need for leadership in managing a crisis is inevitable since leaders are entrusted with the responsibility of making decisions as regards the whole process of managing the crises (Adamu et al., 2016; Wooten and James, 2008; Heide and Simonsson, 2011). Burnett (2002) cited in Lee et al. (2020), indicates that leadership is crucial in managing a crisis as it plays the role of disseminating information aimed at mitigating the crisis and recovering from its effects when the crisis comes to an end. Johansen et al. (2012) highlighted that due to the nature of a crisis which is mostly characterised by higher levels of doubts, the need for a continuous flow of information is inevitable, and it is incumbent upon the leadership to ensure that this is being done.
However, according to Kalev (2020) cited in Lee et al. (2020), not every leadership behavior brings about positive impact in managing a crisis but those capable of inducing knowledge sharing practices among employees. To this effect, the author posited that diversity-centered leadership is critical for effective management of knowledge during a crisis as it encourages critical thinking that can positively respond to the status quo. Other researchers (e.g.Carmeli et al., 2010; Nishii and Mayer, 2009; Luu et al., 2019; Nembhard and Edmondson, 2006) also found that based on the characteristics of employee centeredness, diversity-centered leaders are able to cultivate the best knowledge among the employees since they feel being part of the team. Such leadership styles have the potential to enhance the degree of trust among employees since they are perceived to have the much-needed know-how. According to the author, this trust is positively related to creating new knowledge and disseminating the same.
As it has already been highlighted that knowledge management is critical in all stages of managing a crisis (Jennex and Raman, 2009), it, therefore, followsthat managers should always be thinking about how to sustain it in such a way that it continues to serve the purpose as a strategic asset. To this effect, the need for leadership that takes onboard views of employees is inevitable.
Another leadership behaviour, which, according to Adamu et al. (2016) has been proved to have positive impact towards effective management of a crisis is that which encourages active participation. According to the authors, this leadership behaviour is demonstrated by transformational leaders and is characterised by, among others, the ability to activate employees to take part in identifying the problems, analysing them, and coming up with possible solutions (Wooten and James, 2004). From these submissions, it can be established that despite leadership being critical in managing a crisis, not all leadership behaviours are capable of effectively managing a crisis through knowledge management but those that are able to engage. This leads to the following proposition:
P4 Employee involvement in decision making is positively related to effective crisis management
Culture
In an organisational set-up, culture incorporates beliefs or norms which have the potential of either influencing or hindering the creation and sharing of knowledge that can benefit the organisation through, among other things, innovativeness (Janz and Prasarnphanich, 2003; Alavi and Leidner, 2001; Michailova and Minbaeva, 2012; Razmerita et al., 2016). Du Plesis (2007) alludes that organisational knowledge management efforts cannot be successful without a culture that supports its implementation, as there is a lack of ownership and support. To this effect, the author emphasisesthe need for organisations to build a culture that values knowledge as a critical asset, a culture that is capable of creating new knowledge, sharing it, and utilising it for various innovations that eventually lead to effectiveness. Several other researchers have so far proved the positive linkage between culture and effective management of knowledge in an organisational set-up (e.g. De long and Fahey, 2000; Egan et al., 2004; Mittal and Kumar, 2019; Connelly and Kelloway, 2003; Cabrera et al., 2006)
According to Pauchant and Mitroff (1992) cited in Boin (2008, p.14),“perceptions of senior managers in an enterprise as regards crisis management plays a key role in shaping the cultural beliefs in the organisation about the value and need for crisis management interventions.” Therefore, enterprises where senior managers believe that crisis management is one of the priority areas worthy of pursuing put in place tailor-made plans aimed at managing potential crises (Boin, 2008). This is not the case in enterprises where the senior managers believe that they are exempted from any potential crises; hence, they do not realizethe need for planning (D’Aveni and MacMillan, 1990; Dutton and Duncan, 1987; Kiesler and Sproull, 1992). The authors further posited that in an enterprise, readily available crisis management plans and policies are also not enoughif there is no culture that supports those ideologies. To this effect, the need for senior managers’ efforts to nurture that culture cannot be overemphasised as posited by Behme and Becker (2021) who highlighted the need for leaders to provide incentives to their employees to enhance culture of knowledge sharing. According to the authors, providing incentives to employees does not only help in enhancing their knowledge sharing behaviour, but also ensures the credibility of the shared knowledge. It has to be emphasised here that only credible knowledge can help make quality decisions suitable for managing a crisis. In their survey, Behme and Becker (2021) established elements of inaccuracies in some knowledge shared which is attributed to culture. To this end, the need for senior managers to nurture a culture capable of sharing credible knowledge during a crisis cannot be overemphasised. This has led to the following proposition:
P5a Senior management’s positive attitude towards crisis management interventions is positively related to building a culture that supports the efforts of managing the crisis
P5b Incentives are positively related to knowledge sharing culture
Information and Communication Technology
ICT facilitates the processes associated with effective and efficient knowledge management in an organisational set-up (Wang et al., 2006; Dang et al., 2018). Specifically, ICT is used to link people together, create, disseminate, store and retrieval of knowledge as and when required (Al Alawi et al., 2007; Lee and Lee, 2007; Mehta, 2008; Podrug et al., 2017). ICT encompasses such things as “hardwares, softwares, databases, and platforms that can be usedfor capturing, storing, and disseminating knowledge with minimum barriers (Casimir et al., 2012; Pérez-Lópezand Alegre, 2012; Podrug et al., 2017; Rathi et al., 2014; Lee and Lee, 2007; Kim and Choi, 2018; Dang et al., 2018)
According to Becker and Behme (2021), looking at how Covid-19 has impacted various enterprises where some employees are working from home, the need for knowledge sharing is of paramount importance during the period of the crisis. The authors further indicate that the implications brought about by the Covid-19 crisis, which, among other things, have led to changes in normal working practices and adoption of new ways of conducting businesses, have also necessitated a paradigm shift from the usual ways of transferring knowledge to digitisationto effectively link together employees who are working from home as a way of limiting physical contact at workplaces. A study by KPMG (2000) cited in Cabrera et al. (2006) further highlights the significance of ICT in managing knowledge;the study’s findingsestablished the key rolesthat the technological tools and platforms are playing in bringing employees together virtually with the possibility of exchanging information.
It should also be indicated that during the detection stage of crisis management process, there is a need for crisis response systems that can detect any potential disaster and provide early warnings for necessary interventions (Jennex and Raman, 2009). According to the authors, from the same technological systems, we can retrieve stored information regarding how similar crises were solved in the past and that information can now be utilised to makean informed decision regarding mitigation of the present crisis. This again justifies the essence of knowledge managementto mitigatea crisis through ICT.
Despite the need for technological infrastructure to capture and store knowledge that can be used to manage a crisis, Becher and Behme (2021, p.2) posit that “technology on its own is not enough to enable effective harvest of the value of knowledge management.” Based on their research findings, it has been established that the future of organisations now depends on their capability to migrate from “knowledge acquisition to knowledge creation and transfer with emphasis on establishing a knowledge sharing culture alongside the need for tools and platforms” (p.2). Davenport and Prusak (1998) cited in Cabrera et al. (2006), posited that ICT inventions for the management of knowledge are mainly customised as knowledge repositories. According to the authors, the repositories allow employees to access various information captured and stored in the database, interact, and share best practices, challenges, and other relevant information. However, many other researchers have also found several challenges associated with the knowledge management systems which limit their effectiveness in terms of knowledge sharing as they mostly depend on motivation of the knowledge holders for quality and content (Cabrera and Cabrera, 2002, 2005).
Having established the need for knowledge repositories as a way of sustaining knowledge during the Covid-19 crisis, the point of emphasis is now being put on ensuring the quality of information as well as ease of accessibility (Howard, 2021). Much as technology has been proven to play a critical role in managing knowledge during the Covid-19 crisis, the need for building a culture that embraces credible knowledge sharing still remains of paramount importance. This has led to the following proposition.
P6 ICT is positively related to knowledge sustainability during the Covid-19 crisis
Discussion
Covid-19 crisis has indeed posed a significant challenge to various knowledge workers around the globe, as evidenced by the fact that it still has no cure since its inception; it still remains vague regarding the lead time for the crisis while the virus keeps emerging in various forms. However, since the crisis has posed challenges to knowledge workers, its solutions rely on creating and sharing knowledge, as revealed through the reviewed literature. As the lasting solution for the pandemicis yet to be identified and the lead period still remains vague, fears for further curfews, which among other things, lead to the dispersion of employees from their workplaces, cannot be completely allayed. Therefore, as various stakeholders continue to work round the clock to find a lasting solution for the Covid-19 pandemic, there is also a need for continuity of operations in various enterprises. It is, therefore, evident that there is a need for knowledge to fight against the pandemic on the one hand and a need for knowledge to enable enterprises to continue operating in the midst of the pandemic on the other hand.
Based on the focused antecedents of this paper, sustainability of knowledge as a critical resource that can be used to mitigate the Covid-19 pandemic largely depends upon leadership, culture, and ICT infrastructure. Figure 2 below is a model depicting the relationships of the variables in question.Figure 2. A proposed conceptual model by the authors.
Key variables
It should be emphasised that despite the uniqueness of Covid-19 pandemic as compared to previous pandemics, organisations can still learn from previous knowledge on how such pandemics were being managed. This knowledge can either be in the form of tacit, which is acquired through experiences, or explicit, which is accessible through texts, among other sources. This, however, depends on how the organisation values knowledge as an important resource for problem-solving in general, Covid-19 in particular.
From Figure 2 above, it can be established that there is an interrelationship between the variables of leadership, culture, and ICT, and a combination of these leads to sustainable knowledge management, which in the end, helps in mitigating the crisis. Within an organisational set-up, it is incumbent upon the leadership to set the standards for the rest of the employees regarding the essence of knowledge management. When the leaders set that precedence and employees practice it, it becomes their way of doing business. During the Covid-19 pandemic, the culture of acquiring knowledge, storing it, and sharing it is essential for decision-making and day-to-day operations. Further, leadership needs to support a culture of positivity towards the utilisation of ICT as an alternative to the normal working styles.
It should also be emphasised that during the Covid-19 pandemic, leadership has a role to play in imparting a culture that embraces digitisation. This is the case because some deviations from normalcy call for quick migration from normal ways of conducting business and adopt new ways that respond to the status quo. Much as it may be appreciated that ICT has its own bottlenecks, the need to adopt it as an alternative cannot be overemphasised because failure to do so would mean a complete shutdown of operations in various enterprises. This, therefore, calls for the need for enterprises to ensure limiting the challenges associated with digitisation.
Study implications
The current study avails literature that has been carefully synthesisedso that it aids in a thorough understanding of how best knowledge can be sustained as a strategic resource during a crisis like Covid-19. Our view is that various stakeholders will find the contents of this paper valuable for decision making.
The proposed model that has been developed through this study is believed to be of special importance as it provides further clarity on the relationship between the antecedents of leadership, culture and ICT in managing knowledge as a strategic resource for crisis mitigation. The model further adds value to the already existing knowledge regarding the importance of knowledge management in managing a crisis, Covid-19 in particular.
The distinctiveness of this study does not only lie in the comprehensive exposition and analysis of the literature made but also the development of the proposed model that demonstrates the relationships of the variables under consideration.
Study limitations and conclusion
Through the study, which involved a wider review of study materials related to the management of knowledge as a strategic resource towards mitigation of Covid-19 crisis, several revelations have been made. Key among the revelation made is the proof that knowledge management is indeedan essential strategic resource for effective management of the Covid-19 crisis. Therefore, the need to sustain this strategic resource within the Covid-19 crisis period cannot be overemphasised. To this effect, it has been established through the reviewed literature that the sustainability of knowledge management during the Covid-19 crisis largely depends upon an employee-centred style of leadership, a knowledge embracing culture, and relevant information and technology infrastructure. It has further been revealed that the relevance of ICT as an antecedent of managing knowledge hinges on a culture that embraces knowledge as a critical resource. Some limitations of ICT have been discussed where the emphasis has been put on migration from knowledge capture to knowledge creation and sharing as the best option for moving forward.
The reviewed literature has led to the development of a proposed model that aims to understand further the relationships between the key determinants of leadership, culture, ICT, and knowledge management during the Covid-19 crisis. We believe that the proposed model, together with the propositions put forward in this paper, will spark a desire for further in-depth investigations that relevant statistics and quantitative studies can support. Furthermore, future research attempts can also focus on the other determinants that this study has not considered.
Author biographies
Mr Austin Milward Nyoni is currently a doctoral candidate at the Sharda School of Business Studies, Sharda University, Greater Noida. His research interests include knowledge management and strategic management in public sector organizations.
Dr Sanjay Kaushal is currently engaged as an Assistant Professor at the Sharda School of Business Studies, Sharda University, Greater Noida. He earned his PhD from Jawaharlal Nehru University and his broad area of research include general management, leadership, and discourse analysis.
ORCID iD
Sanjay Kaushal https://orcid.org/0000-0001-5938-166X
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding: The author(s) received no financial support for the research, authorship, and/or publication of this article.
==== Refs
References
Adamu AA Mohamad B Rahman NAA (2016) Antecedents of internal crisis communication and its consequences on employee performance. International Review of Management and Marketing 6 (7 ): 33–41.
Al‐Alawi AI Al‐Marzooqi NY Mohammed YF (2007) Organisational culture and knowledge sharing: critical success factors. Journal of Knowledge Management 11 (2 ): 22–42.
Alavi M Leidner DE (2001) Knowledge management and knowledge management systems: conceptual foundations and research issues. MIS quarterly 25 : 107–136.
Ammirato S Linzalone R Felicetti AM (2021) Knowledge management in pandemics. A critical literature review. Knowledge Management Research & Practice 19 (4 ): 415–426.
Baldwin R , Weder di (2020) Mauro B. Mitigating the COVID economic crisis: Act fast and do whatever it takes. London, UK: CEPR Press.
Behme F Becker S (2021) The New Knowledge Management” Available at: https://www2.deloitte.com/content/dam/insights/articles/emea103993_the-new-knowledge-management/DI_The-new-knowledge-management.pdf (accessed on 22 April 2022).
Boin A (2008) Crisis management. Los Angeles, CA: SAGE, Vol. 2 .
Bratianu C (2020) A knowledge management approach to complex crises. Management Dynamics in the Knowledge Economy 8 (4 ): 345–356.
Burnett J (2002) Managing Business Crises: From Anticipation to Implementation, Quorum Books Westport CTClark R. A. , 2016. Business continuity and the pandemic threat - Potentially the biggest survival challenge facing organizations. IT Governance Publishing.
Cabrera A Cabrera EF (2002) Knowledge-sharing dilemmas. Organisation Studies 23 (5 ): 687–710.
Cabrera EF Cabrera A (2005) Fostering knowledge sharing through people management practices. The International Journal of Human Resource Management 16 (5 ): 720–735.
Cabrera A Collins WC Salgado JF (2006) Determinants of individual engagement in knowledge sharing. The International Journal of Human Resource Management 17 (2 ): 245–264.
Carmeli A Reiter-Palmon R Ziv E (2010) Inclusive leadership and employee involvement in creative tasks in the workplace: the mediating role of psychological safety. Creativity Research Journal 22 (3 ): 250–260.
Casimir G Ng YNK Cheng CLP (2012) Using IT to share knowledge and the TRA. Journal of Knowledge Management 16 (3 ): 461–479.
Connelly CE Kelloway EK (2003) Predictors of employees’ perceptions of knowledge sharing cultures. Leadership and Organization Development Journal 24 (5 ): 294–301.
Dang CN Le-Hoai L Kim SY (2018) Impact of knowledge enabling factors on organisational effectiveness in construction companies. Journal of Knowledge Management 22 (4 ): 759–780.
D’Aveni RA MacMillan IC (1990) Crisis and the content of managerial communications: a study of the focus of attention of top managers in surviving and failing firms. Administrative Science Quarterly 35 : 634–657.
Davenport TH Prusak L (1998) Working Knowledge: How Organizations Manage What They Know. Cambridge, MA: Harvard Business School Press.
De Long DW Fahey L (2000) Diagnosing cultural barriers to knowledge management. Academy of Management Perspectives 14 (4 ): 113–127.
Du Plessis M (2007) Knowledge management: what makes complex implementations successful? Journal of Knowledge Management 11 (2 ): 91–101.
Dutton JE Duncan RB (1987) The creation of momentum for change through the process of strategic issue diagnosis. Strategic Management Journal 8 (3 ): 279–295.
Egan TM Yang B Bartlett KR (2004) The effects of organisational learning culture and job satisfaction on motivation to transfer learning and turnover intention. Human Resource Development Quarterly 15 (3 ): 279–301.
Heide M Simonsson C (2011) Putting coworkers in the limelight: new challenges for communication professionals. International Journal of Strategic Communication 5 (4 ): 201–220.
Howard M (2021) Knowledge management challenges and changes during COVID-19. Available at: https://imanage.com/blog/knowledge-management-challenges-and-changes-during-covid-19/ (Accessed on 22 April 2022).
Janz BD Prasarnphanich P (2003) Understanding the antecedents of effective knowledge management: the importance of a knowledge‐centered culture. Decision Sciences 34 (2 ): 351–384.
Jasko O Popovic N Prokic S (2012) The importance knowledge management for the improvement of crisis management. China-USA Business Review 11 (2 ).
Jennex ME Raman M (2009) Knowledge management in support of crisis response. International Journal of Information Systems for Crisis Response and Management 1 (3 ): 69–83.
Johansen W Aggerholm HK Frandsen F (2012) Entering new territory: a study of internal crisis management and crisis communication in organisations. Public Relations Review 38 (2 ): 270–279.
Kalev A (2020) Research: US unemployment rising faster for women and people of color, Harvard Business Review, available at: https://hbr.org/2020/04/research-u-s-unemployment-rising-faster-for-women-and-people-of-color (accessed 19 Jul 2022).
Kaushik M Guleria N (2020) The impact of pandemic COVID-19 in workplace. European Journal of Business and Management 12 (15 ): 1–10.
Kiesler S Sproull L (1982) Managerial response to changing environments: perspectives on problem sensing from social cognition. Administrative Science Quarterly 27 : 548–570.
Kim DG Choi SO (2018) Impact of construction IT technology convergence innovation on business performance. Sustainability 10 (11 ): 1–16.
Kniffin KM Narayanan J Anseel F , et al. (2021) COVID-19 and the workplace: implications, issues, and insights for future research and action. American Psychologist 76 (1 ): 63–77.32772537
KPMG (2000) KPMG Knowledge Management Survey 2000. Available at http://www.providersedge.com/docs/km_articles/kpmg_km_research_report_2000.pdf (accessed on 19 Jul 2022)
Lee YC Lee SK (2007) Capabilities, processes, and performance of knowledge management: a structural approach. Human Factors and Ergonomics in Manufacturing and Service Industries 17 (1 ): 21–41.
Lee Y Tao W Li JYQ , et al. (2020) Enhancing employees’ knowledge sharing through diversity-oriented leadership and strategic internal communication during the COVID-19 outbreak. Journal of Knowledge Management 25 (6 ): 1526–1549.
Linderman K Schroeder RG Zaheer S , et al. (2004) Integrating quality management practices with knowledge creation processes. Journal of Operations Management 22 (6 ): 589–607.
Luu TT Rowley C Vo TT (2019) Addressing employee diversity to foster their work engagement. Journal of Business Research 95 : 303–315.
Mahdi OR Nassar IA Almsafir MK (2019) Knowledge management processes and sustainable competitive advantage: an empirical examination in private universities. Journal of Business Research 94 : 320–334.
Mehta N (2008) Successful knowledge management implementation in global software companies. Journal of Knowledge Management 12 (2 ): 42–56.
Michailova S Minbaeva DB (2012) Organisational values and knowledge sharing in multinational corporations: the danisco case. International Business Review 21 (1 ): 59–70.
Mitroff II Shrivastava P Udwadia FE (1987) Effective crisis management. Academy of Management Perspectives 1 (4 ): 283–292.
Mittal S Kumar V (2019) Study of knowledge management models and their relevance in organisations. International Journal of Knowledge Management Studies 10 (3 ): 322–335.
Moe TL Gehbauer F Senitz S , et al. (2007) Balanced scorecard for natural disaster management projects. Disaster Prevention and Management: An International Journal 16 : 785–806.
Nembhard IM Edmondson AC (2006) Making it safe: the effects of leader inclusiveness and professional status on psychological safety and improvement efforts in health care teams. Journal of Organizational Behavior: The International Journal of Industrial, Occupational and Organisational Psychology and Behavior 27 (7 ): 941–966.
Nishii LH Mayer DM (2009) Do inclusive leaders help to reduce turnover in diverse groups? The moderating role of leader–member exchange in the diversity to turnover relationship. Journal of Applied Psychology 94 (6 ): 1412–1426.19916652
Pauchant T Mitroff I (1992) Transforming the crisis-prone organization. San Francisco: Jossey-Bass.
Pearson CM Clair JA (1998) Reframing crisis management. Academy of Management Review 23 (1 ): 59–76.
Peeri NC Shrestha N Rahman MS , et al. (2020) The SARS, MERS and novel coronavirus (COVID-19) epidemics, the newest and biggest global health threats: what lessons have we learned? International Journal of Epidemiology 49 (3 ): 717–726.32086938
Pérez‐López S Alegre J (2012) Information technology competency, knowledge processes and firm performance. Industrial Management and Data Systems 112 (4 ): 644–662.
Podrug N Filipović D Kovač M (2017) Knowledge sharing and firm innovation capability in croatian ICT companies. International Journal of Manpowe 38 (4 ): 632–644.
Rathi D Given LM Forcier E (2014) Interorganisational partnerships and knowledge sharing: the perspective of non-profit organisations (NPOs). Journal of Knowledge Management 18 (5 ): 867–885.
Razmerita L Kirchner K Nielsen P (2016) What factors influence knowledge sharing in organisations? A social dilemma perspective of social media communication. Journal of Knowledge Management 20 (6 ): 1225–1246.
Roberts KH (1990) Some characteristics of one type of high reliability organisation. Organisation Science 1 (2 ): 160–176.
Rosenbusch N. Rauch A. Bausch A. 2013. The mediating role of entrepreneurial orientation in the task environment–performance relationship: A meta-analysis. Journal of management, 39 (3 ), pp.633–659.
Sayyadi M (2019) How effective leadership of knowledge management impacts organisational performance. Business Information Review 36 (1 ): 30–38.
Schiuma G Carlucci D Lerro A , et al. (2012) Managing knowledge processes for value creation managing knowledge processes for value creation. VINE 42 (1 ): 4–14.
Schwartz HS (1987) On the psychodynamics of organisational disaster: the case of the space shuttle challenger. Columbia Journal of World Business 22 (1 ): 59–67.
Seneviratne K Amaratunga D Haigh R Pathirage C (2010) Knowledge Management for Disaster Resilience: Identification of Key Success Factors. Salford, United Kingdom: CIB World Congress, 138–150.
Shahzad K Bajwa SU Siddiqi AFI , et al. (2016) Integrating knowledge management (KM) strategies and processes to enhance organisational creativity and performance: an empirical investigation. Journal of Modelling in Management 11 (1 ): 154–179.
Shrivastava P Mitroff II Miller D , et al. (1988) Understanding industrial crises [1]. Journal of Management Studies 25 (4 ): 285–303.
Snyder H. 2019. Literature review as a research methodology: An overview and guidelines. Journal of business research, 104 , pp.333–339.
Solnit R (2020) The Impossible has Already Happened: What Coronavirus Can Teach us About Hope. The Guardian, 7 . Available at: https://www.theguardian.com/world/2020/apr/07/what-coronavirus-can-teach-us-about-hope-rebecca-solnit (accessed on 22 April 2022).
Tomé E Gromova E Hatch A (2022) Knowledge Management and COVID‐19: Technology, People and Processes. Knowledge and Process Management. Available at: https://onlinelibrary.wiley.com/doi/full/10.1002/kpm.1699 (accessed on 22 April 2022).
Torraco R.J. 2005. Writing integrative literature reviews: Guidelines and examples. Human resource development review, 4 (3 ), pp.356–367.
Wang WT Belardo S (2005) Strategic integration: a knowledge management approach to crisis management. IEEE, pp. 252a–252a.Proceedings of the 38th Annual Hawaii International Conference on System Sciences.
Wang J Peters HP Guan J (2006) Factors influencing knowledge productivity in German research groups: lessons for developing countries. Journal of Knowledge Management 10 (4 ): 113–126.
Wang C Cheng Z Yue XG , et al. (2020) Risk management of COVID-19 by universities in China. Journal of Risk and Financial Management 13 (2 ): 36.
Wee JCN Chua AYK (2013) The peculiarities of knowledge management processes in SMEs: the case of Singapore. Journal of Knowledge Management 17 (6 ): 958–972.
Weick KE (1988) Enacted sensemaking in crisis situations [1]. Journal of Management Studies 25 (4 ): 305–317.
Wooten LP James EH (2004) When firms fail to learn: the perpetuation of discrimination in the workplace. Journal of Management Inquiry 13 (1 ): 23–33.
Wooten LP James EH (2008) Linking crisis management and leadership competencies: the role of human resource development. Advances in Developing Human Resources 10 (3 ): 352–379.
Wu WW Lee YT (2007) Selecting knowledge management strategies by using the analytic network process. Expert Systems with Applications 32 (3 ): 841–847.
Yin J Ma Z Yu H , et al. (2019) Transformational leadership and employee knowledge sharing: explore the mediating roles of psychological safety and team efficacy. Journal of Knowledge Management 24 : 150–171.
| 0 | PMC9702974 | NO-CC CODE | 2022-12-01 23:19:02 | no | 2022 Dec; 39(4):136-146 | utf-8 | null | null | null | oa_other |
==== Front
Asia Pac J Public Health
Asia Pac J Public Health
APH
spaph
Asia-Pacific Journal of Public Health
1010-5395
1941-2479
SAGE Publications Sage CA: Los Angeles, CA
36420928
10.1177/10105395221139346
10.1177_10105395221139346
Original Manuscript
Association Between Changes in Family Life Due to COVID-19 and Depressive Mood and Stress Perception
Kim Min-Su MPH 12
https://orcid.org/0000-0003-1213-6952
Han Mi Ah MD, PhD 3
https://orcid.org/0000-0002-5988-6825
Park Jong MD, PhD 3
Ryu So Yeon MD, PhD 3
1 Department of Public Health, Graduate School of Health Science, Chosun University, Gwangju, Republic of Korea
2 Department of Planning Team, Chosun University Hospital, Gwangju, Republic of Korea
3 Department of Preventive Medicine, College of Medicine, Chosun University, Gwangju, Republic of Korea
Mi Ah Han, Department of Preventive Medicine, College of Medicine, Chosun University, 309 Philmum-daero, Dong-gu, Gwangju 61452, Republic of Korea. Email: [email protected]
24 11 2022
24 11 2022
10105395221139346© 2022 APJPH
2022
Asia-Pacific Academic Consortium for Public Health
This article is made available via the PMC Open Access Subset for unrestricted re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the COVID-19 pandemic or until permissions are revoked in writing. Upon expiration of these permissions, PMC is granted a perpetual license to make this article available via PMC and Europe PMC, consistent with existing copyright protections.
Coronavirus disease 2019 (COVID-19) is a new infectious disease that has had a significant impact on daily life. This study investigated the effect of changes in family life due to COVID-19 on depressive mood and stress perception. We used data from the “Survey on changes in family life due to COVID-19” in Korea. The final study population comprised 1500 adults with children aged ≤19 years. Of the total respondents, 59.3% responded that depressive mood and stress had increased due to COVID-19. However, among them, 46.6% did not attempt to resolve or could not find a way to resolve their depressive mood and stress. In multiple logistic regression analyses, a decrease in household income and increased household expenditure due to COVID-19 were significantly associated with an increased risk of depressive mood and stress perception. Depressive mood and stress were significantly higher in respondents who answered that they had experienced family conflicts. These results could be used to assess changes in family life and manage mental health when a new infectious disease occurs. Therefore, it is necessary to assess the long-term effects of changes in family life due to COVID-19 on mental health.
child care
COVID-19
depression
family relations
psychological stress
edited-statecorrected-proof
typesetterts1
==== Body
pmcWhat We Already Know
COVID-19 is a new infectious disease that has had a significant impact on daily life.
The mental health of community members has been negatively affected by the experience of the COVID-19 outbreak.
Data on the effects of changes in family life due to COVID-19 on mental health in Korea is limited.
What This Article Adds
We provided the evidence of change in family life and mood or stress due to COVID-19.
Of the total participants, 59.3% responded that depressive mood and stress had increased due to COVID-19.
Most participants did not have an appropriate way to deal with increased depressive mood and stress.
Introduction
Coronavirus disease 2019 (COVID-19) has become a significant health burden since it was first detected in Korea in January 2020. As of January 2022, there were 53.08 million confirmed cases and 6.31 million deaths worldwide. In Korea, as of August 2022, 20 million confirmed cases and 25 000 deaths have been recorded.1
To prevent the occurrence and spread of COVID-19, infection control strategies, including social distancing, isolation, and restrictions on facility use, have been implemented and have caused many changes in daily and family life. Telecommuting and online learning at home owing to restrictions on going to work and school closures have been introduced.2 Moreover, because the care provided by childcare institutions was limited during the pandemic, the need to take care of children or help with education at home increased. Therefore, as the time spent with the family increased, changes in family life occurred, which improved family relationships, such as increased communication between families, but also had a negative effect on deepening family conflicts.3
In addition, as the new term “Corona Blue” suggests, the mental health of community members has been negatively affected by the experience of the COVID-19 outbreak, the continued feeling of isolation due to social distancing, and the deepening of economic difficulties.4 COVID-19-related concerns and anxiety, such as fear of infection, fear of being criticized or harmed by those around you when infected, fear that someone in the family may be infected, and fear of economic damage, might negatively affect mental health.5
Therefore, previous studies have focused on changes in family life and their effects on mental health during the COVID-19 outbreak. A study examining the influence of the pandemic on family relationships reported that those who reported negative changes in their family relationships also reported worse mental health.6 Youth mental health problems were significantly associated with parental psychological symptoms during the COVID-19 pandemic,7 and family support was positively associated with life satisfaction.8 COVID-19-related stressors and the perceived negative effects of the pandemic on family life increased the risk of depression and anxiety, while more family health resources reduced the risk of depression and anxiety symptoms.9 These results suggest the need to consider the impact of family life on mental health during a new outbreak of infectious disease.
However, there are limited data on the effects of family characteristics or changes in family life due to COVID-19 on mental health in Korea. Therefore, the purpose of this study was to investigate the prevalence of depression and stress perception due to COVID-19 and to investigate related factors.
Methods
Data Source and Study Population
This study used the data of the “Survey on changes in family life due to COVID-19.” The survey aimed to identify changes in family life due to COVID-19 for the Ministry of Gender Equality and Family to establish family policies in Korea.10 Data were collected by Embrain, a specialized social research institute (https://embrain.com/).
The survey population comprised 1500 adults with children under the high school age. The source population was based on population statistics from the resident registration of May 2020, and income level and gender were reflected in the sample design. According to the household trend survey by the National Statistical Office, income classes constituted 20% of the first quartile, 25% of the second quartile, 25% of the third quartile, and 30% of the fourth quartile; regarding sex, there were 600 men and 900 women.
Embrain asks consent for collection of personal information when recruiting panels and collects personal information such as gender, age, education level, occupation, residence area, marital status, family structure, and income. Based on the information provided when panels registered, male and female panel members with children were randomly invited to participate in the survey through text messages, emails, KakaoTalk, and website advertisements.
Only one family member was allowed to participate in the survey. To determine eligibility, all children living together and the number of children in each age group were entered. A total of 4068 agreed and participated in the survey. Among them, those who did not complete the survey have been excluded, and the survey was discontinued if there were children who did not live together or if there were only adult children. The survey period was from June 1 to 7, 2020, and data were collected through online survey. The survey was conducted until all 1500 adults had completed their questionnaire by setting income level and gender in advance.
Ethical approval of the survey could not be obtained due to the urgency and time-sensitive nature of the survey. However, the survey was at a minimal risk or less based on self-reported online survey by voluntary participation. This study was based on publicly available deidentified data and would not require ethical approval. All participants were informed about the survey, and informed consent was obtained for all study participants.
General Characteristics
General characteristics included gender (male, female), age (20-39, 40-49, ≥50 years), respondent’s academic background (≤high school, university, graduate school), type of residence (house, apartment, etc), type of home ownership (owned house, lease house), marital status (with spouse, without spouse), number of children (one, two, three or more), level of first-child education (preschool, elementary school, middle school, high school), and household income (first, second, third, or fourth quartile).
Changes in Family Life Due to COVID-19
Changes in family life due to COVID-19 included primary caregiver for children before COVID-19 (myself, spouse, both equally provide care, without spouse), changes in child care after COVID-19 (more than before, no change, spouse does it more, without spouse), use of family care leave to care for children (used, did not use), household income change due to COVID-19 (decreased, no change, increased), household expenditure change due to COVID-19 (decreased, no change, increased), time spent at home due to COVID-19 (decreased, no change, increased), leisure activity time change due to COVID-19 (decreased, no change, increased), change in the number of family who eat together due to COVID-19 (decreased, no change, increased), and family conflicts due to COVID-19 (yes or no).
Family conflict was defined as a yes/no answer to the following question: “After the spread of COVID-19, do you think that family members experience conflict with each other more often than before?” For those experiencing family conflicts, family members of the conflict (spouse, preschool children, elementary school children, middle school children, high school children, other family members, etc) and the main reason (housework such as preparing meals, compliance with hygiene rules to prevent infection such as washing hands after going out, disagreements regarding going out for family leisure, differences of opinion about leisure activities, individual lifestyles, childcare and care-sharing issues, and economic problems) were investigated.
Depressive Mood and Stress Perception Due to COVID-19
Changes in depressive mood and stress due to COVID-19 were assessed using the following question: “Do you think your depressive mood or stress has increased after the spread of COVID-19 compared to before?” For those who responded yes, the resolution methods for depressive mood or stress were assessed with the question, “How did you resolve family conflict, depressive mood, and stress after COVID-19?” The response options were consultations with public institutions such as the national and local governments, consultations with private organizations such as private institutions and companies, resolution within the family such as through family conversations, resolution with advice from friends, and did not try to solve or could not find a way to resolve.
Analysis
The collected data were analyzed using the statistical analysis program SPSS 26.0. The status of depressive mood and stress perception due to COVID-19 and the general characteristics of the study population were presented as frequency and percentage. χ2 tests were performed to determine the proportions of change in depressive mood and stress according to the participants’ general characteristics and the characteristics of changes in family life due to COVID-19. Finally, a binary multiple logistic regression analysis was performed to determine the effects of changes in family life related to depressive mood and increased stress.
Results
Depressive Mood and Stress Perception Due to COVID-19
After the spread of COVID-19, 59.3% of participants perceived that their depressive mood and stress had increased (Table 1). Among them, 46.6% answered that they did not attempt to resolve or could not find a solution to resolve their depressive mood and stress, 42.6% answered that they resolved them within the family, and 8.5% said that they solved them through advice from friends (Supplementary Table 1).
Table 1. Depressive Mood and Stress Perception by General Characteristics of the Subject Due to COVID-19.
Variables Category Total Depressive mood and stress perception due to COVID-19 P value
Yes No
Total 1500 890 (59.3) 610 (40.7)
Gender Male 600 (40.0) 296 (49.3) 304 (50.7) <.001
Female 900 (60.0) 594 (66.0) 306 (34.0)
Age 20-39 541 (36.1) 329 (60.8) 212 (39.2) .249
40-49 777 (51.8) 463 (59.6) 314 (40.4)
≥50 182 (12.1) 98 (53.8) 84 (46.2)
Respondent’s academic background ≤High school 260 (17.3) 122 (46.9) 138 (53.1) <.001
University 1033 (68.9) 627 (60.7) 406 (39.3)
Graduate school 207 (13.8) 141 (68.1) 66 (31.9)
The type of residence House 106 (7.1) 55 (51.9) 51 (48.1) .142
Apartment 1113 (74.2) 675 (60.6) 438 (39.4)
Other 281 (18.7) 160 (56.9) 121 (43.1)
Type of home ownership Owned house 965 (64.3) 557 (57.7) 408 (42.3) .088
Lease house 535 (35.7) 333 (62.2) 202 (37.8)
Marital status With spouse 1466 (97.7) 870 (59.3) 596 (40.7) .951
Without spouse 34 (2.3) 20 (58.8) 14 (41.2)
Number of children 1 569 (37.9) 334 (58.7) 235 (41.3) .882
2 774 (51.6) 464 (59.9) 310 (40.1)
≥3 157 (10.5) 92 (58.6) 65 (41.4)
Level of first-child education Preschool 616 (41.1) 373 (60.6) 243 (39.4) .025
Elementary school 275 (18.3) 179 (65.1) 96 (34.9)
Middle school 234 (15.6) 137 (58.5) 97 (41.5)
High school 375 (25.0) 201 (53.6) 174 (46.4)
Household income First quartile 300 (20.0) 180 (60.0) 120 (40.0) .983
Second quartile 375 (25.0) 222 (59.2) 153 (40.8)
Third quartile 375 (25.0) 224 (59.7) 151 (40.3)
Fourth quartile 450 (30.0) 264 (58.7) 186 (41.3)
Data were expressed as number (%).
Abbreviation: COVID-19, Coronavirus disease 2019.
Depressive Mood and Stress Perception by General Characteristics
Of the total number of participants, 40% were male and 60% were female; regarding age, 36.1% were aged 20-39 years, 51.8% were aged 40-49 years, and 12.1% were aged 50 years or older. Regarding the number of children, 51.6% had two children, 37.9% had one child, and 10.5% had three or more children. For the level of first-child education, 41.1% were preschoolers, 25.0% were high school students, 18.3% were elementary school students, and 15.6% were middle school students. The prevalence of depressive mood and stress was 66.0% and 49.3% in women and men, respectively (P < .001). The perception of depressive mood or stress according to children’s school age level was 60.6% for those with preschool children, 65.1% for those with elementary school students, 58.5% for those with middle school students, and 53.6% for those with high school students (P = .025) (Table 1).
Depressive Mood and Stress Perception by Characteristics of Changes in Family Life Due to COVID-19
Regarding changes in child care after the COVID-19 outbreak, 31.3% responded that they had more responsibilities than before, and 15.3% said that the burden on spouses had increased. Moreover, 46.7% responded that household income decreased, and 32.9% responded that household expenditure increased due to COVID-19. After the spread of COVID-19, 75.1% responded that the time spent at home with their families increased, and 37.4% experienced conflict between family members after the COVID-19 outbreak. Among family members, the majority of participants who experienced conflict were their spouses (39.7%) and elementary school children (19.1%). The most common causes of family conflict were an increase in housework (8%), lifestyle problems (21.7%), differences in opinion about leisure activities (14.1%), childcare and care-sharing issues (11.4%), disagreements regarding going out for family leisure (8.4%), problems related to compliance with hygiene rules for infection prevention (8.0%), conflicts due to economic problems (8.0%), and others (0.6%). The perception rate of depressive mood or stress was 71.1% in those caring for children more than before COVID-19, 54.3% in those without change, and 52.2% in those who responded that the spouse had more childcare duties. The prevalence of depressive mood and stress according to changes in the economic environment due to COVID-19 was 69.6% in those with decreased household income, 50.2% in those without change, and 56.5 in those with increased household income. Regarding household expenditure, the prevalence was 64.1% in those with decreased household expenditure, 46.8% in those without change, and 68.4% in those with increased household expenditure. Regarding family conflict experience, the prevalence was 87.0% in those with conflicts and 42.8% in those without conflicts (Table 2).
Table 2. Depressive Mood and Stress Perception by Characteristics of Changes in Family Life Due to COVID-19.
Variables Category Total Depressive mood and stress perception due to COVID-19 P value
Yes No
Primary caregiver for children before COVID-19 Myself 715 (47.7) 485 (67.8) 230 (32.2) <.001
Spouse 396 (26.4) 193 (48.7) 203 (51.3)
Both equally provide care 355 (23.6) 192 (54.1) 163 (45.9)
Without spouse 34 (2.3) 20 (58.8) 14 (41.2)
Changes in child care after COVID-19 More than before 470 (31.3) 334 (71.1) 136 (28.9) <.001
No change 766 (51.1) 416 (54.3) 350 (45.7)
Spouse does it more 230 (15.3) 120 (52.2) 110 (47.8)
Without spouse 34 (2.3) 20 (58.8) 14 (41.2)
Use of family care leave to care for children Used 125 (8.3) 76 (60.8) 49 (39.2) .727
Did not use 1375 (91.7) 814 (59.2) 561 (40.8)
Household income change Decreased 700 (46.7) 487 (69.6) 213 (30.4) <.001
No change 777 (51.8) 390 (50.2) 387 (49.8)
Increased 23 (1.5) 13 (56.5) 10 (43.5)
Household expenditure change Decreased 468 (31.2) 300 (64.1) 168 (35.9) <.001
No change 538 (35.9) 252 (46.8) 286 (53.2)
Increased 494 (32.9) 338 (68.4) 156 (31.6)
Time spent at home Decreased 19 (1.3) 10 (52.6) 9 (47.4) <.001
No change 354 (23.6) 157 (44.4) 197 (55.6)
Increased 1127 (75.1) 723 (64.2) 404 (35.8)
Leisure activity time change Decreased 463 (30.9) 315 (68.0) 148 (32.0) <.001
No change 565 (37.7) 296 (52.4) 269 (47.6)
Increased 472 (31.5) 279 (59.1) 193 (40.9)
Change in the number of family who eat together Decreased 43 (2.9) 33 (76.7) 10 (23.3) <.001
No change 450 (30.0) 210 (46.7) 240 (53.3)
Increased 1007 (67.1) 647 (64.3) 360 (35.7)
Family conflict experience Yes 561 (37.4) 488 (87.0) 73 (13.0) <.001
No 939 (62.6) 402 (42.8) 537 (57.2)
Data were expressed as number (%).
Abbreviation: COVID-19, Coronavirus disease 2019.
Factors Related to Increased Depressive Mood and Stress Perception Due to COVID-19
In the multiple regression analysis, the odds ratio (OR) for depressive mood and stress perception was significantly higher in those at a graduate school education level than in those at a high school education level (OR = 2.72, 95% CI [1.74, 4.26]). Decreased household income (OR = 2.01, 95% CI [1.54, 2.62]) was associated with the risk of depressive mood and stress. In the case of household expenditure, both participants with decreased (OR = 1.48, 95% CI [1.09, 2.01]) and increased (OR = 1.73, 95% CI [1.28, 2.34]) household expenditure showed a higher OR for depressive mood and stress perception than did those with no change. Family conflict was significantly associated with a higher risk of depressive mood and stress perception (OR = 7.95, 95% CI [5.92, 10.68]) (Table 3).
Table 3. Factors Related to Increased Depressive Mood and Stress Perception Due to COVID-19.
Variables Category OR 95% CI
Gender (/male) Female 1.38 [0.93, 2.04]
Age (/≥50) 40-49 1.28 [0.84, 1.96]
20-39 1.09 [0.74, 1.62]
Respondent’s academic background (/≤high school) University 1.88 [1.35, 2.61]
Graduate school 2.72 [1.74, 4.26]
The type of residence (/house) Apartment 1.42 [0.89, 2.28]
Other 1.07 [0.63, 1.83]
Type of home ownership (/owned house) Lease house 1.18 [0.91, 1.53]
Marital status (/with spouse) Without spouse 1.20 [0.47, 3.10]
Number of children (/≥3) 2 1.30 [0.84, 2.02]
1 1.28 [0.84, 1.95]
Household income (/first quartile) Second quartile 1.04 [0.71, 1.52]
Third quartile 1.10 [0.75, 1.61]
Fourth quartile 1.13 [0.78, 1.64]
Primary caregiver for children before COVID-19 (/spouse) Myself 1.53 [0.96, 2.44]
Take care of each other equally 1.05 [0.71, 1.53]
Without spouse 0.82 [0.32, 2.12]
Changes in child care after COVID-19 (/no change) More than before 1.15 [0.86, 1.55]
Spouse does it more 0.97 [0.67, 1.40]
Without spouse 0.82 [0.32, 2.12]
Use of family care leave (/do not use) Use 1.13 [0.72, 1.78]
Household income change (/no change) Decreased 2.01 [1.54, 2.62]
Increased 1.09 [0.39, 2.99]
Household expenditure change (/no change) Decreased 1.48 [1.09, 2.01]
Increased 1.73 [1.28, 2.34]
Time spent at home (/no change) Decreased 1.43 [1.01, 2.02]
Increased 1.21 [0.78, 1.89]
Leisure activity time change (/no change) Decreased 1.20 [0.91, 1.57]
Increased 0.93 [0.46, 1.86]
Change in the number of family who eat together (/no change) Decreased 1.27 [0.92, 1.75]
Increased 1.59 [1.00, 2.51]
Family conflict experience (/no) Yes 7.95 [5.92, 10.68]
Abbreviations: COVID-19, Coronavirus disease 2019; OR, odds ratio.
Discussion
Various strategies, such as social distancing and contact tracing, were implemented to prevent the occurrence and spread of COVID-19. The outbreak of a new infectious disease is known to cause various psychological problems, such as depression and anxiety, not only in patients but also in the entire community. In addition, it can cause various changes in family life, such as increased childcare burden and family conflicts. Therefore, this study aimed to investigate the effects of changes in family life due to COVID-19 on depressive mood and stress.
In this study, 59.3% of participants responded that they experienced an increase in depressive mood and stress caused by COVID-19. This is similar to previous systematic reviews addressing the prevalence of mental health problems during the COVID-19 pandemic. In the United Kingdom, the prevalence of anxiety increased by 26.35%, from 4.65% in the prepandemic period to 31% in the pandemic period. Similarly, while the prevalence of prepandemic depression was 4.12%, that during the pandemic was 32.0%, indicating a 27.88% increase.11 In China, the prevalence of depressive symptoms was 27%, 26%, and 61% in COVID-19, general illness, and chronic illness patients, respectively, and the prevalence of anxiety symptoms was 14%, 23%, and 85%, respectively.12 The prevalence of depression was 14.6%, and depression was more prevalent among health care workers than among non–health care workers in Vietnam.13
These results suggest that COVID-19 increased the prevalence of mental health problems among the general population compared with that in the prepandemic period. Because of the prolonged COVID-19 outbreak, communities might suffer significant long-term mental health consequences, and policymakers and mental health services would need to make efforts to monitor mental health and provide interventions to support those in need.
Depressive mood and stress were significantly higher when household income decreased and household expenditure increased than when household income did not change. These results are consistent with those of previous studies that showed that those with low economic status were more likely to be at risk of depression. A huge decrease in income and low current income during COVID-19 are significantly associated with more anxiety/depression symptoms among non–health care workers.14 Families with unstable incomes are more likely to experience severe anxiety during the COVID-19 pandemic.15 Because economic level is a well-known risk factor for mental health,16-18 when the economic situation worsens due to COVID-19, psychological and mental health problems may follow.
Experience of family conflict was significantly associated with increased depressive mood and stress due to COVID-19. Moreover, 75.1% of participants indicated that the time spent with family at home increased after the spread of COVID-19, and 37.4% responded that conflicts among family members increased after the COVID-19 outbreak. The reasons for family conflict were the increase in household work (27.8%), followed by lifestyle (21.7%) and differences in opinions on leisure activities (14.1%). However, among those experiencing family conflicts, depressive mood, and stress due to COVID-19, 46.6% did not attempt to resolve or could not find a way to resolve them, and 42.6% resolved them within the family, such as through family conversations. An increase in the time spent at home due to the prolonged COVID-19 outbreak would increase conflict among family members, but there are often no adequate ways to relieve them. This suggests that the outbreak of new infectious diseases, such as COVID-19, or the risk thereof could cause stress, depression, and anxiety in the home. Therefore, in the event of an outbreak of a new infectious disease, it is necessary not only to manage the outbreak or spread of the infectious disease but also to strengthen the psychological counseling approach to depression and stress in household members. There is a need for psychological support through efforts and measures to promote counseling in schools, educational institutions, and workplace counseling centers for office workers.
This study has several limitations. First, since we used data collected in June 2020, the evolving government infection control policy and fatigue caused by the prolonged COVID-19 pandemic might not be reflected in this study. Variants of COVID-19, such as Delta and Omicron, are discovered, and COVID-19 is continuously spreading. Future studies are needed to evaluate the relationship between changes in family life and mental health, reflecting the prolonged COVID-19 pandemic. Second, when investigating changes in family life due to COVID-19, external factors such as the community environment for emergency care, education, child care, and workplace flexibility should also be considered. There may be differences in emergency care programs and financial support depending on the situation of local governments. Similarly, there may be differences in workplace COVID-19 quarantine guidelines, such as telecommuting and work-time flexibility. These factors should be included in future studies investigating changes in family life due to new infectious diseases. Third, this survey used the data of the “Survey on changes in family life due to COVID-19.” The survey did not include depressive symptoms or history of depression before the COVID-19 outbreak. In addition, COVID-19-related factors did not include a history of COVID-19 infection, experience of being classified as a contact person, and self-quarantine experience. Finally, there was no validated questionnaire because it was a new topic of changes in family life due to COVID-19 when the survey was conducted, and the developed questionnaire was not validated due to the urgency and time constraints of evaluation. Because depressive mood was assessed relying on a single question, further studies would be needed to assess its association with clinically important depression.
Conclusion
This study investigated various factors that may cause depressive mood or stress due to COVID-19, including general characteristics and family life changes. According to the results of this study, after the spread of COVID-19, the proportion of increase in depressive mood and stress was high; however, most participants did not have an appropriate way to resolve them. The general characteristics of the study population and various aspects of family life changes due to COVID-19 were related to depressive mood and stress. The results of this study can be used as evidence for changes in family life and mental health management in the event of a new infectious disease.
Supplemental Material
sj-docx-1-aph-10.1177_10105395221139346 – Supplemental material for Association Between Changes in Family Life Due to COVID-19 and Depressive Mood and Stress Perception
Click here for additional data file.
Supplemental material, sj-docx-1-aph-10.1177_10105395221139346 for Association Between Changes in Family Life Due to COVID-19 and Depressive Mood and Stress Perception by Min-Su Kim, Mi Ah Han, Jong Park and So Yeon Ryu in Asia Pacific Journal of Public Health
Authors’ Note: This article is a condensed form of the first author’s master’s thesis from Chosun University.
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding: The author(s) received no financial support for the research, authorship, and/or publication of this article.
ORCID iDs: Mi Ah Han https://orcid.org/0000-0003-1213-6952
Jong Park https://orcid.org/0000-0002-5988-6825
Supplemental Material: Supplemental material for this article is available online.
==== Refs
References
1 Korea Disease Control and Prevention Agency. Coronavirus (COVID-19), Republic of Korea. Accessed August 3, 2022. https://ncov.kdca.go.kr/
2 Kim EA. Social distancing and public health guidelines at workplaces in Korea: responses to coronavirus disease-19. Saf Health Work. 2020;11 (3 ):275-283. doi:10.1016/j.shaw.2020.07.006 32837740
3 Lee J Chin M Sung M. How has COVID-19 changed family life and well-being in Korea? J Comp Fam Stud. 2020;51 (3-4 ):301-313.
4 Shin YJ Lee JY. South Korea’s proactive approach to the COVID-19 global crisis. Psychol Trauma. 2020;12 (5 ):475-477. doi:10.1037/tra0000651 32551765
5 Kim W Ju YJ Lee SY. Does having various types of fear related to COVID-19 disrupt individuals’ daily life? Findings from a nationwide survey in Korea. Epidemiol Health. 2022;44 :e2022004. doi:10.4178/epih.e2022004
6 Volpert-Esmond HI Marquez ED Camacho AA. Family relationships and familism among Mexican Americans on the U.S.-Mexico border during the COVID-19 pandemic. Cultur Divers Ethnic Minor Psychol. Published online April 21, 2022. doi:10.1037/cdp0000547
7 Bai Y Liu X Zhang B , et al . Associations of youth mental health, parental psychological distress, and family relationships during the COVID-19 outbreak in China. BMC Psychiatry. 2022;22 (1 ):275. doi:10.1186/s12888-022-03938-8 35439972
8 Rodriguez-Rivas ME Varela JJ González C Chuecas MJ. The role of family support and conflict in cyberbullying and subjective well-being among Chilean adolescents during the Covid-19 period. Heliyon. 2022;8 (4 ):e09243. doi:10.1016/j.heliyon.2022.e09243
9 Crandall A Daines C Hanson CL Barnes MD. The effects of COVID-19 stressors and family life on anxiety and depression one-year into the COVID-19 pandemic. Fam Process. Published online March 29, 2022. doi:10.1111/famp.12771
10 Korea Ministry of Gender Equality and Family, Korean Women’s Development Institute. Changes in families and policy tasks due to COVID-19. Seoul: Korean Women’s Development Institute; 2020.
11 Dettmann LM Adams S Taylor G. Investigating the prevalence of anxiety and depression during the first COVID-19 lockdown in the United Kingdom: systematic review and meta-analyses. Br J Clin Psychol. 2022;61 :757-780. doi:10.1111/bjc.12360 35137427
12 Yan Y Du X Lai L Ren Z Li H. Prevalence of depressive and anxiety symptoms among Chinese older adults during the COVID-19 pandemic: a systematic review and meta-analysis. J Geriatr Psychiatry Neurol. 2022;35 (2 ):182-195. doi:10.1177/08919887221078556 35245999
13 Tran QD Vu TQC Phan NQ. Depression prevalence in Vietnam during the Covid-19 pandemic: a systematic review and meta-analysis. Ethics Med Public Health. 2022;23 :100806. doi:10.1016/j.jemep.2022.100806 35676918
14 Lee PMY Wang D Li Y , et al . Association of current income and reduction in income during the COVID-19 pandemic with anxiety and depression among non-healthcare workers. J Ment Health. 2022;31 :585-596. doi:10.1080/09638237.2022.2091768 35786138
15 Cao W Fang Z Hou G , et al . The psychological impact of the COVID-19 epidemic on college students in China. Psychiatry Res. 2020;287 :112934. doi:10.1016/j.psychres.2020.112934 32229390
16 Muntaner C Eaton WW Miech R O’Campo P. Socioeconomic position and major mental disorders. Epidemiol Rev. 2004;26 :53-62. doi:10.1093/epirev/mxh001 15234947
17 Reiss F. Socioeconomic inequalities and mental health problems in children and adolescents: a systematic review. Soc Sci Med. 2013;90 :24-31. doi:10.1016/j.socscimed.2013.04.026 23746605
18 Li Z Page A Martin G Taylor R. Attributable risk of psychiatric and socio-economic factors for suicide from individual-level, population-based studies: a systematic review. Soc Sci Med. 2011;72 (4 ):608-616. doi:10.1016/j.socscimed.2010.11.008 21211874
| 36420928 | PMC9703011 | NO-CC CODE | 2022-11-29 23:21:05 | no | Asia Pac J Public Health. 2022 Nov 24;:10105395221139346 | utf-8 | Asia Pac J Public Health | 2,022 | 10.1177/10105395221139346 | oa_other |
==== Front
Health Promot Pract
Health Promot Pract
HPP
sphpp
Health Promotion Practice
1524-8399
1552-6372
SAGE Publications Sage CA: Los Angeles, CA
36419256
10.1177/15248399221132581
10.1177_15248399221132581
Article
The Role of Anti-Racist Community-Partnered Praxis in Implementing Restorative Circles Within Marginalized Communities in Southern California During the COVID-19 Pandemic
https://orcid.org/0000-0003-3618-449X
Adkins-Jackson Paris B. PhD, MPH 1
Vázquez Evelyn PhD 2
Henry-Ala Frank K. MPH 3
Ison Juliana M. BA 4
Cheney Ann PhD 2
Akingbulu Josephine MPH 3
Starks Christian MPA, BS 5
Slay Lindsay MSW 6
Dorsey Alexander MA, LPCC 7
Marmolejo Connie DrPH, MPH 2
Stafford Alvin 8
Wen James 9
McCauley Margaret H. 9
Summers Latrese 10
Bermudez Llendy 8
Cruz-Roman Zitlaly L. MSW 11
Castillo Itzel 8
Kipke Michele D. PhD 5
Brown Arleen F. MD, MPH 12
The STOP COVID-19 CA Vaccine Hesitancy Workgroup
1 Columbia University, New York, NY, USA
2 University of California, Riverside, Riverside, CA, USA
3 Claremont Graduate University, Claremont, CA, USA
4 Massachusetts General Hospital, Boston, MA, USA
5 University of Southern California, Los Angeles, CA, USA
6 Children’s Hospital Los Angeles, Los Angeles, CA, USA
7 Mending Minds Professional Clinical Counseling, Inc, Los Angeles, CA, USA
8 Independent Researcher
9 St. John’s Cathedral, Los Angeles, CA, USA
10 St. John’s Well Child and Family Center, Inc., Los Angeles, CA, USA
11 California State University, San Bernardino, San Bernardino, CA, USA
12 University of California, Los Angeles, Los Angeles, CA, USA
Paris B. Adkins-Jackson, PhD, MPH, Departments of Epidemiology & Sociomedical Sciences, Mailman School of Public Health, Columbia University, 722 W. 168th Street, New York, NY 10032, USA; e-mail: [email protected]
23 11 2022
23 11 2022
15248399221132581© 2022 Society for Public Health Education
2022
Society for Public Health Education
This article is made available via the PMC Open Access Subset for unrestricted re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the COVID-19 pandemic or until permissions are revoked in writing. Upon expiration of these permissions, PMC is granted a perpetual license to make this article available via PMC and Europe PMC, consistent with existing copyright protections.
The COVID-19 pandemic has exacerbated the adverse influence of structural racism and discrimination experienced by historically marginalized communities (e.g., Black, Latino/a/x, Indigenous, and transgender people). Structural racism contributes to trauma-induced health behaviors, increasing exposure to COVID-19 and restricting access to testing and vaccination. This intersection of multiple disadvantages has a negative impact on the mental health of these communities, and interventions addressing collective healing are needed in general and in the context of the COVID-19 pandemic. The Share, Trust, Organize, and Partner COVID-19 California Alliance (STOP COVID-19 CA), a statewide collaborative of 11 universities and 75 community partners, includes several workgroups to address gaps in COVID-19 information, vaccine trial participation, and access. One of these workgroups, the Vaccine Hesitancy Workgroup, adopted an anti-racist community-partnered praxis to implement restorative circles in historically marginalized communities to facilitate collective healing due to structural racism and the COVID-19 pandemic. The project resulted in the development of a multilevel pre-intervention restorative process to build or strengthen community–institutional partnerships when procurement of funds has been sought prior to community partnership. This article discusses this workgroup’s role in advancing health justice by providing a community-based mental health intervention to marginalized communities in Southern California while using an antiracist praxis tool to develop a successful community–institutional partnership and to live up to the vision of community-based participatory research.
community-based
mental health
restorative
antiracism
CBPR
partnerships
community–academic partnerships
intervention planning
edited-statecorrected-proof
typesetterts1
==== Body
pmcCommunity-based participatory research (CBPR) is an approach that facilitates conducting research and partnerships between academic research institutions and community-based organizations and entities (CBOEs). The aim of this approach is to develop and implement collaborative interventions that directly address the salient needs of communities (Israel et al., 1998; Schulz et al., 1998; Zimmerman, 2020), with emphasis on those affected by health disparities (National Institute on Minority Health and Health Disparities, 2018). Community–academic partnerships benefit both partners by increasing funding for CBOEs to implement innovative multisector programs that bring diversity and inclusion to academic research institutions. In academic research institutions, such diversity and inclusion help to make scientific research by those institutions more robust, offer students connections to community through service-learning programs that are grounded in real-world knowledge, and actively contribute to the improvement of local and national social conditions (Zimmerman, 2020).
Despite the clear principles of CBPR—community engagement, partnership, action, and change—there continue to be pitfalls in the implementation of this approach that further perpetuate structural and institutional racism (Adkins-Jackson et al., 2022). Structural racism is produced by systems of oppression that discriminate against racialized populations with the goal of maintaining white supremacy (Bailey et al., 2017; Gee & Hicken, 2021). Institutions, like academic research institutions, reproduce structural racism via discriminatory policies and procedures toward marginalized individuals and communities (Adkins-Jackson et al., 2021). Although CBPR, in its purest form, engages community and academic partners in shared decision making, resource allocation, and power distribution (Minkler & Wallerstein, 2008; Van de Sande & Schwartz, 2017), the application of this approach often falls short in addressing the inequitable distribution of power and resources among community–academic partnerships. Traditionally, power (e.g., decision-making) and resources (e.g., grant funds) are consolidated within academic research institutions—and other power-holding institutions (e.g., government)—as agencies often fund research projects where the principal investigator (PI) is from a scientific institution (Heaney et al., 2007). Although CBPR’s approach calls for shared partnership, it does not prohibit the power hoarding that can occur when institutional partners are imbued with decision-making authority.
There are alternative university-managed research models like the Community-Owned and -Managed Research (COMR) Model that was developed by West End Revitalization Association, a CBOE, to address this specific issue (Heaney et al., 2007). The COMR model necessitates that the PI of an award be the community partner to facilitate the CBOE’s authority on the project relating to decision-making, project management, and data ownership. The model also encourages long-term commitment to solving health justice issues, another implementation pitfall that reveals how structural and institutional racism assign power to institutional partners in ways that undermine CBPR approaches.
Despite the innovation of the dynamic COMR model and growing literature on CBPR, disproportionate distribution of power and resources between community–academic partners persists. CBOEs are often included in research late in the process such as once research questions and agendas have been established (Adkins-Jackson et al., 2022). Some academics using CBPR reproduce structural racism through emphasizing end products like scientific publications that do not directly benefit community partners. Most dangerously, structural racism is (re)produced when partnerships dissolve when funding for the institutional partners ends. While publishing in scientific journals can be fruitful for securing future funding for the partnership, these actions and end products do not readily provide benefit to the CBOE or community at-large. Moreover, academic institutions reproduce institutional racism when they do not recognize nor establish institutional mechanisms to compensate community members from marginalized communities (e.g., undocumented, low-income, people from rural backgrounds).
Part of the allure of CBPR, COMR, and similar models is that such approaches are ideal to address health disparities in marginalized communities with limited resources. Given institutional partners are equipped with sources of funding, personnel, and access to research infrastructure, there is a strategic opportunity to assist disenfranchised community partners with large catchment areas, while addressing salient health concerns. Yet, too often, structural and institutional racism shapes the relationship by viewing community partners as objects rather than partners and investigators co-producing knowledge in research (Ahmed et al., 2004). It is clear that CBPR models like COMR can address inequitable distribution of power and resources throughout the research process. However, not all partnerships begin as COMR describes. The COMR model is ideal to implement before the research process has begun as it entails building partnerships with CBOEs in ways that result in pursuing a research project, collaborative intervention, and/or long-term commitment with CBOEs and the community at-large.
We, as community–academic partners, agree with the intentions of CBPR and the practice of COMR. Engagement of community throughout the research process, from development of research questions, study design, to proposal submissions to project implementation and dissemination, is fundamental. In this article, we put forth a model to engage when funding for a pre-existing project has already been procured. Our approach can be used with established collaborations or new partnerships that come out of a community need for collective healing. Our model tasked community–academic partnerships with: (a) working together toward equitable partnership and establishing (or restoring) trust by implementing a project together; and (b) implementing a community-based mental health intervention (CBMHIs; i.e., restorative circle) to provide a direct service to a historically marginalized community hard hit by structural racism and resulting syndemics during the COVID-19 pandemic (Gravlee, 2020; Mendenhall et al., 2021). These components subsequently encompass a multilevel pre-intervention restorative process.
The multilevel pre-intervention restorative process reduces the time from implementation of research to direct benefit to the community by centering community needs and emphasizing healing and restoration at multiple levels of injury. We engaged CBPR to implement this restorative program. To intervene on structural racism in COVID-19-related health disparities and to achieve CBPR’s vision, we adopted an anti-racist praxis as a tool to engage in reflexive relational practices that named and addressed racist institutional actions that prohibited equitable partnership, ultimately facilitating shared decision making, resource sharing, and knowledge co-creation.
Theoretical framework: antiracist praxis as a tool
An antiracist praxis draws on a cross-section of theoretical concepts relating to structural racism (Bailey et al., 2017), institutional racism (Adkins-Jackson et al., 2021), and anti-racism (Came & Griffith, 2018). Structural racism is facilitated by an institution, like an academic research institution, claiming authority over decision-making regarding research with historically marginalized communities. Performing institutional racism, academic institutional partners force CBOEs to abide by their regulations regarding stipends, documentation needed for incentives, and other harmful approaches that result from an institutional partner housing the grant funds for a project (Adkins-Jackson et al., 2022).
Anti-racism is an advocacy-based approach rooted in acknowledging and intervening in structural and institutional racism by increasing inclusivity, representation, and dismantling power structures (Griffith et al., 2007; Legha & Miranda, 2020). Key to anti-racism framing is the position that racism is a modifiable social construct; thus, anti-racism is a practice and not a fixed goal. Came and Griffith (2018) outline five components of an anti-racism praxis: a reflexive relational practice, socio-political education, structural power analysis, systems change, and monitoring and evaluating.
A reflexive relational practice refers to active relationship building where accountability for those in power is vital to the success of the partnership. In a community–institutional partnership, this might take the form of decentering the research needs of the institutional partner and encouraging the community partner’s preferences to be the deciding factor in an actionable step. Socio-political education involves a decolonization process of “unlearning and relearning conscientization” (Came & Griffith, 2018, p. 183), which is similar to how Legha and Miranda (2020) describe naming the racist legacies of institutions and their harmful impact on society. Through socio-political education, institutional partners can lessen the chance of repeating harm by unlearning the behaviors that have traditionally perpetuated structural and institutional racism in community partnerships; and relearning to critically analyze racism and inequities within their institution—while learning new partnership approaches and collaborative skills (Came & Griffith, 2018; Legha et al., 2020; Legha & Miranda, 2020).
The remaining components of an anti-racism praxis are core to the implementation of our restorative program. A structural power analysis is a process where the pathways through which racism operates are identified and opportunities for anti-racist intervention are targeted (Came & Griffith, 2018). A structural power analysis of community–institutional partnerships reveals striking power imbalances like policies that place restrictions on subawards and position institutional partners as key decision-makers, and ultimately, the sole responsibility for knowledge creation and dissemination. System change is where an anti-racist intervention occurs as the knowledge gained through the structural analysis is brought together with socio-political education and reflexive practices. Thus, as inequitable policies and practices are identified, a system change necessitates that institutional partners develop advocacy-based systems to resist power imbalances by advocating for fair, equitable, and just payment. Monitoring and evaluating institutional change is the key to ensure the accountability of a practice of anti-racism over time.
Given the utility of the Came and Griffith (2018) anti-racism praxis, we used CBPR as our engagement framework and anti-racism praxis as a tool to develop our multilevel pre-intervention restorative process. As we show in this article, our approach and antiracist praxis addressed the inequitable distribution of power and resources between these community–academic partnerships. The following question guided our work: Can CBPR-informed community–academic partnerships that employ an anti-racist praxis, foremost intervene on the role of structural racism on the mental health of historically marginalized communities, and also build an equitable partnership that establishes trust among partners?
Our approach: multilevel pre-intervention restorative process
Setting
The Share, Trust, Organize, and Partner COVID-19 California Alliance (STOP COVID-19 CA), a statewide collaborative of 11 universities and over 75 community partners, carried out this study from Fall 2020 to Fall 2021. This alliance included several workgroups, including a vaccine hesitancy workgroup (VHW), to address salient COVID-19 concerns in historically marginalized communities in California. Community and institutional representatives across sites formed the VHW, under the leadership of the first author, Dr. Adkins-Jackson, to identify barriers and facilitators to vaccine trial participation and vaccine uptake for marginalized communities throughout California (Cheney et al., 2021).
Reflexive Relational Praxis
Within the VHW, the first two authors, Drs. Adkins-Jackson and Vazquez, developed the multilevel pre-intervention restorative process to address the dual need of building relationships between community and institutional partners, and providing a safe space and service to historically marginalized communities that have been hard hit by the COVID-19 pandemic. Drs. Adkins-Jackson and Vazquez were frustrated with the constant extraction of stories and data from historically marginalized communities without the acknowledgment from researchers for the need to protect the mental health and well-being of their communities, particularly their grief and need for healing. Discrimination and oppression among these communities were amplified during the pandemic for research purposes. At the time of the study, Drs, Adkins-Jackson and Vazquez, both members of historically marginalized communities, were postdoctoral scholars. Together they developed the multilevel pre-intervention restorative process as an evolved approach rooted in the core principles of CBPR and applied anti-racism praxis as a tool to engage in equitable community–institutional partnerships to improve health in historically marginalized communities. The STOP COVID-19 CA project created an opportunity to use this approach.
Restoring Health Within Communities
Marginalized communities experience structural racism, oppression, and discrimination that are reflected in stressed immune systems, trauma-induced health behaviors, and income dependence that further increase exposure to COVID-19 (Bailey et al., 2017; Geronimus et al., 2010; Glymour & Manly, 2008; Stuifbergen & Im, 2008; Webb Hooper et al., 2020). Intersecting forms of disadvantage (e.g., racism, sexism, transphobia, lack of health insurance, having an undocumented status, having limited proficiency in English, being a part of the essential labor workforce) place some marginalized communities at greater risk for COVID-19 exacerbating the physical, psychological, and emotional well-being of already marginalized communities (Carson et al., 2021; Gehlbach et al., 2021; Hill et al., 2021). Discourses about the pandemic origin, spread, and low vaccination rates stigmatize and blame marginalized groups contributing to depressive and post-traumatic stress symptoms, substance use, and diminished life satisfaction (Bor et al., 2018; Cokley et al., 2022; Garcini et al., 2021; Stoller, 2021; Cheney et al., 2021).
One way to address collective mental health is through CBMHIs that meet the specific needs of the whole community through structurally and culturally responsive approaches (i.e., addressing anti-Black racism and context-driven trauma; Safe Black Space, 2021). CBMHIs involve multi-sector partnerships and emphasize community members as the designers, providers, facilitators (e.g., local practitioners, community members, and activists, faith leaders, educators, etc.), and recipients of the intervention in community settings (Castillo et al., 2019). CBMHIs provide trauma-informed mental health support in safe community settings and studies have shown that community-led interventions provide more culturally responsive information that lead to successful adoption of the service and necessary changes in health (Corbin et al., 2015; Makhay, 2021; McNeish et al., 2019; Plevin, 2019).
A restorative circle is a healing circle conducted in a safe space for community members to discuss their concerns regarding health, health care, COVID-19, and other related topics. Inspired by the Truth and Reconciliation Commission in South Africa and Safe Black Space in Sacramento, California (Brahm, 2007; Safe Black Space, 2021), Dr. Adkins-Jackson proposed a combination of these approaches in a restorative circle that allowed community members space to discuss their feelings about structural racism, COVID-19, and related events, but placed an emphasis on the needs of the community and the conflicts pertinent for them to discuss. Similar, to talking circles, umoja circles, emancipation circles, and sister circles, (Community Healing Network, 2021; Makhay, 2021; Plevin, 2019; Safe Black Space, 2021; Sister Circle, 2021), restorative circles provide space for individuals to repair harm through a facilitated dialogue (Ortega et al., 2016). Table 1 provides a list of these structurally and culturally responsive circles that are traditionally used in marginalized communities as safe spaces for the discussion of trauma and healing and the cultivation of resilience (Cowan et al., 2022). Like other structurally and culturally responsive circles, restorative circles center on collective healing through prompts that encourage participants to share common experiences. Given the successful implementation of the aforementioned CBMHIs, Drs. Adkins-Jackson and Vazquez believed that these novel restorative circles would be an effective strategy to reduce mental health burden due to structural racism and the COVID-19 pandemic, and thus, they were selected as the CBMHI. Although the multilevel pre-intervention restorative process was led by Drs. Adkins-Jackson and Vazquez as a part of the VHW, the restorative circles were organized by the community partners of the community–institutional partnerships of this study. This emic (insider) approach allows for self-agency among communities and yields ownership of a process where meaningful dialogue and connections could be made (Ortega et al., 2016).
Table 1 Examples of Culturally and Structurally Responsive Restorative Circles
Restorative circle Description Citation
Sister Circles Provides culturally responsive mental health counseling, social support, and collective healing practices that meet the unique needs of Black women (Sister Circle, 2021)
Umoja Circles Provides a space to express experiences related to health inequities and anti-Black racism (Makhay, 2021)
Safe Black Space Combines African-centered healing strategies (e.g., libation, drumming, etc.) within practices of mindfulness and other self-care exercises to overcome the traumas associated with structural racism (Safe Black Space, 2021)
Emotional Emancipation Circles Provides a deep level of healing by focusing on Black circle participants identifying the traumas they experience that are rooted in anti-Black racism and learning essential emotional wellness skills to overcome them (Community Healing Network, 2021)
Talking Healing Circle The gathering resembles a group counseling session infused with Mexican ancestral traditions. The facilitator burns the incense, beat an elk-skin drum and sing in Nahuatl, a Mexican indigenous language, in preparation for a practice they call a talking healing circle. (Plevin, 2019)
Restorative Circles
The 11 sites across California were invited to implement restorative circles in their respective regions. Sites were required to collaborate with community partners. Institutional partners provided planning support resources such as venues for in-person events, meeting platforms suitable for virtual circles, and other community resources referenced by CBOE’s as needed for their local community such as mental health professionals and informational pamphlets (e.g., on intimate partner violence). Drs. Adkins-Jackson and Vazquez provided 30-min training sessions to facilitators and mental health professionals to guide them on conducting restorative circles, which included four key components: setting the agenda, reviewing shared agreements, open discussion, and takeaways (described in Table 2).
Table 2 Restorative Circle Guidelines
Circle section Description Example
Setting (or introduction) Encompasses the physical and psychoemotional environment of the circle including the ambiance of the in-person or virtual space and the mood of the facilitator. During the setting, key components include land acknowledgments and the recognition of past and present abuses and discrimination inflicted upon the communities from which attendees descend. These acknowledgments establish solidarity within the circle.
Shared agreements The facilitator, mental health provider, and attendees set agreements. Attendees share best practices and approaches that promote safe spaces e.g., speaking one at a time, active listening, and confidentiality of information shared. These agreements affirms confidentiality and establish a safe space for everyone
Open discussion Facilitator leads attendees in discussions about trauma, coping, grief, resilience, and related topics. Storytelling can be used to inspire attendees to share their own experiences. Facilitators can also engaged attendees by setting a question e.g., “What food got you through this experience?” These strategies help attendees to reflect and share.
Takeaways Reflections from attendees on the benefits gained from attending the restorative circle. Attendees share one thing they are taking away from participating in the circle. This reflection helps attendees to be mindful of helpful resilience strategies.
As a core component of the multilevel pre-intervention restorative process, the restorative circles were to benefit historically marginalized communities directly. Therefore, the circles were not to be treated like a focus group where semi-structured questions guide responses and data is extracted from the discussion. Theremore, parameters to ensure the community was centered were set on the circles where minimal descriptive statistics were recorded (i.e., number of attendees), a pre- and post-test was not given, and institutional partners could not attend the circle.
In alignment with the CBPR and anti-racism praxis that we utilized (i.e., anti-racist community-partnered praxis), the community–institutional partnerships made decisions together regarding the characteristics, purpose, and styles of the restorative circles implemented within diverse communities. Thus, restorative circles varied by site, modality (virtual or in-person), time (anywhere from 90 to 120 min), and community characteristics. But all restorative circles were consistent in that they included facilitators and mental health professionals that reflected the community (e.g., Spanish speaking). In addition, the topics discussed varied per site as did the material shared with attendees at the conclusion of the restorative circles. For example, one site sought to engage adolescents and families preparing to return to in-person classrooms and the stress of this transition—participants were provided material on the signs and symptoms of anxiety among children and adolescents. During the sessions, the mental health professional helped facilitate the session and was available to provide one-on-one or group-level support, as needed, during the circle. Community partners were asked to set the monetary value for facilitators and mental health professionals to hold the circles.
Data Collection and Analysis
Data collection involved a group interview with community–institutional partners that implemented the restorative circles. In September and November 2021, Drs. Adkins-Jackson and Vasquez facilitated two group interviews with community partners, institutional partners, facilitators, and mental health professionals using a semi-structured interview guide with questions aimed to elicit information on characteristics of anti-racist community partnerships (shared decision-making, resource distribution, relational praxis, knowledge co-creation). These topics informed a deductive analysis of the group interview data in which analysts sought out examples of anti-racist praxis and examined the degree to which each partnership engaged in the four practices.
Results
We evaluated five sites or community–institutional partnerships in Southern California that collectively conducted six restorative circles. As described in Table 3, community partners varied, including one large recreational program, two local community-based organizations, and two local churches. One of the partnership teams included an academic research institution, an academic health center, a community-based organization, and a local church. Other partnership teams included an academic research institution and a CBOE.
The six restorative circles were conducted between June and October 2021: Four planned among existing partnerships and two with new partnerships. Two circles were conducted virtually and held in Spanish; four were conducted in-person and held in English. The circles ranged from 90 to 120 min with an average of 17 attendees. Table 3 provides attendee characteristics per circle and a list of resources provided to attendees.
Table 3 Restorative Circle Characteristics
Restorative circle Community Ages Number of attendees Resources provided Type of partnership
Online Latinx youth 16–25 12 How to cope with grief Community organization and university
Online Latinx Promotores 30–55 18 COVID-19 and children; Cognitive behavioral therapy Community organization and university
In-person Black Men 18–35 25 Coping with racial stress Church and university
In-person LGBTQ and allies 25–55 18 Coping with COVID-19-related stress Church, academic health center, and university
In-person LGBTQ and allies 25–55 13 Coping with COVID-19-related stress Church, academic health center, and university
In-person Transgender persons 17–26 5 Intimate partner violence; journals Community organization, academic health center, and university
Establishing an Anti-Racist Community Partnership
These findings highlight the ways community–institutional partners engaged in anti-racist praxis, evidenced by shared decision-making, equitable distribution of resources, reflexive relational practices, and knowledge co-creation, as well as how restorative circles offered a healing space for community members. These circles provided a safe space for discussing stress and sharing collective grief as attendees openly discussed their fears, concerns, and experiences during the COVID-19 pandemic.
Shared decision-making
Some CBOEs had specific ideas about how to organize and implement the restorative circles, whereas others would have liked more structure. For instance, one CBOE provided the institutional partner with a multi-page proposal of structured ideas for the circles. Whereas, another CBOE struggled with the lack of structure provided by the institutional partners and direction in determining the appropriate audience. This partnership required more follow-up meetings than other partnerships with their institutional partners and Drs. Adkins-Jackson and Vazquez for brainstorming and planning.
Equitable distribution of resources
Resources varied by partnerships. For instance, community partners serving as facilitators or mental health providers set the monetary value for their role in the restorative circles, thus their compensation varied. Community partners generally appreciated the opportunity to determine their compensation; however, this varied across partners as some institutional partners were concerned with the variation in pay across sites. One institutional partner argued for a set range to ensure pay equity across sites. Though this same institutional partner noted that this feature (i.e., setting their value) was important because community partners may work with institutions for free. This was evidenced by a community partner that responded positively: “I was shocked I was getting paid to do this.” This quote highlights the immense free labor that members of these communities provide.
Community partners encouraged compensation to restorative circle attendees, which was viewed as a genuine approach to community health and collective healing. Although discussed, the partners decided not to incentivize participation because attendees were not participants in the research study—the focus of data collection was on the partnership, not the restorative circles. However, one community partner disagreed with this decision. During the evaluation circle, it was noted that payment to attendees would have been cumbersome given institutional policies requiring participant identification and completion of W-9s or remunerating with gift cards instead of cash, which can deter participation from undocumented individuals.
Reflexive relational praxis
A key component of anti-racist praxis was transparent communication from the institutional partner. Both community and institutional partners described clear communication as critical to the success of implementing the restorative circles. Partners met a minimum of five times to plan and organize for the restorative circles. In addition to scheduled meetings, regular communication occurred via email and text message. Email exchanges with Dr. Adkins-Jackson and Vazquez facilitated payment procedures (e.g., invoicing, maneuvering the institutional payroll website, etc.) for facilitators and mental health professionals.
Both community and institutional partners described communication as transparent. During the group interview, one community partner described “loving” the institutional partners after working successfully with them. Transparent communication between partners was not observed by the community at-large. Only one of the restorative circles was hosted at an institutional partner’s site and none directly introduced the institutional partner. One institutional partner described uncertainty as to whether the communities would know they (i.e., the institutional partners) were involved in the service provided.
Trust. The process of planning and organizing the restorative circles presented an opportunity for active relationships and building trust between partners. For the pre-established partnerships, planning the restorative circles strengthened their relationships. One institutional partner shared how the community partner had been “the boss of this project.” Other institutional partners shared their appreciation for how the multilevel pre-intervention restorative process encouraged the community “to lead when working with a research institution.” Another community partner said the “trust and confidence” built with their institutional partner was appreciated.
Knowledge co-creation
Although institutional partners did not attend the circles, they held conversations with community partners to debrief and plan for upcoming circles. Institutional partners commented on the depth of knowledge generated during the circles. An institutional partner commented how they felt they had been “guessing what was needed but hearing from the community partner directly helped.” Both community and institutional partners described learning significant aspects about the community’s needs during the pandemic through the restorative circles. These needs varied and included access to public health information about the COVID-19 vaccine and children, a subject of concern among many circle attendees, as well as “tips and tools” to cope with racial stress induced by the pandemic. Community partners who attended the circles reported conversations from attendees shifting from discussions of depression, grief, and the COVID-19 vaccine for children to macro-level inequities, including racism, discrimination, and social injustices. In a restorative circle for Black men, attendees detailed lived experiences and traumas that affected their mental well-being stemming from injustices from local law enforcement agencies. These attendees recounted instances of police brutality, discrimination, and undiagnosed post-traumatic stress that negatively influenced their trust of government, health care, and public health more generally. In a restorative circle for self-identified transgender persons, discussion of domestic violence, trauma, and interpersonal violence occurred. Attendees also discussed violence and trauma as side effects of the pandemic not openly discussed or acknowledged elsewhere.
Benefits of community-led intervention
Community–institutional partners greatly appreciated and valued implementing restorative circles as a community-led intervention. For example, during the group interviews, community partners commented on how the process eased concerns about incorporating formal mental health services into the intervention. Community partners leading efforts for the Black male and Latinx/Hispanic restorative circles had been unsure about including a mental health professional given a history of distrust of health care systems in Black and Latinx communities. One community partner ultimately decided to have a pastor to facilitate the restorative circle for Black men in addition to a mental health provider from the same racialized group, city, and age group. For the circles held with the Indigenous Latin American community, a community mental health educator co-facilitated the circles. A deacon trained by Drs. Adkins-Jackson and Vazquez in psychoemotional support facilitated the LGBTQ and allies restorative circle.
Discussion
This study explored the implementation of an anti-racist community-partnered praxis in Southern California. The process yielded a multilevel pre-intervention restorative process that can be used in existing and new partnerships to address the role of structural and institutional racism in CBPR partnerships when grants have already been procured.
Using a structural power analysis from Came and Griffith’s (2018) anti-racism praxis, the community–institutional partners engaged in shared decision-making in which the voice of the community was centered. Through collaboratively planning the restorative circles (i.e., Level 1 of the multilevel pre-intervention restorative process), power was shifted from the institution to the community. Community partners shaped the restorative circle curriculum by making salient changes—not emphasizing the acknowledgments section because these were often addressed by circle attendees—and determining who facilitated the circle and what resources were provided to attendees.
However, despite community partners setting their value and the use of a third party to process payments (a university not conducting a partnership in this process), there was not sufficient system change to intervene on institutional inequities in the distribution of resources. Although this grant was received prospectively, institutional payment processes required retrospective reimbursement payments to CBOEs—meaning all community partners had to pay for their time and effort and wait for reimbursement. Some community partners struggled to fund their time upfront, delaying the implementation of their circles for months. Although this delayed the onset of implementation, the restorative circles still occurred due to the commitment of the community partners.
As described by Heaney and colleagues (2007), relationship building and trust between partnerships were strengthened when community partners led the collaboration. The partnerships were strengthened by transparent and continuous communication. Institutional partners openly shared the institutional restrictions they faced and community partners shared the impact of such restrictions on their ability to execute the circles. This transparency built trust between partners, allowing for a restorative collaboration (Cowan et al., 2022). However, the trustworthiness of specific individuals from an institution may not have translated to the institution’s trustworthiness. Although the collaboration was an effective first step at establishing or strengthening the partnerships, continued partnership may be needed to restore institutional trustworthiness.
The co-created knowledge gained by both partners helped solidify the partnership, even after the end of funding. Community partners learned more about the needs of their community without sacrificing attendees’ data. Institutional partners learned about concerns specific to these marginalized communities, and the role of restorative circles and mental health professionals in providing support. All partners gained knowledge about structuring CBMHIs to promote collective healing among marginalized communities (i.e., implementation of the CBMHI is Level 2 of the multilevel pre-intervention restorative process). With the exception of one partnership, plans were made to continue partnerships and pursue further funding, which suggests this pre-intervention process facilitates the community-building component of CBPR that allow further collaborations, possibly using a COMR model, to flourish. With continued partnership, brings more opportunities to address community concerns, obtain social change, and evaluate that change over time to ensure it occurs (Came & Griffith, 2018)—all of which successfully employ CBPR as initially envisioned.
Limitations
We designed the restorative circles to be a research-free safe space focused on acknowledgment, listening, respect, and collective healing. As such, data were not collected from attendees of restorative circles. This created a space in which attendees could openly share their stress, fears, and collective grief. Community partners were present at each of the circles and attended the group interviews where they shared their insights and observations. Our analysis is, thus, limited to observational data and feedback from community partners and does not include input from restorative circle attendees.
Implications for practice and policy
Our findings have implications for public health practice and institutional policy settings. First, there is a need for community-led interventions to address collective grief and trauma during the COVID-19 pandemic. Grief therapy and trauma-based counseling may provide safe spaces for therapeutic, innovative, and culturally responsive interventions that facilitate collective bereavement and healing that are needed, especially within historically marginalized communities. Interventions such as restorative circles may also assist communities in the management of collective trauma, stress, and inequities exacerbated by the pandemic. Many such communities bear the burden of limited access to COVID-19 testing, vaccination, and related health care services during the pandemic. Community-led interventions such as restorative circles, present structurally and culturally relevant ways to address existing and exacerbated unmet community mental health needs.
Second, by engaging in anti-racist practices—such as power sharing (i.e., shared decision-making and equitable distribution of resources)—community–institutional partnerships can begin to address the effects of structural racism on the community health of marginalized communities and foster deeper trust between these groups as seen in this work. Such praxis may prompt power-holding institutions like academic research institutions to consider ways to proactively integrate and operationalize anti-racist business, management operations, and policies as well as eliminate ongoing discriminatory policies and procedures in order to create equity between community and institutional partners. Our work shows the value of anti-racist praxis in the implementation of community-led interventions and advocates for funding research and programs led by community–institutional partnerships that embrace decolonizing methodologies and advocate for equity and social justice.
Despite a rich history of partnership, structural and institutional policy changes in the implementation of CBPR, COMR, and similar models are still needed to advance bidirectional partnerships among academics and CBOEs. As discussed, academic research institutions hold inequitable amounts of power over the resources often needed to build and sustain community–academic partnerships. Moreover, community–academic partnerships remain hard to implement in the traditional institutional workflows of many academic research centers (Nkimbeng et al., 2022; Strike et al., 2016). Thus, developing and mandating institutional guidelines and policies which fundamentally consider these partnerships and unequivocally reframe these as equitable is imperative if this work is to continue. As seen in our work, the multilevel pre-intervention restorative process with harmonized implementation between community–institutional partners working collaboratively on a project centering historically marginalized communities is possible. Therefore, reflecting this paradigm in institutional policy rhetoric and implementation may lead to greater trust between academic research institutions and historically marginalized communities as well as advance health equity in a manner previously unmatched in traditional research spaces.
ORCID iD: Paris B. Adkins-Jackson https://orcid.org/0000-0003-3618-449X
==== Refs
References
Adkins-Jackson P. B. Burke N. J. Espinosa P. R. Ison J. M. Goold S. D. Rosas L. G. Doubeni A. (2022). Inclusionary trials: A review of lessons not learned. Epidemiologic Reviews. Advance online publiscation. 10.1093/epirev/mxac007
Adkins-Jackson P. B. Legha R. K. Jones K. A. (2021). How to measure racism in academic health centers. AMA Journal of Ethics, 23 (2 ), 144–149. 10.1001/amajethics.2021.140
Ahmed S. M. Beck B. Maurana C. A. Newton G. (2004). Overcoming barriers to effective community-based participatory research in U.S. medical schools. Education for Health: Change in Learning & Practice, 17 (2 ), 141–151. 10.1080/13576280410001710969
Bailey Z. D. Krieger N. Agénor M. Graves J. Linos N. Bassett M. T. (2017). Structural racism and health inequities in the USA: Evidence and interventions. The Lancet, 389 (10077 ), 1453–1463. 10.1016/s0140-6736(17)30569-x
Bor J. Venkataramani A. S. Williams D. R. Tsai A. C. (2018). Police killings and their spillover effects on the mental health of Black Americans: A population-based, quasi-experimental study. The Lancet, 392 (10144 ), 302–310. 10.1016/s0140-6736(18)31130-9
Brahm E. (2007). Uncovering the truth: Examining truth commission success and impact. International Studies Perspectives, 8 , 16–35.
Came H. Griffith D. (2018). Tackling racism as a “wicked” public health problem: Enabling allies in anti-racism praxis. Social Science & Medicine, 199 , 181–188. 10.1016/j.socscimed.2017.03.028 28342562
Carson S. L. Casillas A. Castellon-Lopez Y. Mansfield L. N. Morris D. Barron J. Ntekume E. Landovitz R. Vassar S. D. Norris K. C. Dubinett S. M. Garrison N. A. Brown A. F. (2021). COVID-19 vaccine decision-making factors in racial and ethnic minority communities in Los Angeles, California. JAMA Network Open, 4 (9 ), e2127582. 10.1001/jamanetworkopen.2021.27582
Castillo E. G. Ijadi-Maghsoodi R. Shadravan S. Moore E. Mensah I. I. I. Docherty M. Aguilera Nunez M. G. Barcelo N. Goodsmith N. Halpin L. E. Morton I. Mango J. Montero A. E. Rahmanian Koushkaki S. Bromley E. Chung B. Jones F. Gabrielian S. Gelberg L. . . .Wells K. B. (2019). Community interventions to promote mental health and social equity. Current Psychiatry Reports, 21 (5 ), 35. 10.1007/s11920-019-1017-0 30927093
Cheney A. McMullin J. Vázquez E. Burroughs M. Ortiz G. Moran A. Trehan S. LeComte-Hinely J. Smith S. Stell R. P. Pozar M. Salgado S. Juturu P. Gonzalez C. Lo D. (2021). The STOP COVID-19 CA Project: Factors shaping COVID-19 testing and vaccination in vulnerable communities in Inland Southern California: 2021 qualitative report. https://drive.google.com/file/d/10-Z8ObLfQbDnrieA2JPHvvKcVw5ccl3s/view
Cokley K. Krueger N. Cunningham S. R. Burlew K. Hall S. Harris K. Castelin S. Coleman C. (2022). The COVID-19/racial injustice syndemic and mental health among Black Americans: The roles of general and race-related COVID worry, cultural mistrust, and perceived discrimination. Journal of Community Psychology, 50 , 2542–2561. 10.1002/jcop.22747 34797928
Community Healing Network. (2021, October 7). Emotional emancipation circles. https://communityhealingnet.org/emotional-emancipation-circle/
Corbin J. H. Fernandez M. E. Mullen P. D. (2015). Evaluation of a community-academic partnership: Lessons from Latinos in a network for cancer control. Health Promotion Practice, 16 (3 ), 345–353. 10.1177/1524839914558514 25395057
Cowan E. S. Dill L. J. Sutton S. (2022). Collective healing: A framework for building transformative collaborations in public health. Health Promotion Practice, 23 , 356–360. 10.1177/15248399211032607 34549646
Garcini L. M. Daly R. Chen N. Mehl J. Pham T. Phan T. Hansen B. Kothare A. (2021). Undocumented immigrants and mental health: A systematic review of recent methodology and findings in the United States. Journal of Migration and Health, 4 , 100058. 10.1016/j.jmh.2021.100058 34405198
Gee G. C. Hicken M. T. (2021). Structural racism: The rules and relations of inequity. Ethnicity & Disease, 31(Suppl. 1 ), 293–300. 10.18865/ed.31.s1.293
Gehlbach D. Vázquez E. Ortiz G. Li E. Sánchez C. B. Rodríguez S. Pozar M. Cheney A. M. (2021). COVID-19 testing and vaccine hesitancy in Latinx farm-working communities in the Eastern Coachella Valley. Preprint. Res Sq. 2021;rs.3.rs-587686. Published 2021 Jun 25. 10.21203/rs.3.rs-587686/v1
Geronimus A. T. Hicken M. T. Pearson J. A. Seashols S. J. Brown K. L. Cruz T. D. (2010). Do U.S. Black women experience stress-related accelerated biological aging? Human Nature, 21 (1 ), 19–38. 10.1007/s12110-010-9078-0 20436780
Glymour M. M. Manly J. J. (2008). Lifecourse social conditions and racial and ethnic patterns of cognitive aging. Neuropsychology Review, 18 (3 ), 223–254. 10.1007/s11065-008-9064-z 18815889
Gravlee C. C. (2020). Systemic racism, chronic health inequities, and COVID-19: A syndemic in the making? American Journal of Human Biology, 32 (5 ), e23482. 10.1002/ajhb.23482
Griffith D. M. Mason M. Yonas M. Eng E. Jeffries V. Plihcik S. Parks B. (2007). Dismantling institutional racism: Theory and action. American Journal of Community Psychology, 39 (3–4 ), 381–392. 10.1007/s10464-007-9117-0 17404829
Heaney C. D. Wilson S. M. S.M. Wilson O. R. (2007). The West End Revitalization Association’s community-owned and -managed research model: Development, implementation, and action. Progress in Community Health Partnerships: Research, Education, and Action, 1 (4 ), 339–349. 10.1353/cpr.2007.0037 20208213
Hill J. Rodriguez D. X. McDaniel P. N. (2021). Immigration status as a health care barrier in the USA during COVID-19. Journal of Migration and Health, 4 , 100036. 10.1016/j.jmh.2021.100036 33778797
Israel B. A. Schulz A. J. Parker E. A. Becker A. B. (1998). Review of community-based research: Assessing partnership approaches to improve public health. Annual Review of Public Health, 19 , 173–202.
Legha R. K. Miranda J. (2020). An anti-racist approach to achieving mental health equity in clinical care. Psychiatric Clinics of North America, 43 (3 ), 451–469. 10.1016/j.psc.2020.05.002 32773074
Legha R. K. Williams D. R. Snowden L. Miranda J. (2020, November 4). Getting our knees off Black people’s necks: An anti-racist approach to medical care. Health Affairs Blog. https://www.healthaffairs.org/do/10.1377/hblog20201029.167296/full/
Makhay B. (2021, January 19). Umoja circles: Processing support for Black community members. Columbia University Neighbors. https://neighbors.columbia.edu/news/umoja-circles-processing-support-black-community-members
McNeish R. Rigg K. K. Tran Q. Hodges S. (2019). Community-based behavioral health interventions: Developing Strong Community Partnerships. Evaluation and Program Planning, 73 , 111–115. 10.1016/j.evalprogplan.2018.12.005 30580000
Mendenhall E. Newfield T. Tsai A. C. (2021). Syndemic theory, methods, and data. Social Science & Medicine, 295 , 114656. 10.1016/j.socscimed.2021.114656 34949486
Minkler M. Wallerstein N. (2008). Community-based participatory research for health: From process to outcomes (2nd ed.). Jossey-Bass.
National Institute on Minority Health and Health Disparities. (2018). Community-Based Participatory Research Program (CBPR). https://www.nimhd.nih.gov/programs/extramural/community-based-participatory.html
Nkimbeng M. Han H. R. Szanton S. L. Alexander K. A. Davey-Rothwell M. Giger J. T. Gitlin L. N. Joo J. H. Koeuth S. Marx K. A. Mingo C. A. (2022). Exploring challenges and strategies in partnering with community-based organizations to advance intervention development and implementation with older adults. The Gerontologist, 62 , 1104–1111. 10.1093/geront/gnab190 34958098
Ortega L. Lyubansky M. Nettles S. Espelage D. (2016). Outcomes of a restorative circles program in a high school setting. Psychology of Violence, 6 , 459–468. https://psycnet.apa.org/buy/2016-21281-001
Plevin R. (2019, October 23). “We’re just here to listen”: After El Paso shooting, some Latinos turn to Mexican ancestral traditions for healing. Palm Springs Desert Sun. https://eu.desertsun.com/story/news/2019/10/22/after-el-paso-shooting-some-latinos-turn-mexican-ancestral-traditions-healing/2431074001/
Safe Black Space. (2021). About us. https://www.safeblackspace.org/about-us
Schulz A. J. Parker E. A. Israel B. A. Becker A. B. Maciak B. J. Hollis R. (1998). Conducting a participatory community-based survey for a community health intervention on Detroit’s east side. Journal of Public Health Management & Practice, 4 , 10–24.
Sister Circle. (2021). Sister circle. The Women’s Initiative. https://thewomensinitiative.org/sister-circle
Strike C. Guta A. De Prinse K. Switzer S. Carusone S. C. (2016). Opportunities, challenges and ethical issues associated with conducting community-based participatory research in a hospital setting. Research Ethics, 12 (3 ), 149–157. 10.1177/1747016115626496
Stoller C. (2021, March 29). What COVID-19 can teach us: Violence and marginalization. UNH Today. https://www.unh.edu/unhtoday/2021/03/what-covid-19-can-teach-us-violence-and-marginalization
Stuifbergen A. K. Im E. O. (2008). An online forum on cancer patients’ needs for help: Gender and ethnic differences. Oncology Nursing Forum, 35 (4 ), 653–660. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2713991/ 18591169
Van de Sande A. Schwartz K. (2017). Research for social justice: A community-based participatory approach (2nd ed.). Fernwood Publishing.
Webb Hooper M. Nápoles A. M. Pérez-Stable E. J . (2020). COVID-19 and racial/ethnic disparities. Journal of the American Medical Association, 323 (24 ), 2466–2467. 10.1001/jama.2020.8598 32391864
Zimmerman E. (2020). Researching health together: Engaging patients and stakeholders, from topic identification to policy change (1st ed.). SAGE.
| 36419256 | PMC9703012 | NO-CC CODE | 2022-11-29 23:21:05 | no | Health Promot Pract. 2022 Nov 23;:15248399221132581 | utf-8 | Health Promot Pract | 2,022 | 10.1177/15248399221132581 | oa_other |
==== Front
Br J Pain
Br J Pain
spbjp
BJP
British Journal of Pain
2049-4637
2049-4645
SAGE Publications Sage UK: London, England
10.1177_20494637221121703
10.1177/20494637221121703
Original Manuscript
How has the COVID-19 pandemic affected patients’ experience of pain management therapy?
https://orcid.org/0000-0003-4869-5035
French Olivia MSc 1
Mattacola Emily MSc, PhD 1
1 School of Psychology, 2060 University of Buckingham , Buckingham, UK
Olivia French, School of Psychology, University of Buckingham, Hunter Street, Buckingham MK18 1EG, UK. Email: [email protected]
24 11 2022
24 11 2022
20494637221121703© The Author(s) 2022
2022
British Pain Society
This article is made available via the PMC Open Access Subset for unrestricted re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the COVID-19 pandemic or until permissions are revoked in writing. Upon expiration of these permissions, PMC is granted a perpetual license to make this article available via PMC and Europe PMC, consistent with existing copyright protections.
Objectives
The current service evaluation aimed to explore the impact of COVID-19 on patients' experiences of pain management therapy. The study examined the barriers and benefits of the move from face-to-face to eHealth methods of delivery.
Design
A service evaluation was conducted in an outpatient pain clinic in an NHS Trust in the East of England. A qualitative approach was taken using semi-structured interviews.
Methods
Participants were recruited through a health psychology service operating as part of a multidisciplinary pain management clinic. Six patients, aged 39–67, were interviewed one-to-one using the online platform ZoomTM. During COVID-19, participants had individual or group pain management therapy via telephone or video conferencing. All interviews were transcribed using Otter.aiTM and thematic analysis was performed. The study was approved by internal clinical governance for service evaluations and the authors adhered to the BPS Code of Human Research Ethics.
Results
Three key themes emerged from the analysis; Benefits Aside From Pain Relief, Limited Their Experience, and COVID-19: A Double-Edged Sword.
Conclusion
Findings suggested patients were able to benefit from pain management therapy despite the impact of COVID-19 on daily routines and pain experience. Adopting eHealth methods during the pandemic was an effective means of accessing pain management therapy. These methods allowed patients to continue to benefit from peer support and learn about skills and resources regarding self-management, whilst also improving accessibility for those with chronic pain. Yet, these methods are not without their limitations. Technical issues and difficulties creating therapeutic connections with psychologists limited patients' experience of pain management therapy.
pain management
COVID-19
patient experience
edited-statecorrected-proof
typesetterts10
==== Body
pmcStatement of Contribution
What is already known on this subject?
1. Pain management programmes are more effective than individual therapy.
2. Existing online pain management programmes are designed specifically for online platforms.
3. COVID-19 restrictions and regulations increased pain experience for people with chronic pain.
What does this study add?
COVID-19 did not impact patients’ ability to gain support and knowledge from pain management therapy.
eHealth methods have limitations and contingencies should be put in place by healthcare services.
Patients should be given a choice of delivery method for pain management post-COVID.
Introduction
Chronic pain is defined as an experience of pain for longer than 3 months. It is both an accompanying symptom of existing conditions and a condition in its own right.1 Currently, chronic pain affects one-third to one-half of the UK population,2 is the leading cause of disability on a global scale,3 and is significantly associated with unemployment and low household income.4 Hence, governments across the world are beginning to treat chronic pain as a serious challenge to public health.2 The literature argues chronic pain interventions should use a biopsychosocial framework, as chronic pain has considerable biological factors and detrimental psychological and social impacts.5 Evidence supporting psychological interventions has found improvements to patient experience and reduced healthcare system costs.6 Furthermore, interventions using cognitive behavioural therapy [CBT] can reduce pain experience and catastrophizing beliefs.7 Psychological interventions have also been linked to improved wellbeing and lower rates of depression in chronic pain patients.8 Thus, psychological interventions should be integrated into the treatment and management of chronic pain.
Pain management programmes [PMPs] are multidisciplinary treatment plans which comprise physical, educational, and psychological components.9 They aim to improve patients' quality of life and increase self-management.10 Pain management programmes typically last between six and eight weeks and range in intensity, from one hour sessions to eight hour days.10 Pain management programmes are often delivered to people with heterogeneous pain conditions11 and are typically delivered in a group environment. However, evidence suggests group settings are not always an appropriate mode of delivery.10 Individuals suffering from complex psychological conditions or multi-comorbidities may benefit from individual pain management therapy, instead of the group setting.12 To establish which method is appropriate for patients, healthcare professionals [HCPs] conduct initial assessments and consider the circumstances of a patient’s pain experience.12 However, during COVID-19, both individual pain management therapy and group PMPs had to adapt to the regulations enforced by the government to continue providing treatment to patients living with chronic pain.
The COVID-19 pandemic directed all non-essential healthcare services to a remote setting following government restrictions. To reduce the spread of the virus, assessments and therapies that originally took place in-person were forced to take place remotely. This change saw HCPs quickly accommodate new delivery methods to maintain care and contact with patients. During the pandemic, there were no guidelines provided to HCPs on the management of chronic pain patients,13 and so pain management services were left to decide the best delivery method to continue providing pain management therapy during the crisis. Depending on the alternative delivery methods chosen, HCPs had to adapt their communication and clinical skills to compensate for the lack of visual and physical contact with patients.14 Recent evidence suggests this caused HCPs to feel high levels of self-doubt, anxiety, and worry.15 Additionally, HCPs may have experienced trauma or loss during the pandemic which impacted their ability to provide care to patients.16 Hence, the sudden move away from in-person sessions to alternative methods may have caused difficulties and extra pressure for psychologists providing pain management interventions.
Yet, despite the difficulties for HCPs, it was extremely important for pain management therapy to continue. Evidence suggests pain experience likely increased for many chronic pain patients during COVID-19 as a result of government restrictions. A multitude of psychosocial factors potentially caused this increase, such as personal loss, worry about financial loss, and increased anxiety.17 Restrictions such as self-isolation impacted pain experience through increasing people’s feelings of loneliness; a feeling strongly associated with increased pain experience, depression, and fatigue.18 Self-isolation has also been associated with a heightened awareness of pain symptoms and fear of infection, which can cause patients to fear travelling to a public medical setting to seek medical attention.19 Previous evidence proposes untreated chronic pain can increase depression rates by 50% and leads to a 35% increase in suicide ideation,20 feelings that are linked to increased pain. Further to this, patients were likely to adopt passive coping strategies for their pain during lockdown, which can increase depressive symptoms and suicide ideation, therefore increasing the likelihood of pain morbidity and mortality occurring.21 Therefore, it was extremely important for healthcare services to continue providing pain management therapy using alternative delivery methods to reduce the negative impacts of the restrictions implemented during COVID-19.
As a result of this importance, alternative delivery methods needed to be chosen to continue delivering pain management therapy during the pandemic. Before COVID-19, eHealth delivery methods, such as telephone and video conferencing, were being trialled to deliver pain management interventions.22 However, during the pandemic, eHealth methods were the only alternative option that adhered to government restrictions. Research indicates eHealth delivery methods are beneficial for patients with chronic pain. They are associated with improved quality of life and reduced healthcare costs for patients with pain conditions.23 Telephone communication provides an inexpensive alternative to face-to-face sessions22 and enables HCPs to reach a large population. Additionally, this type of intervention has been found to reduce pain intensity in patients with chronic pain.24 Video conferencing was also a popular delivery method during the pandemic, with the use of platforms such as Attending Anywhere and Zoom.25 These video conferencing platforms allowed HCPs to physically see patients and communicate in real-time whilst adhering to social distancing rules.26 Therefore eHealth methods provided HCPs with sufficient alternative options to deliver pain management therapy to patients during the pandemic.
Furthermore, evidence demonstrates eHealth methods are effective in reducing barriers for patients. Existing online PMPs have been found to reduce stigma, provide timely information, and provide accessibility for isolated groups.27 Additionally, patients attending online PMPs have shown reduced levels of disability, anxiety, and depression after treatment, compared to those who did not attend a programme.28 A reason for this could be because online PMPs allow patients to complete programmes without feeling pressured to meet HCPs expectations, a feeling that may occur during face-to-face sessions. Additionally, patients have self-reported increased life satisfaction at 3 and 6 months when they have completed an online PMP.29 This long-term efficacy is arguably caused by patients learning to self-manage their pain in a comfortable and familiar environment that allows them to maintain these newly learnt behaviours.30 Hence, online PMPs may be beneficial for providing long-term pain management for patients with chronic pain, presenting a viable alternative to face-to-face delivery.
However, contradictory evidence suggests online interventions may not be appropriate for patients who experience chronic pain. These types of programmes have little involvement from psychologists and provide generalised educational resources for patients. Nevedal, et al.31 found patients with complex comorbidities were more likely to benefit from a personally tailored, disease-specific approach that comes from individual face-to-face therapy. Face-to-face individual interventions provide patients with the opportunity to receive unique care, tailored to the demands of their condition. However, despite the effectiveness of in-person sessions, uptake of face-to-face sessions is significantly higher in patients from affluent areas and in patients of older ages,32 whereas online interventions are accessible across socioeconomic backgrounds and ages. General internet-based programmes are also cost-effective for the healthcare system compared to individually tailored sessions.28 Thus, despite online pain management tools providing generalised educational information for patients, eHealth can reduce barriers associated with face-to-face sessions.
Even though there are evidence-based benefits of using online PMPs, research tends to focus on the experience of patients who have actively selected this mode of delivery or participated in a randomised trial where they knew online delivery was a possibility. Patients accessing face-to-face pain management sessions prior to COVID-19 were forced to continue their therapy using eHealth methods, such as telephone or video appointments due to government restrictions. This patient population potentially differs from those who specifically choose to have pain management online or participate in a trial. Individuals who self-select online programmes may be choosing them for specific reasons, such as feeling reluctant to seek help in primary care settings, and financial barriers.33 The remote sessions offered during COVID-19 were arguably different to existing online PMPs, which are readily available to patients and designed for an online platform. These programmes are focussed on self-guided symptom management, and are self-directed and self-paced.33 However, remote sessions during COVID-19 developed by HCPs were structured to mirror live face-to-face appointments. Healthcare professionals ensured patients attended the sessions at specific times and dates, and the sessions lasted a specific timeframe. Additionally, HCPs using video conferencing expected patients to attend with their video cameras turned on, giving the session a face-to-face and live quality that other internet-based interventions do not require. This allowed HCPs to maintain as much normality within the sessions as possible to ensure the transition from face-to-face to eHealth methods was comfortable for patients.
Furthermore, patients who began with face-to-face sessions, with the hope of joining group PMPs, were also moved to an online setting. It is likely this method was not their preferred choice and may have affected their experience of pain management sessions. In-person settings promote social interaction between patients and can encourage mobility in patients with chronic pain. Consequently, the experience of patients who were forced to an online environment will differ from that of patients who consciously selected an online programme. Understanding patients' experiences of pain management therapy during COVID-19 and how the use of eHealth has impacted patients' experience of pain management therapy can inform pain management services development post-COVID-19.
Aim and objectives
The current study aims to explore the impact COVID-19 has had on patients' experience of pain management sessions. To examine this impact, the study will analyse the barriers and benefits that developed as a consequence of moving from face-to-face to eHealth delivery methods.
Research question
How has the COVID-19 pandemic affected patients' experience of pain management therapy?
Method
Design
The researchers conducted a service evaluation of pain psychology sessions delivered by a Clinical Health Psychology service as part of a multidisciplinary pain management team in an NHS Trust during the COVID-19 pandemic. Braun and Clarke34 suggest qualitative research methods produce in-depth, exploratory data, hence this methodology was implemented as the study aimed to explore individuals’ perspectives and subjective experiences. The researcher conducted online semi-structured interviews with individuals who began pain psychology sessions face-to-face and continued with eHealth delivery during COVID-19. The use of semi-structured interviews to explore experiences has been supported by DeJonckheere and Vaughn,35 who propose semi-structured interviews allow researchers to explore experiences, thoughts, and feelings whilst collecting open data around research topics. The researchers used an interview schedule during data collection, allowing them to maintain high reliability throughout the study.34
Participants
All participants had previously been referred to Clinical Health Psychology and after initial assessment were offered individual face-to-face sessions, some with the potential to join a group setting after several individual sessions. The inclusion criteria for the study stipulated participants being over the age of 18 years old and having begun individual sessions face-to-face that were moved to an eHealth mode of delivery due to COVID-19 (see Table 1). The pain clinic provided a list of potential participants who met the inclusion criteria. Individuals were excluded if they had serious psychological comorbidities or the Pain Psychology team felt they were too vulnerable to participate. Sixteen potential participants were approached. Ten declined to participate or did not respond, leaving six participants consenting to take part. Despite the small number, this met the suggested criteria proposed by Braun and Clarke34 for a study of this size (6–15 participants). Table 1. Participant biographies.
Participant pseudonym Age Gender Pain condition (if disclosed) eHealth format during COVID-19
Ken 67 Male Degenerative disc disease, diabetes, osteoarthritis and dystonia. Individual video and telephone sessions.
Maria 53 Female Fibromyalgia, osteoarthritis and degenerative disc disease. Video PMP.
Daniel 55 Male Nerve damage (after an accident) Video PMP.
Taylor 60 Female Morton’s Neuroma. Individual telephone sessions.
Video PMP.
Sarah 39 Female Fibromyalgia Video PMP.
Mark 60 Male Vasculitis and kidney failure. Video PMP.
Note. The table demonstrates participant demographic information, such as age and gender. It also discloses the participant’s pain condition to provide an understanding of their pain experience. It also states what type of eHealth method each participant was provided during the COVID-19 pandemic.
Note. PMP: Pain management programmes
Data collection procedure
Participants who met the inclusion criteria were approached with an expression of interest email and information sheet (see Supporting Information). A follow-up email was sent a week later to participants who did not respond. For those who did not respond to the email invitations, follow-up phone calls were made to ensure participants had received the invite and per their request, invites were sent again with an attached consent form. If patients wished to take part, they were asked to respond to the email with a completed consent form, then a mutually convenient date and time for interview were agreed upon.
Before the interview, participants were sent a Zoom video conferencing link, with participants asked to join from a confidential space where no one could overhear the conversation. The interviews took between 30 and 40 min. After the interview, participants were emailed a debrief form and thanked for their participation. Otter.ai and an android device were used to record the interviews. Recordings were then transcribed using Otter.ai. All recordings and transcriptions were anonymised and saved on a password-protected laptop. Analysis was conducted by the first author by hand, and checked by the second author.
Data collection materials
An interview schedule was written by the researchers to help guide the semi-structured interviews (see Supporting Information). As suggested by Gill et al.,36 the questions were designed to yield the most information possible whilst focussing on the aims and objectives of the service evaluation. The schedule was split into four focal points: initial thoughts around pain psychology and face-to-face sessions, the experience of COVID-19, feelings towards online/telephone sessions, and future implications and delivery of pain psychology. The schedule consisted of 8 main questions, each question was clear, open-ended, avoided jargon and moved from easy to in-depth focus.35 Further to the main questions, 12 potential probes were also included, however, based on the advice of Roulston and Choi,37 the probes were formulated and used relative to what the participant said during the interview. The questions were reviewed by the Lead for Clinical Health Psychology in the pain management team before use with participants.
Along with the interview schedule, an information sheet, consent form, and debrief sheet were used throughout the research process for ethical and informative purposes.
Analytic approach
Inductive thematic analysis (TA), as developed by Braun and Clarke,38,39 was used to analyse the data set. TA allows researchers to identify, organise, and make sense of patterns across data sets. King40 proposed TA is a useful method for highlighting similarities across data and generating unanticipated themes from an unknown phenomenon. Hence, TA was an appropriate choice for the current study as little is known about how COVID-19 has affected patients' experience of pain management therapy. Inductive TA, commonly known as a bottom-up approach, is where the analytical process is guided by the data itself.39 This type of analysis was appropriate for the current study as it allows basic observations to develop into complex understandings of experiences. Braun and Clarke41 argue committing to an inductive approach leads to a stronger focus on the meanings that participants have made of the world around them. Therefore, taking an inductive approach was appropriate for the current research question due to the prominent focus on patient experience and their unique perspectives and understandings. In contrast, a deductive approach relies on researchers’ pre-existing theoretical understandings to drive analysis.42 This approach has been associated with quantitative methods34 and does not provide a deep description of an overall dataset.38
Braun and Clarke’s34 six-phase approach to analysis was adopted. The first author began by familiarising themselves with the data set by reading and re-reading the transcripts, making initial annotations to highlight key points and areas of interest. Initial codes were generated for each data set, taking a semantic focus. Smith43 proposes semantic meaning refers to the explicit meaning the person is communicating with the researchers. Thus, as the current study focuses on explicit experiences, semantic coding was used.39 Each interview was coded until themes related to patient experience of pain management therapy during COVID-19 were clearly identifiable across each transcribed interview, and no further new themes were generated. These themes were then reviewed, named and defined across the data set as a whole (see Figure 1). During the write-up process, the final themes were related back to a broader theoretical base and current understanding.Figure 1. Thematic Map.
Ethical considerations
NHS ethical approval was not required as the study was a service evaluation within an NHS trust. Instead, the appropriate research governance approvals for a service evaluation were achieved. However, the study adhered to The British Psychological Society Code of Human Research Ethics,44 which states researchers should respect the dignity and autonomy of participants and ensure they minimise any potential impact on the participants. Consent forms, audio recordings, transcripts and analysis were all saved in separate protected folders to maintain participant confidentiality and these files were only accessible to the research team. After each interview, data were saved anonymously and identifiable data, such as names and places, in the transcripts were replaced with simple descriptors. During the analysis, pseudonyms were used to retain anonymity.
An information sheet was provided to all potential participants to explain the purpose of the study, what was expected of the participants, and their right to withdraw. The consent form required participants to confirm they understood all the information provided to them and consented to their participation. A debrief was sent to participants after their interview, reminding them of the study aims and their right to withdraw. Although it was felt that there was minimal risk of harm to participants, the study focused on personal experience, and therefore the debrief signposted to educational resources and appropriate sources of suppport should further information be required.
Reflexivity
The first author, who was the primary coder for the data, is a female trainee Health Psychologist with a diagnosed chronic pain condition. I acknowledge my ability to relate to participants’ experiences of chronic pain and also the impact that COVID-19 has had on self-management. I understood that I could potentially have an emotional response to participants during the interviews or have difficulty asking specific questions, especially if participants had similar conditions to mine. I recognised that my interpretation of participants’ experiences during analysis may be biased based on my own experiences with chronic pain. Hence, the interpretation should be viewed through the lens of an insider researcher. I devised the interview schedule with support from the second author and the pain psychology team, based on the operation of the pain psychology service, the restrictions COVID-19 initiated on the healthcare system, and existing literature around eHealth and patient experiences of pain management. Therefore I feel the schedule guided the interviews and focused on the research’s aims and objectives. Throughout the analysis, I was conscious of my empathy with the participants but focussed on the data itself and what it showed.
Analysis
Six patients from the pain psychology clinic participated in the service evaluation. Reasons for declining the invitation included lack of confidence in using Zoom and lack of time. During the analysis, three key themes were identified; Benefits Aside from Pain Relief, Limited the Experience, and COVID-19: A Double-Edged Sword. All of the themes and their sub-themes can be seen in Figure 1 and are described below.
Theme 1 – Benefits aside from pain relief
The aim of pain management therapy is not just to help individuals learn how to self-manage their pain condition. The online group PMPs aimed to help bring people with similar experiences together to share knowledge and to help individuals feel less alone in their pain experiences, ‘It was positive to discuss with other people and they were all open and friendly...and to understand how other people approached things. So all of that was, was helpful’. (Mark, p.4;152). These connections between patients extended after the PMP, to patient support groups, ‘So we’re still in touch with each other...it’s really really useful, and I’m glad that, that we were able to do that so we still continue to support each other…’ (Daniel, p.6; 273). Support from others experiencing similar situations can help patients feel less alone in their pain experience, improve their quality of life, and increase understanding of pain through sharing knowledge and experiences. Due to the pandemic restrictions, this would not have been possible for patients during COVID-19 without the adoption of eHealth methods and the continuation of PMPs.
Despite having to attend the pain management sessions in a different environment than originally expected, patients felt they continued to benefit from the sessions and gained an in-depth understanding and knowledge around pain management. Patients felt they were still able to learn about specific self-management skills and topics they may not have been aware of, ‘everything, every issue, every topic was covered on that course, it was utterly brilliant… It blew me away’. (Taylor, p.7;299). This better equipped them with their pain management moving forward as they were able to implement skills they had learnt into their self-management. Patients who attended the online PMP expressed learning about services available to them which they were previously unaware of, ‘...to know that there are others things that are available and that there’s help. ..that was sort of really good…And just knowing you’ve got access to people…’ (Sarah, p.4;172). This was helpful to ensure patients didn’t feel alone in their experience and highlighted resources that can help improve their self-management of their pain, for instance, access to medication reviews or disability-related facilities.
Theme 2 – Limited the Experience
Despite the benefits identified by participants for eHealth modes of delivery, the majority of patients felt eHealth methods also limited their experience. Participants found it difficult to communicate with the psychologists when using eHealth methods, which caused them to feel incapable of building sufficient therapeutic connections with psychologists, ‘my issue, my, my, problem is that I can’t portray over the phone, sort of what I’m going through’. (Taylor, p.4;147). This inability to create open and comfortable relationships between HCPs and their patients potentially impacted patients' self-management. Arguably this impact on experiences would not have occurred during face-to-face sessions and therefore COVID-19 caused this limitation.
The move to eHealth delivery relied on patients owning the technological equipment to allow them to participate. Yet, some of the patients did not have the equipment needed to participate fully in the online sessions. This consequently impacted their experience. ‘...so I don’t have a laptop. So I used to get my phone...But to get it to work with these different systems has been a nightmare...I managed to get by the second group meeting…’ (Daniel, p.2;93). Daniel's reflection suggests his lack of equipment and technical knowledge limited his access to the PMP and caused frustration. Furthermore, every patient who attended online sessions recalled experiencing technological difficulties. Participants commented on problems with their internet connection and lack of technological knowledge which caused them to experience problems during their sessions, ‘...they kind of throw me slightly...there’s a lot of erm feedback kind of sounds and strange noises…’ (Ken, p.10;407). Some participants also experienced difficulties engaging with the online PMP material because of technical difficulties experienced by the HCP, ‘She wasn’t always able to transfer from the picture of her to the, to the slides, she wanted to show. I think that happened on quite a few occasions’. (Mark, p.4;166). Hence, technical difficulties with eHealth can cause difficulties with engagement and learning for participants, especially if they have a lack of technical knowledge.
Theme 3 – COVID-19: a double-edged sword
The pandemic enforced restrictions that stopped participants from engaging in their daily routines and activities which were part of their pain self-management. For instance, the lockdowns forced some participants to stop being physically active, an important tool for pain management. ‘I was getting lots of exercises and doing really well and then when the lockdown came all that stopped…’ (Daniel, p.2;73). This suggests the COVID-19 restrictions halted participants' ability to engage in self-management behaviours, such as physical exercise, which consequently affected their ability to manage their pain.
Many participants felt COVID-19 introduced an easier way to access pain management therapy. Despite the negative impact of the pandemic, the majority of participants expressed that eHealth methods allowed them to be comfortable during sessions. They enjoyed the lack of travel, cost-effectiveness, and limited demand on their bodies, compared to face-to-face sessions, ‘...the amount of travel time was nil, which was good. The comfort part of it, I was able to put my feet up not be bothered about anybody else’. (Mark, p.4;177). Further to this, participants felt more relaxed being able to choose their surroundings for pain management therapy, ‘...instead of being sat up, or anything and being in discomfort. I was able to go between laying down...and actually been comfortable in my own surroundings’. (Sarah, p.4;144). Being relaxed during the pain management therapy allowed participants to reduce their pain experience, gain more from the sessions, and increased engagement. Hence, it can be argued that eHealth introduced a superior mode of delivery for pain management therapy as it reduced barriers such as travel, discomfort, and increased anxiety.
Discussion
The study examined the effect of the COVID-19 pandemic on patients' experiences of pain management therapy. The authors identified three key themes (see Figure 1). The first, Benefits Aside from Pain Relief, demonstrated how beneficial pain management therapy can be for patients' quality of life, as well as their pain experience. Creating connections, sharing mutual experiences and understanding of pain with their peers helped participants feel less alone in their pain experience. Chronic pain conditions can cause disabling symptoms which potentially force individuals to disconnect from others and increase feelings of loneliness, which increase pain experience.19 Evidence also suggests self-isolation restrictions implemented during COVID-19 increased the likelihood of loneliness. Isolation from others and lack of access to social support during the pandemic consequently increased pain experience for people with chronic pain.45 Thus, it was important to continue offering PMPs during the pandemic, as decreasing loneliness and increasing peer support are important for self-management of chronic pain.
Furthermore, this theme highlighted patients’ ability to gain knowledge and understanding during pain management therapy, despite being affected by COVID-19. Education is an aspect of PMPs that the British Pain Society10 guidelines state should be a low-intensity method to help promote self-management behaviour change. Participants who attended the adapted online PMPs gained knowledge from the educational aspects of the course which provided them with new skills and resources. These findings align with previous evidence which suggests existing eHealth pain management interventions improve pain experience46 and that patients feel eHealth is an effective way to gain knowledge about pain management.47 Therefore, although COVID-19 forced pain management services towards eHealth methods, this did not hinder patients’ ability to learn and gain a better quality of life from the content of the programme. Moreover, evidence suggests patients who have individual therapy do not receive the same level of educational resources and self-management skills-based knowledge. Therefore, moving forward, HCPs should ensure patients receiving individual therapy are also provided with the same education.
Nonetheless, the second theme, Limited the Experience, suggests eHealth delivery methods are not without their limitations. eHealth methods cause difficulties in building a therapeutic relationship between patients and HCPs. The therapeutic relationship is an interactive relationship that is professional yet caring, with clear boundaries that help make patients feel safe and comfortable.48 However, evidence suggests long periods on screen have been linked to disconnection49 and technical issues common in eHealth methods can be interpreted by patients as a lack of presence.17 This may limit participants’ experience, as therapeutic presence helps build stronger therapeutic alliances and increases the effectiveness of therapy.50 This differs from existing online PMPs where people have little to no live interaction. Existing literature suggests psychologists provide distanced weekly support instead of leading the programme for people in existing PMPs.29,30 Therefore, arguably, existing programmes do not create a strong therapeutic relationship, unlike the programmes created during COVID-19, which may impact patients' pain management outcomes.
Additionally, as participants had not chosen eHealth delivery methods, they were less prepared to use them. Some participants reported having a lack of equipment and technical knowledge which limited their experience of pain management therapy. eHealth tools only work if the individual is in a position to utilise them.51 If they are not, the risk of social health inequality increases for those unable to access support. Healthcare professionals should be aware of the potential risk of inequality and ensure when offering eHealth delivery methods that they provide patients with multiple access options, such as online conferencing with and without video, or telephone communication. However, during COVID-19, patients' access to face-to-face therapy was restricted and patients had no choice but to rely on the technical equipment and knowledge patients had on short notice.
Furthermore, the majority of participants voiced concerns about technical difficulties limiting their experience by causing difficulties in engagement. Previous literature proposes dropped connections and delays to audio and visual components during a PMP have been linked to increased difficulties.52 To reduce these issues, healthcare services should consider alternatives when using eHealth methods, for example, acess to technical support, to ensure patients gain the most out of eHealth methods of delivery.26 This type of contingency plan was not implemented in the pain psychology service in which the data were conducted., though this is likely to be a result of the rapid speed at which the service had to adapt during COVID-19. Moving forward, healthcare services using eHealth should consider the importane of access to technical support to reduce the barriers experienced by eHealth methods.
The third theme, COVID-19: A Double-Edged Sword identified participants feeling COVID-19 impacted them both positively and negatively. Daily activities that chronic pain patients used to manage their pain symptoms, such as gentle exercise or meeting with support groups, were impacted by the government lockdowns.53 Effects on daily activities, such as increased inactivity and self-isolation, potentially worsened patients' pain experience and increased the need for pain management therapy. However, despite the negative impacts on participants ability to continue self-management activities, participants felt COVID-19 introduced a new way to access pain management therapy with the use of eHealth delivery methods. Video and telephone communication allowed participants to gain the most out of their experience by enabling them to engage with the pain management therapy from a safe and comfortable environment, and reduced the cost and practical barriers associated with travel. These findings support previous literature which proposes eHealth improves patient’s comfort54 and increases accessibility.47 Therefore, many patients benefited from the move to eHealth delivery and embraced this change, as they felt it improved their pain experience.
Moving forward, it has been argued that patient preference towards the delivery method of their pain management therapy should be considered55 by healthcare services to allow them to gain the most out of their sessions. Further to this, eHealth methods reduced transmission of the virus56 and provided appropriate resources that protected vulnerable patients with severe, chronic conditions.53 Hence, in the aftermath of COVID-19, healthcare services should not rule out the use of eHealth delivery methods for pain management therapy, especially to those who are classed as vulnerable. Providing patients with the choice between face-to-face and online sessions potentially gives chronic pain patients more control over their self-management and pain experience outcomes.
Implications
eHealth modes of delivery for pain management therapy are constructive and valuable methods of delivery that can reduce the barriers that patients experience during face-to-face sessions. However, technical issues and issues with creating connections suggest healthcare services should be aware of how eHealth methods can create barriers to accessing sufficient pain management therapy and should build in contingencies against these. The results of this service evaluation encourage pain management services to consider providing patients with a choice between eHealth and face-to-face delivery methods post-COVID-19.
Limitations
The current study is not without its limitations. One limitation of the study is the lack of diversity in the age amongst participants. Chronic pain does not discriminate against age and evidence suggests the prevalence of chronic pain in 18–39-year-olds may be as high as 30% in the UK.2 The average age of participants was 56 years old, thus, the findings may not be representative of the pain population.
Future directions
Further studies may wish to compare the experiences and self-management outcomes of patients who attended the adopted PMPs during COVID-19 to those who attended existing online PMPs by choice during the pandemic. This could potentially advise healthcare services considering the use of eHealth for their pain management in factors that are effective in existing online PMPs. Future research should also look to the support groups that form after patients attend PMPs and consider their effectiveness.
Conclusion
The research has shown that eHealth methods are an appropriate form of delivering pain management therapy and allowed chronic pain patients to benefit from pain management therapy during the COVID-19 pandemic. The evidence supports existing literature that suggests there was an increased need for pain management therapy to continue during COVID-19, as a result of the negative impacts of lockdown restrictions and guidelines. This is useful for healthcare services as they move forward after the pandemic, and aim to increase accessibility and reduce barriers to pain management therapy for chronic pain patients. Overall, the COVID-19 pandemic demonstrated that eHealth methods can be used for pain management therapy and that patients are still able to gain important information and support in their pain management journey. It suggests that, moving forward, healthcare services should take into consideration eHealth methods when deciding how to deliver individual pain management therapy and PMPs after the pandemic.
Acknowledgements
The research team would like to give special thanks to Dr Rochelle Pinner who was instrumental in the development of the research question and provided invaluable support and information along the way. This research was not funded by any external institution.
ORCID iD
Olivia French https://orcid.org/0000-0003-4869-5035
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding: The author(s) received no financial support for the research, authorship, and/or publication of this article.
Data availability statement: The data that support the findings of this study are available from the corresponding author upon reasonable request.
==== Refs
References
1 Mills SEE Nicolson KP Smith BH . Chronic pain: a review of its epidemiology and associated factors in population-based studies. Br J Anaesth 2019; 123 (2 ): e273–e283.31079836
2 Fayaz A Croft P Langford RM , et al. Prevalence of chronic pain in the UK: a systematic review and meta-analysis of population studies. BMJ Open 2016: 6 (6 ): e010364.
3 GBD 2017 Disease and Injury Incidence and Prevalence Collaborators. Global, regional, and national incidence, prevalence, and years lived with disability for 354 diseases and injuries for 195 countries and territories, 1990–2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet 2018; 392 (10159 ): 1789–1858.30496104
4 Johannes CB Le TK Zhou X , et al. The prevalence of chronic pain in United States adults: results of an Internet-based survey. J Pain 2010; 11 (11 ): 1230–1239.20797916
5 Clauw DJ Essex MN Pitman V , et al. Reframing chronic pain as a disease, not a symptom: rationale and implications for pain management. Postgrad Med 2019; 131 (3 ): 185–198.30700198
6 Giusti E Pietrabissa G Manzoni GM , et al. The economic utility of clinical psychology in the multidisciplinary management of pain. Front Psychol 2017; 8 : 1860.29163260
7 Niknejad B Bolier R Henderson CR , et al. Association between psychological interventions and chronic pain outcomes in older adults: a systematic review and meta-analysis. JAMA Intern Med 2018; 178 (6 ): 830–839.29801109
8 Peters A Siripong N Wang L , et al. Enhanced recovery after surgery outcomes in minimally invasive nonhysterectomy gynecologic procedures. Am J Obstet Gynecol 2020; 223 (2 ): 234.e8.e1–234.e8.
9 Toelle TR Utpadel-Fischler DA Haas K-K , et al. App-based multidisciplinary back pain treatment versus combined physiotherapy plus online education: a randomized controlled trial. Npj Digit Med 2019; 2 (1 ): 1–9.31304351
10 The British Pain SocietyLocation. https://www.britishpainsociety.org/static/uploads/resources/files/pmp2013_main_FINAL_v6.pdf
11 Wilson IR . Management of chronic pain through pain management programmes. Br Med Bull 2017; 124 (1 ): 55–64.28927228
12 Knight LD Guildford BJ Daly-Eichenhardt A , et al. Assessment and patient selection process for a pain management programme: a case study in specialty care. Br J Pain 2019; 13 (2 ): 74–81.31019688
13 Shanthanna H Strand NH Provenzano DA , et al. Caring for patients with pain during the COVID-19 pandemic: consensus recommendations from an international expert panel. Anaesthesia 2020; 75 (7 ): 935–944.32259288
14 Burgess R . Moving from face-to-face to telephone assessments with patients during the COVID-19 pandemic. Nurs Stand 2021; 36 : 45–50.
15 Aafjes-van Doorn K Békés V Prout TA . Grappling with our therapeutic relationship and professional self-doubt during COVID-19: will we use video therapy again? Counselling Psychol Q 2020; 0 (0 ): 1–12.
16 Geller S . Cultivating online therapeutic presence: strengthening therapeutic relationships in teletherapy sessions. Counselling Psychol Q 2020; 34 : 1–17.
17 Javed S Hung J Huh BK . Impact COVID-19 on chronic pain patients: a pain physician’s Perspective. Pain Manag 2020; 10 (5 ):275–277. DOI: 10.2217/pmt-2020-0035.32772631
18 Jaremka LM Andridge RR Fagundes CP , et al. Pain, depression, and fatigue: loneliness as a longitudinal risk factor. Health Psychol 2014; 33 (9 ): 948–957.23957903
19 Clauw DJ Häuser W Cohen SP , et al. Considering the potential for an increase in chronic pain after the COVID-19 pandemic. Pain 2020; 161 (8 ): 1694–1697.32701829
20 Coluzzi F Marinangeli F Pergolizzi J . Managing chronic pain patients at the time of COVID-19 pandemic. Minerva Anestesiol 2020; 86 (8 ): 797–799.32400999
21 Puntillo F Giglio M Brienza N , et al. Impact of COVID-19 pandemic on chronic pain management: looking for the best way to deliver care. Best Pract Res Clin Anaesthesiol 2020; 34 (3 ): 529–537.33004164
22 Eccleston C Blyth FM Dear BF , et al. Managing patients with chronic pain during the COVID-19 outbreak: considerations for the rapid introduction of remotely supported (eHealth) pain management services. Pain 2020; 161 (5 ): 889–893.32251203
23 Fiani B Siddiqi I Lee SC , et al. Telerehabilitation: development, application, and need for increased usage in the COVID-19 era for patients with spinal pathology. Cureus 2020; 12 (9 ): e10563.33101809
24 O’Brien KM Hodder RK Wiggers J , et al. Effectiveness of telephone-based interventions for managing osteoarthritis and spinal pain: a systematic review and meta-analysis. PeerJ 2018; 6 : e5846.30397549
25 Walumbe J Belton J Denneny D . Pain management programmes via video conferencing: a rapid review. Scand J Pain 2021; 21 (1 ): 32–40.33027056
26 Greenhalgh T Wherton J Shaw S , et al. Video consultations for covid-19. BMJ 2020; 368 : m998.32165352
27 Griffiths F Lindenmeyer A Powell J , et al. Why are health care interventions delivered over the internet? A systematic review of the published literature. J Med Internet Res 2006; 8 (2 ): e10.16867965
28 Dear BF Titov N Perry KN , et al. The Pain Course: a randomised controlled trial of a clinician-guided Internet-delivered cognitive behaviour therapy program for managing chronic pain and emotional well-being. Pain 2013; 154 (6 ): 942–950.23688830
29 Dear BF Nicholson Perry K Siddall P , et al. The pain course: exploring the feasibility of an internet-delivered pain management programme for adults with spinal cord injury. Spinal Cord 2018; 56 (10 ): 931–939.29855612
30 Turner-Stokes L Erkeller-Yuksel F Miles A , et al. Outpatient cognitive behavioral pain management programs: a randomized comparison of a group-based multidisciplinary versus an individual therapy model. Arch Phys Med Rehabil 2003; 84 (6 ): 781–788.12808527
31 Nevedal DC Wang C Oberleitner L , et al. Effects of an individually tailored web-based chronic pain management program on pain severity, psychological health, and functioning. J Med Internet Res 2013; 15 (9 ): e2296.
32 Moore F Williams L Dunbar M . Sociodemographic predictors of attendance at a Scottish pain management programme. Br J Pain 2020; 15 : 2049463720970579.
33 Ruehlman LS Karoly P Enders C . A randomized controlled evaluation of an online chronic pain self management program. Pain 2012; 153 (2 ): 319–330.22133450
34 Braun V Clarke V . Successful qualitative research: a practical guide for beginners. 2013.
35 DeJonckheere M Vaughn LM . Semistructured interviewing in primary care research: a balance of relationship and rigour. Fam Med Com Health 2019; 7 (2 ): e000057.
36 Gill P Stewart K Treasure E , et al. Methods of data collection in qualitative research: interviews and focus groups. Br Dent J 2008; 204 (6 ): 291–295.18356873
37 Roulston K Choi M . Qualitative interviews. in: the SAGE handbook of qualitative data collection [Internet]. 55 City Road: SAGE Publications Ltd; 2018: 233–249. Available from: https://sk.sagepub.com/reference/the-sage-handbook-of-qualitative-data-collection/i1709.xml.[cited 2021 Aug 18]
38 Braun V Clarke V . Using thematic analysis in psychology. Qual Res Psychol 2006; 3 (2 ): 77–101.
39 Braun V Clarke V Weate P . Using thematic analysis in sport and exercise research. Available from: https://uwe-repository.worktribe.com/output/3123247/using-thematic-analysis-in-sport-and-exercise-research. 2016 Sep 1.[cited 2021 Aug 18].
40 King N . Using templates in the thematic analysis of text. In: Essential Guide to Qualitative Methods in Organizational Research [Internet]. London, UK: SAGE Publications Ltd, 2021, pp. 256–270. Available from: https://sk.sagepub.com/books/essential-guide-to-qualitative-methods-in-organizational-research/n21.xml.
41 Braun V Clarke V . Thematic analysis. Available from: https://uwe-repository.worktribe.com/output/841297/thematic-analysis 2015 Jan 1.[cited 2021 Aug 18]
42 Nowell LS Norris JM White DE , et al. Thematic analysis: striving to meet the trustworthiness criteria. Int J Qual Methods 2017; 16 (1 ): 1609406917733847.
43 Johnathan AS . Qualitative psychology: a practical guide to research methods [Internet]. 3rd edition. Thousand Oaks, CA: SAGE Publications Ltd. Available from: https://us.sagepub.com/en-us/nam/qualitative-psychology/book242733.
44 The British Psychological Society. BPS Code of Human Research Ethics [Internet]. The British Psychological Society. https://www.bps.org.uk/sites/www.bps.org.uk/files/Policy/Policy%20-%20Files/BPS%20Code%20of%20Human%20Research%20Ethics.pdf (2021.Available from
45 Nieto R Pardo R Sora B , et al. Impact of COVID-19 lockdown measures on spanish people with chronic pain: an online study survey. J Clin Med 2020; 9 (11 ): 3558.33167322
46 Du S Liu W Cai S , et al. The efficacy of e-health in the self-management of chronic low back pain: a meta analysis. Int J Nurs Stud 2020; 106 : 103507.32320936
47 Ledel Solem IK Varsi C Eide H , et al Patients’ needs and requirements for ehealth pain management interventions: qualitative study. J Med Internet Res 2019; 21 (4 ): e13205.30877780
48 Ridling DA Lewis-Newby M Lindsey D . Chapter 9-family-centered care in the pediatric intensive care unit. In: Fuhrman BP Zimmerman JJ (eds). Pediatric Critical Care. 4th Edition. Saint Louis, MO: Mosby, 2021, pp. 92–101. Available from: https://www.sciencedirect.com/science/article/pii/B9780323073073100096.
49 Dodgen-Magee D . Deviced! balancing life and technology in a digital world. Lanham, MD: Rowman & Littlefield Publishers, 2018.
50 Dunn R Callahan JL Swift JK , et al. Effects of pre-session centering for therapists on session presence and effectiveness. Psychother Res 2013; 23 (1 ): 78–85.23066649
51 Latulippe K Hamel C Giroux D . Social health inequalities and ehealth: a literature review with qualitative synthesis of theoretical and empirical studies. J Med Internet Res 2017; 19 (4 ): e136.28450271
52 Palyo SA Schopmeyer KA McQuaid JR . Tele-pain management: use of videoconferencing technology in the delivery of an integrated cognitive-behavioral and physical therapy group intervention. Psychol Serv 2012; 9 (2 ): 200–202.22662734
53 El-Tallawy SN Nalamasu R Pergolizzi JV , et al. Pain management during the COVID-19 pandemic. Pain Ther 2020; 9 (2 ): 453–466.32840756
54 Widberg C Wiklund B Klarare A . Patients’ experiences of eHealth in palliative care: an integrative review. BMC Palliat Care 2020; 19 : 158.33054746
55 So H Szeto C-C Tam L-S . Patient acceptance of using telemedicine for follow-up of lupus nephritis in the COVID-19 outbreak. Ann Rheum Dis 2021; 80 (6 ): e97–e97.32581085
56 Monaghesh E Hajizadeh A . The role of telehealth during COVID-19 outbreak: a systematic review based on current evidence. BMC Public Health 2020; 20 (1 ): 1193.32738884
| 0 | PMC9703013 | NO-CC CODE | 2022-11-29 23:21:05 | no | Br J Pain. 2022 Nov 24;:20494637221121703 | utf-8 | Br J Pain | 2,022 | 10.1177/20494637221121703 | oa_other |
==== Front
spthr
THR
Tourism and Hospitality Research
1467-3584
1742-9692
SAGE Publications Sage UK: London, England
10.1177_14673584221141294
10.1177/14673584221141294
Article
Retaining talented employees during COVID-19 pandemic: The leverage of hotel pandemic response strategies
https://orcid.org/0000-0002-9483-3550
Salem Islam Elbayoumi
Business Administration Department, College of Economics and Business Administration, 144882 University of Technology and Applied Sciences , Salalah, Oman; 496089 Alexandria University, Faculty of Tourism and Hotels , Alexandria, Egypt
Aideed Hassan
Alkathiri Nasser A
Business Administration Department, College of Economics and Business Administration, 144882 University of Technology and Applied Sciences , Salalah, Oman
https://orcid.org/0000-0002-8836-8939
Ghazi Karam Mansour
High Institute of Tourism and Hotels in Alexandria (EGOTH) , Alexandria, Egypt
Islam E Salem, Alexandria University, Faculty of Tourism and Hotels, Mostafa Musharafa, Alexandria 340, Egypt. Emails: [email protected]; [email protected]
24 11 2022
24 11 2022
14673584221141294© The Author(s) 2022
2022
SAGE Publications
This article is made available via the PMC Open Access Subset for unrestricted re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the COVID-19 pandemic or until permissions are revoked in writing. Upon expiration of these permissions, PMC is granted a perpetual license to make this article available via PMC and Europe PMC, consistent with existing copyright protections.
This study aims to examine the impact of hotels' pandemic response strategies (service automation, downsizing, restructuring, health protection, and training) on talent retention intentions with the mediation of talent satisfaction and moderation of job insecurity in 4- and 5-star hotels. The sample was composed of 357 talented hotel employees. Findings reveal that automation services, health and safety, and training support were found to positively affect talents' satisfaction with response strategies and favourably enhance talents' retention intention through the mediating role of talents' satisfaction. The findings also suggest that high job insecurity would undermine the positive impact of talent satisfaction on retention intentions. The study contributes to the existing literature by providing theoretical and practical implications in the hotel context and directions for future research.
Talented employee
COVID-19
pandemic response strategies
talents' satisfaction
job insecurity
talents retention intention
edited-statecorrected-proof
typesetterts10
==== Body
pmcIntroduction
The global hospitality industry faces several challenges such as talent shortage and high staff turnover, which directly impact costs, profitability, competitiveness, service quality, brand, investment, and ultimately future growth (Baum et al., 2020; Elbaz et al., 2022; Fuentes, 2021; Jooss et al., 2019, 2022; Kravariti et al., 2022). An extensive survey of 1317 hospitality talents found that 9 out of 10 hospitality talents are actively looking for a job, 27% of senior talents and 31% of current hospitality students are unsure or unwilling to work in hospitality again even after rebounding (Fuentes, 2021). The talent shortage problem could cost the global economy 14 million jobs and nearly US$610 billion in GDP by 2024 (WTTC, 2015). It could mean a significant loss of presently qualified talent and the loss of an irreplaceable future asset for all stakeholders (Fuentes, 2021; Jooss et al., 2019, 2022; Kravariti et al., 2022).
In addition to the talent shortage and high turnover problem, the COVID-19 pandemic has heavily hit the hospitality industry, resulting in millions of employees being laid off (victims), and other remaining employees (survivors) being challenged by hard drops in their yield and different changes in their work and lives with extremely high levels of job insecurity and physical and psychological risks (Jung et al., 2021; Vo-Thanh et al., 2021). Fundamental crises (like the COVID-19 pandemic) inevitably lead to major changes in an organisation’s work environment and force most hospitality organisations to apply more flexible and innovative transitional strategies that include downsizing to labour cost reduction, and flexible programmes of using employees (e.g. part-time or temporary jobs), carrier automation, and health protection policies in their attempt to overcome the pandemic (Elkhwesky et al. 2022).
These temporary strategies that are conducted because of the pandemic of COVID-19 are generally risky and foster survivors’ negative feelings of anxiety, stress, job insecurity, distrust, powerlessness, loss of morale and motivation, and even burnout regarding their jobs (Bajrami et al., 2021; Jung et al., 2021; Vo-Thanh et al., 2021).
The talent shortage and uncertainty of employment in the hospitality industry, which has been increased by the COVID-19 pandemic, therefore poses an immediate threat to organisational performance and viability (survival), an unprecedented situation that requires hospitality organisations to seek a variety of solutions to best retain their skilled/talented staff to remain competitive and sustained in the future (Jooss et al., 2019, 2022; Jung et al., 2021; Kravariti et al., 2022; WTTC, 2015). Despite retaining skilled (talented) staff in the hospitality industry having received increased attention from both academics and practitioners (Deery and Jago, 2015), research on talent management topics in the hospitality sector remains limited (Jooss et al., 2019, 2022; Kravariti et al., 2022; Sheehan et al., 2018; Vasquez, 2014). Further, although the impact of the COVID-19 pandemic on the hospitality industry is fairly well documented from customers’ perspectives (Ritchie and Jiang, 2019), there is a lack of studies on the impact of hotels' pandemic response strategies on employees’ (survivors’) attitudes to remain competitive and sustained in the future (Baum et al., 2020; Jung et al., 2021; Vo-Thanh et al., 2021).
The current study fills this gap by using a nonlinear Partial Least Squares Structural Equation Modeling (PLS-SEM) to examine the impact of the hotel’s pandemic response strategies (service automation, responsible downsizing, restructuring, health protection, and training) as a part of Corporate Social Responsibility (CSR) on talent retention intentions with the mediation of talents’ satisfaction and moderation of job insecurity in 4- and 5-star hotels. Research literature (Deery and Jago, 2015; Jooss et al., 2019, 2022; Kravariti et al., 2022; Sheehan et al., 2018; Vasquez, 2014) states that despite rising interest from academics and practitioners in keeping competent (talented) workers in the hotel industry, research on talent management (retention) in the hospitality sector remains scarce. Furthermore, Ritchie and Jiang (2019); Baum et al. (2020); Jung et al. (2021); Vo-Thanh et al. (2021) claim that while the impact of the COVID-19 pandemic on the hospitality industry from the perspective of customers is fairly well documented, studies on the impact of hotel pandemic response strategies on employees' (survivors') attitudes to remain competitive and sustained in the future are lacking. CSR is defined by Font and Lynes (2018), p. 1028) as “a process whereby individuals identify stakeholder demands on their organisations and negotiate their level of responsibility towards the collective wellbeing of society, environment, and economy”.
Limited investigations on CSR place considerable priority on internal stakeholders, instead the majority focused on organizational and macro-level inquiry (Zainee and Puteh, 2020). However, prior work has recommended that in order to retain talented employees and secure long-term sustainability, firms should integrate their employee retention practises with CSR, such as providing training and certain steps/protocols related to safety and health (Glavas, 2016; Zainee and Puteh, 2020). In light of this, implementing CSR procedures within the company could be considered as a key element in enhancing talent retention. Drawing on multiple theoretical perspectives, the Conservation of Resources Theory (Hobfoll, 1989), Motivation Theory (Maslow, 1943), Organisational Support Theory (Eisenberger et al., 1986), Social Exchange Theory, and Self-Determination Theory (Ryan and Deci 2000), the study proposed that the hotel’s effective COVID-19 responses may play an important role in attaining its sustainable development via reducing perceived job insecurity and enhancing retention of talented staff through employees’ satisfaction Salem et al. (2022). Thus, the paper is a narrative study that adds valuable contributions to identifying the most effective pandemic response strategies related to talents’ attitudes of satisfaction, job insecurity, and retention intention. The research conceptual framework is shown in Figure 1.Figure 1. Conceptual framework.
Literature review and hypothesis development
Automation services in the hospitality sector (Automation as a response strategy)
The COVID-19 pandemic has made tourism businesses further accelerate the implementation of smart tourism services — using Artificial Intelligence [AI], robots, and digital/automated services— (Elkhwesky and Elkhwesky, 2022; Fusté-Forné and Jamal, 2021). Such an expansion of automated services in the tourism industry has received both scientific (research) and practical recognition (Fusté-Forné and Jamal, 2021; Koo et al., 2021). The transformation from manpower to automation services/AI is expected to make hotel operations more efficient, reduce costs related to employees, and optimise the experience of customers (Fusté-Forné and Jamal, 2021; Koo et al., 2021). Automation in the hotel industry exists in two forms: substitutive (e.g. robots delivering food and drinks) and supportive - e.g. online payments - (Fusté-Forné and Jamal, 2021). A number of demanding reasons have further stimulated the growth of automated services/AI in the hotel industry, taking the examples of lack of labour, enhancing tourists’ experience, cutting costs, customising guests’ needs/expectations, and providing services around the clock −24/7 (Fusté-Forné and Jamal, 2021; Koo et al., 2021).
Hospitality 5.0, which comprises advanced contactless technology (e.g. automation, robots, mobile phone technology, AI, Augmented Reality, and Virtual Reality), can be an effective tool during (and after) pandemic times by preventing virus transmission (social distancing) and ensuring both guests' and employees' health and safety (Pillai et al., 2021). Renowned international hotel chains such as Hilton Worldwide, Marriott International, and Sheraton have introduced Hospitality 5.0 technologies by using AI and robots (Koo et al., 2021). With a variety of technological options (Hospitality 5.0), choosing the right technology is a challenge in itself, where such technologies should complement hotel employees’ work rather than replace them (Fusté-Forné and Jamal, 2021).
While various departments of the hotel industry are expected to benefit from automated services/AI, there are particular services which are expected to be automated first, relating to maintenance and housekeeping departments (Fusté-Forné and Jamal, 2021). To further add, AI is expected to replace man-powered jobs (service sector) which are of an analytical and mechanical nature, whereas humans will still be wanted (difficult to automate) in jobs where the characteristics of empathy and intuition are needed (Koo et al., 2021). Service automation in the hospitality sector is a double-edged sword where the adoption of Hospitality 5.0 technology will significantly benefit the hotel, customers, and employees (Fusté-Forné and Jamal, 2021; Koo et al., 2021). Still, it can put consumers' personal information at risk of privacy breaches, surveillance, and data mining (Fusté-Forné and Jamal, 2021). Furthermore, hotel low-skilled employees will feel threatened by becoming redundant as a result of AI/robots (Koo et al., 2021). This subsequently stimulates their turnover intent and sparks a sense of job insecurity (Koo et al., 2021). Based on the above discussion, the following hypothesis is proposed:
Hypothesis 1 Automation services are significantly related to talented satisfaction with response strategies.
Downsizing as a response strategy
The COVID-19 outbreak along with the precautionary measures imposed (e.g. travel restrictions) has severely affected the hospitality sector, forcing hotels to downsize (partial or full) the number of employees in an attempt to cut costs (Abuelnasr, 2020). Downsizing is defined by McDevitt et al. (2013, as cited in Abuelnasr (2020), p. 168) as “a global management strategy that is purposively undertaken to reduce organization employees”. Employees who escape termination in the downsizing process (remain in the organization) are referred to as “survivors” (Abuelnasr, 2020). The anticipated success or failure of an organization’s downsizing process relies on how survivors respond to pre and post-downsizing (Abuelnasr, 2020).
The downsizing strategy resulting from COVID-19 is a risky attempt for organizations, where comprehending the potential shift in survivors' attitudes could help reduce this risk (Abuelnasr, 2020). While survivors are the lucky ones to continue working in the organisation in turbulent situations (i.e. COVID-19), they may experience some negative impacts resulting from the downsizing process. To illustrate, downsizing strategies can make employees feel pressured, lack of motivation and morale, discontent, angry, sick, concerned, job insecurity, and nervous, consequently affecting their job performance (Abuelnasr, 2020; Vo-Thanh et al., 2021). Abuelnasr (2020) claimed that survivors after downsizing (including those with investments in the working organisations) might not choose to continue working in the organization (seeking another career), especially when they are faced with salary cuts and compulsory leaves (Abuelnasr, 2020). Based on the above discussion, the following hypothesis is proposed:
Hypothesis 2 Downsizing is significantly related to talented satisfaction with response strategies.
Restructuring as a response strategy
The COVID-19 pandemic has made organizations’ Human Resources (HR) departments make some disruptive alterations, among them employee restructuring (Biron et al., 2021). Such drastic changes (i.e. restructuring) aggravated by the pandemic can be seen by organisations as a viable solution, although it can increase employees’ sense of job insecurity (Biron et al., 2021; Vo-Thanh et al., 2021). Based on the above discussion, the following hypothesis is proposed:
Hypothesis 3 Restructuring is significantly related to talented satisfaction with response strategies
Health and safety as a response strategy
During a crisis, particularly in the case of COVID-19, the issue of health and safety (particularly hotel employees) has received more attention (Salem et al., 2022). In the workplace, employees are occasionally subject to injuries and accidents, whereas during pandemics hotels should implement emergency procedures to ensure employees’ health and safety (Salem et al., 2022). This is done by the tourism/hospitality sector by following certain steps/protocols related to safety, health screening, disinfection, hygiene, and cleanliness (Salem et al., 2021; Pillai et al., 2021). During pandemics, technology can play an important role by providing contactless services that reduce the risk of contamination. These services include smart control of in-room facilities (such as lighting, air conditioning, door entry, and TV), online payments, automated check-in and check-out, and robotic room food delivery (Pillai et al., 2021; Salem et al., 2022).
Throughout pandemics, official competent health and safety authorities and experts can assist hotels by providing guidelines, recommendations, information, procedures, and training to keep employees healthy and safe from disease (Salem et al., 2021). Such initiatives, established by governmental health authorities to provide hotels with guidelines to deal with the COVID-19 situation, have shown their effectiveness in protecting hotel employees from the pandemic (Salem et al., 2021). Furthermore, such authorities can help protect hotel employees’ health and safety by supplying them with virus detection and protection (e.g. masks) equipment, as well as sterilisers (Salem et al., 2021). Through training (preferably online to prevent infection), hotels can ensure that their employees are familiar with such disease-preventive and hygienic measures to safeguard hotel customers and employees’ health and safety (Salem et al., 2021).
When the pandemic begins to subside, the implemented health and hygienic practises should not be halted, and employees should be regularly educated about them through training courses and meetings (Pillai et al., 2021; Salem et al., 2021). In uncertain situations (COVID-19), putting in place health and safety measures does not guarantee that every employee will follow them (Bajrami et al., 2021). Some employees might ignore or only partially follow the rules just to keep their job. Despite the stress and anxiety caused by the COVID-19 pandemic, the stringent health and safety measures have worsened the situation for service sector employees, making them feel more frustrated and worried (Bajrami et al., 2021). As recommended by international health organisations (i.e. World Health Organization) to utilise contactless services during the COVID-19 pandemic, hotels are expected to continue using such contactless services (Hospitality 5.0 technologies) post-pandemic, becoming the new norm in the hotel sector (Pillai et al., 2021). Based on the above discussion, the following hypothesis is proposed:
Hypothesis 4 Health and safety are significantly related to talented satisfaction with response strategies.
Training support as a response strategy
Training is deemed vital if an organisation seeks to retain employees in the long run (Vasquez, 2014). Retention of employees is not the only benefit hotels attain from training courses; it also helps to decrease unsatisfactory performance (improve productivity) and resolve conflicts among employees and between employees and hotel guests (Vasquez, 2014). Despite its high financial cost, hotels’ neglection of training, especially cross-cultural training, can result in the employees' lacking essential skills, consequently leading to the failure of services (Scott, 2016; Vasquez, 2014). As a result (i.e. high cost of training), some hotels during a crisis defer training, especially when there is no government support in this regard (Salem et al., 2021).
Hotel managers can benefit from training by gaining critical knowledge and skills in employee leadership, whereas employee enrollment in training courses improves morale, productivity, and work satisfaction (García et al., 2022; Salem et al., 2021; Vasquez, 2014). In contrast, hotel employees who were offered a limited number of poor training courses showed low levels of job satisfaction (Salem et al., 2021). Recognizing the value of formal education, governments have begun to fund apprenticeship programs, and the number of universities offering hospitality and tourism degrees is increasing (Sheehan et al., 2018). Such official attention towards education and training in the hospitality sector should be viewed as highly important, taking into account the size and the high job opportunities that the hospitality sector offers to the community (Sheehan et al., 2018). Based on the above discussion, the following hypothesis is proposed:
Hypothesis 5 Training support is significantly related to talented satisfaction with response strategies.
Retention of talented employees’
Nowadays, employee retention has become an important aspect of organizations’ business strategies (Chee, 2017; Scott, 2016). Organizations often encounter the issue of high turnover, which leaves them with a more difficult situation — a lack of talented and skilful employees (Chee, 2017; Scott, 2016). Mai and Thuy, 2021: p. 49) describes talented people as “those who have outstanding qualities and abilities to be able to undertake a job or a difficult field of activity, complex and successful, efficient, very high quality, sometimes the highest in a certain range”. As hotels eagerly seek to recruit talented employees (for their future existence), retaining them has become the main challenge, and hotel management should pay great attention to this issue (Chee, 2017). Failing to retain employees can negatively affect the hotel in terms of quality of customer service, social responsibility, organisation morale, job satisfaction, achieving the hotel’s set goals, productivity, competitive advantage, and organisational performance (Chee, 2017; Mai and Thuy, 2021; Scott, 2016; Vasquez, 2014). Despite the challenges hotels face in retaining employees, it can be done if the hotel management demonstrates a sense of cooperation, leadership, and commitment (Vasquez, 2014). In addition, the non-significant association between downsizing strategy and talents' satisfaction and retention intention could indicate that organizations in their downsizing strategy are not affecting the talented staff due to their outstanding qualities and abilities to accomplish sophisticated and productive tasks (Mai and Thuy, 2021).
To retain talented employees in the hospitality sector for longer periods, several retention strategies need to be implemented (Vasquez, 2014). Such employee retention strategies can include development and training programs, showing support and value, aligning employees’ job aspirations with job tasks, increasing wages, establishing a cooperative environment between employees and management, work/life balance; employee involvement, promotions, benefits, incentives, and appropriate work atmosphere (Chee, 2017; Kichuk, 2017; Salem et al., 2021; Scott, 2016; Vasquez, 2014). Based on the above discussion, the following hypothesis is proposed:
Hypothesis 6 Talented satisfaction with response strategies is significantly related to talented retention intention.
Moderating employee job insecurity
The COVID-19 pandemic has played an effective role in aggravating the job loss perception among hotel employees, which does not only affect individuals but extends to include communities and organisations (Abuelnasr, 2020; Biron et al., 2021; Jung et al., 2021). Abuelnasr (2020): p. 169) explains job insecurity as “a perceived threat to the continuity and stability of employment as it is currently experienced”. Despite being survival employees during the pandemic, survivors’ perceptions of job insecurity should be examined, as they may well experience job insecurity symptoms/implications such as salary cuts and a challenging work environment as a result of the imposed COVID-19 preventive measures (Abuelnasr, 2020). A number of reasons are attributed to the feeling of job security among employees, for instance, ineffective downsizing approaches, layoffs, future uncertainty, organisation restructuring, lack of trust and commitment, and the possibility of being replaced by AI (Abuelnasr, 2020; Jung et al., 2021; Koo et al., 2021). Employees holding a sense of job insecurity will affect the organisation and employees in terms of productivity, punctuality, efficiency, commitment, job satisfaction, performance, absenteeism, and turnover intent (Abuelnasr, 2020; Chen and Eyoun, 2021; Jung et al., 2021; Koo et al., 2021). The effects of job insecurity on employees are not confined to their workplace environment, but can negatively affect their psychological, mental, and physical health-e.g. anxiety and frustration- (Chen and Eyoun, 2021; Jung et al., 2021; Koo et al., 2021). Based on the above discussion, the following hypothesis is proposed:
Hypothesis 7 Job insecurity moderates the relation between talent satisfaction with response strategies and talent retention intention.
Mediating role of employee satisfaction
As employees spend long times at work, job satisfaction becomes an important matter for them (Salem et al., 2021). The hotel industry is known for experiencing high turnover, job discontent, a stressful workplace, and lengthy work hours (Kichuk, 2017; Koo et al., 2021). Achieving job satisfaction is among organizations’ top goals for success, where it (i.e. job satisfaction) can be an indicator of employees’ turnover intention (Chee, 2017; Kichuk, 2017; Mai and Thuy, 2021; Scott, 2016; Wong et al., 2021). Employees’ job satisfaction is about retaining talented workforces; therefore, employee satisfaction in the service sector is considered an important element (Ashton, 2017; Kaewsaeng-on, 2016; Kichuk, 2017; Koo et al., 2021). Wong et al. (2021): p. 2) describes job satisfaction as “pleasurable emotional state resulting from the appraisal of one’s job as achieving or facilitating the achievement of one’ s job values”.
Employees become satisfied at work when several elements exist, such as receiving sufficient benefits, life and work balance, fair treatment, a healthy and stress-free environment, and empowerment (Ashton, 2017; Salem et al., 2021; Wong et al., 2021). Hotels which apply retention strategies will have satisfied employees, subsequently showing an improvement in customer service, better profit, satisfied customers, employees demonstrating good behaviour, and the organisation gaining a competitive advantage (Jooss, et al., 2019; Salem et al., 2021; Vasquez, 2014). Hotel managers (including HR) can play a role in making employees feel satisfied at the workplace. This is by exhibiting leadership traits such as commitment to provide outstanding services, inspiring and motivating employees, good attitude, justice, providing health support (e.g. insurance), improving skills through training, promotions, ethical environment, transparency, involvement, rotating positions, being supportive, fostering good moral values, and giving employees’ individual needs appropriate attention (Ashton, 2017; Kaewsaeng-on, 2016; Kichuk, 2017; Salem et al., 2021; Scott, 2016). On the other hand, work colleagues can influence job satisfaction by showing good behavior, which in turn creates a favourable working environment, synergies teamwork and encourages team commitment and collaboration (Ashton, 2017). Having unsatisfied employees at work can result in some negative outcomes. For instance, high turnover, negative attitudes at work, poor performance, and morale drop, subsequently negatively affecting an organization’s profitability (Ashton, 2017; Chee, 2017; Mai and Thuy, 2021).
Based on the above discussion, the following hypotheses are proposed:
Hypothesis 8.1 Satisfaction with response strategies mediates the relation between automation and retention intention.
Hypothesis 8.2 Satisfaction with response strategies mediates the relation between downsizing and retention intention.
Hypothesis 8.3 Satisfaction with response strategies mediates the relation between restructuring and retention intention.
Hypothesis 8.4 Satisfaction with response strategies mediates the relation between response training and retention intention.
Hypothesis 8.5 Satisfaction with response strategies mediates the relation between health and safety and retention intention.
Research methodology
Sampling design and data collection
The data for this study was gathered through a survey of talented employees in Egyptian 4- and 5-star hotels. In order to select the talented respondents, this study applied purposeful sampling, which allows the researcher to choose a case because it fulfils a set of criteria (Patton, 2015; Silverman, 2017). The criteria to select talent respondents were guided by the literature on talent management and by the recommendation of the head of HR or the heads of talent management. Consistent with talent management literature, the study used the following criteria: (a) being a manager or employee with high potential and high performing; (While specific definitions depend on organisational contexts, a strong focus is placed on high performers and/or high potentials); (b) having at least 12 months of experience in the organization, and (c) being retained during the COVID-19 pandemic period consistent with the study context to examine the relationships of response strategies with talent retention (Collings and Mellahi, 2009; Jooss et al., 2019, 2022; Kravariti, et al., 2022; Mai and Thuy 2021; Tansley, et al., 2007; Zhang et al., 2014).
Both the talent management and hospitality literature reveal a lack of clarity, conceptualisation and theorisation, along with a significant lack of empirical evidence with regard to the talent definition and identification processes (criteria) (Jooss, et al., 2019, 2022; Kravariti, et al., 2022). It also found that talents differ to an extent across the various hospitality and tourism sectors and countries, suggesting that talent’s operationalisation is context-dependent. Given that context (e.g. country, industry, sector) impacts the operationalisation of talent and talent management processes (criteria). The definition and identification of “talent” (criteria) depend quite a lot on the approaches of the researcher, the context, and the specifics of the organisation (Mai and Thuy, 2021; Jooss et al., 2022; Kravariti et al., 2022). While talent definition and identification (criteria) depend on organisational contexts, a strong focus is placed on high performers and/or high potentials (Collings and Mellahi, 2009; Jooss et al., 2019, 2022). Talent is predominantly viewed as high-performing, high-potential talent, which can contribute considerably to the organisation.
Thus, the first main criterion is the high-performing, high-potential individual. In line with Tansley et al. (2007), an immediate contribution through high performance or in the longer term by showing high potential was a key factor in their understanding of talent. Based on the data analysis of extant literature, the criteria to identify talent can be clustered into seven broad areas: competency framework, intellectual abilities, education, experience, performance, potential, and readiness (Jooss et al., 2019, 2022; Kravariti et al., 2022). Thus, the second main criterion is the talent experience in the organization. The third criterion has to do with the purpose of the study, which is to test the relationships between pandemic response strategies and talent behaviours of satisfaction and retention during pandemic periods. These criteria were selected as the study aims to examine the impact of the hotel’s pandemic response strategies on talent retention intention during COVID-19, in which the talent respondent should have been there during this pandemic and have at least 12 months of experience in the organisation. In this research case, talented employees are selected as the target sample due to the fact that during the peak outbreak of COVID-19, the hospitality industry severely suffered from the pandemic due to lockdown measures. Such measures did not only impose a major challenge to the hotel from a financial perspective but also resisted the retention of talented employees with whom they had heavily invested in recruiting and training.
Talented hotel employees are essential to the successful operation of a hotel, as hotels rely on them to provide high-quality service to hotel guests. Losing talented employees would have been a great hit for hotels at the time, taking into account the costly and time-consuming process of re-recruiting them. Consequently, examining how hotels used effective retention strategies to retain talented employees during the COVID-19 period would be a significant insight, a valuable contribution to knowledge, and have beneficial practical implications. The COVID-19 period was particularly examined in this research case since the hospitality sector was one of the main service sectors to be severely affected by the pandemic (e.g. lockdowns). Among the inevitable consequences the hotel had to bear, besides financial loss/total collapse, was talented employees’ turnover. This has exacerbated the hotel industry’s challenges, which include mitigating financial risks and retaining talented employees.
Using purposeful sampling, the head of HR or the heads of talent management acted as gatekeepers by approving the research and assuring access to respondents in the hotel (Creswell, 2014). Provided with the three criteria as outlined above, they selected respondents who they deemed appropriate. The researcher relied on the judgement of the gatekeepers of the study to identify suitable participants. An email was sent to HR managers or talent managers of all 4- and 5-star hotels in Egypt, inviting them to take part in the research by sending a link to the survey to all of their talented employees. All the responses were collected via a web-based survey. The managers sent a link to the survey only to those who were still employed during the pandemic period and who met the above three criteria.
The sampling procedure was conducted in May and June 2021 the during COVID-19 pandemic time by email. This method of data collection was chosen to avoid the risks of infection for researchers and participants. All questionnaires were completed voluntarily by respondents. First, the authors went through the Egyptian Hotel Guide (Egyptian Hotel Association, 2020) to identify appropriate hotel properties. Second, the authors contacted the human resources directors (HRDs) of the identified hotels by email. In the email, the authors explained the purpose of the study and the type of employees targeted for the study and requested the HRDs' permission and assistance in distributing the survey to the talented employees in their hotel. Third, the survey was distributed to the HRDs of the participating hotels in the month of May 2021. The HRDs then distributed the survey to employees within their hotels who were meeting the above three criteria. Fourth, a reminder email was sent to the HRDs one and 3 weeks after the survey was distributed to the hotels. The final data for inclusion in the study was collected before the end of June 2021. The survey was sent out to 357 potential candidates, in this case, employees who had been formally identified by the hotel (HR manager) as meeting the above three criteria.
The final sample was composed of 357 talented hotel employees, deemed suitable for studies using a quantitative approach (Hair et al., 2014). The descriptive profile of the respondents is presented in Table 1. In terms of hotels, the majority (63.9%) worked in 5-star hotels, with 36.1% working in 4-star hotels. In terms of the number of rooms, 56% had 150 or more, while 33.1% had 100 or less than 150 rooms. The respondents were from various hotel departments, including the kitchen (34.5%), front office (18.5%), restaurant (16%), accounting (7.6%), marketing (6.4%), housekeeping (4.5%), and HR (2.1%). Regarding gender, 260 were male (72.8%) and 97 were female (27.2%). The majority of respondents were between the ages of 18 and 25 years old (90.8%). As for work experience, the majority (55%) had 3 years or more of professional experience, and others (45%) had 1 year or less than 3 years of experience.Table 1. Sample characteristics (N = 357).
Characteristics Percent Characteristics Percent
Experience Hotel type
One year and less than 3 years 161 (45%) Four stars 129 (36.1%)
3 years and more 196 (55%) Five stars 228 (63.9%)
Age Number of rooms
18–25 Years 324 (90.8%) Less than 100 39 (10.9%)
26–40 Years 28 (7.8%) 100 and less than 150 118 (33.1%)
40–55 Years 2 (0.6%) 150 and more 200 (56%)
More than 55 Years 3 (0.8%)
Gender
Male 260 (72.8%)
Female 97 (27.2%)
Department
Front office 66 (18.5%)
Kitchen 123 (34.5%)
Restaurant 57 (16%)
Housekeeping 16 (4.5%)
Accounting 27 (7.6%)
Marketing 23 (6.4%)
HR 8 (2.1%)
Others 37 (10.4)
Survey measures and development
The survey consisted of eight sections. In the first section, the purpose of the study was presented. Furthermore, a statement about the importance of answering the survey questions was presented. In addition, there was a statement concerning the privacy of the respondent’s answers and the guarantee of the respondents’ anonymity. At the end of this section, the respondents were asked to consent to participating in the study. All the surveys included in the final data collection were from consenting respondents only. The second to seventh sections concentrated on the survey measures, namely automation (AUT), downsizing (DOW), restructuring (RES), training support (TRN), health and safety (HES), talented satisfaction with organisation COVID-19 responses (TSRS), job insecurity (JIS), and talent retention intention (TRET).
These measures were developed from prominent scales to ensure reliability and validity. The AUT construct was created by combining four items from Lukanova and Ilieva (2019), Zeng et al. (2020), and Ivanov et al. (2020). The DOW construct was developed using five items and the RES construct was developed using two items, both adapted from Santana et al. (2017); Thumiki et al. (2019); Bajrami, et al., (2021); Biron et al. (2021), Jung et al. (2021), and Kim and Pomirleanu (2021). The TS construct was developed using four items, adapted from Salem et al. (2021). The HES construct was developed using five items, adapted from Salem et al. (2021) and Robina-Ramírez et al. (2021). For the AUT, DOW, RES, TRN, and HES constructs, the scores were given on a 5-point Likert scale ranging from 1 (never) to 5 (always). The TSRS construct was developed using three items, taken from Vo-Thanh et al. (2021). The JIS construct was developed using three items, taken from Vo-Thanh et al. (2021) and Jung et al. (2021). The TRET was developed using five items, taken from Mai and Thuy (2021). Appendix 1 is accessible via https://drive.google.com/file/d/192bJJ9sz-Waceb6L5Ek75t91z72IS3ou/view.
For the TSRS, JIS, and TRET constructs, the scores were given on a 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). In the eighth section, demographic and occupation questions were asked. These included questions about gender, age, experience, department, hotel type, category, and the number of rooms. The measures were first developed and revised in English and then transliterated into Arabic by a professional transliterator. This standard back-translation method is to ensure the authenticity and identicalness of intent of the measures, as recommended by Brislin (1986). The results of the back-translation demonstrated that there was high compliance between the Arabic and English versions of the measures. Next, the survey was reviewed by three academics and three HRDs to ensure readability and accuracy. The reviewers noted that the survey was readable. Based on their comments, the authors made some changes to the explanatory descriptions at the start of each section, but no changes were made to the measures.
Data analysis and results
To analyse the collected data, PLS-SEM was used by exploiting WarpPLS7 (Kock, 2020). PLS-SEM involves the estimating of two models: the measurement model and the structural model (Hair et al., 2021). To test multicollinearity and Common Method Bias (CMB), Harman’s single factor test was conducted, whose results revealed that the principal factor was below 50% of the variance. Therefore, this dataset had no problem with CMB (Chin et al., 2012). CMB was further tested in WarpPLS using the Average Full Collinearity Variance Inflation Factor (AFVIF) that affirmed all variables had values (1.43) of less than 3.3, which is ideal (Kock, 2020). The PLS-SEM assessment is comprised of a two-step process; the measurement model through Confirmatory Factor Analysis (CFA), followed by testing the hypothesised structural relationships among the key constructs included in the conceptual model.
Measurement model
Hair et al. (2021) provided guidelines for selecting reflective constructs, which were followed in this study. The Composite Reliability (CR), Cronbach’s Alpha, Average Variance Extracted (AVE), and the Variance Inflation Factor (VIF) were all used to determine convergent validity. Table 2 illustrates the CR and Cronbach Alpha values, both of which were higher than the recommended level of 0.7. The AVE values were greater than 0.5 (range: 0.525–0.865), indicating adequate convergent validity (Fornell and Larcker, 1981). In addition, all variables had VIF values of less than 3.3 (range 1.079–3.014), which is ideal, and there was no multicollinearity or common method bias (Kock and Lynn, 2012).Table 2. Convergent validity.
Variable Composite reliability Cronbach’s alpha AVE VIF
Automation services 0.864 0.790 0.616 1.743
Downsizing 0.847 0.774 0.525 2.090
Restructuring 0.864 0.685 0.761 2.079
Health and safety 0.949 0.933 0.789 2.772
Training support 0.927 0.895 0.762 3.014
Talented satisfaction with response strategies 0.950 0.922 0.865 2.732
Job insecurity 0.888 0.811 0.726 1.646
Talented retention intention 0.945 0.927 0.775 2.851
AVE: average variance extracted; VIF: variance inflation factor.
Second, the study’s major constructs' discriminant validity is evaluated. The square root of AVE for each construct was tested with correlations among the latent variables, as depicted in Table 3, demonstrating adequate discriminant validity (Fornell and Larcker, 1981). Henseler et al. (2015) have presented a new method for verifying discriminant validity that focuses on the multitrait-multimethod matrix to assess discriminant validity: the Heterotrait-Monotrait (HTMT) ratio of correlations (see Table 3). All of the study variables had a discriminant validity of less than 0.85, indicating that they were acceptable.Table 3. Discriminant validity and Heterotrait-Monotrait (HTMT) ratios of correlation.
Construct (discriminant validity) 1 2 3 4 5 6 7 8
1. Automation services 0.785
2. Downsizing 0.495 0.725
3. Restructuring 0.508 0.464 0.872
4. Health and safety 0.399 0.458 0.510 0.888
5. Training support 0.491 0.435 0.450 0.443 0.873
6. Talented satisfaction with response strategies 0.385 0.412 0.432 0.713 0.796 0.930
7. Job insecurity 0.464 0.410 0.337 0.733 0.448 0.726 0.852
8. Talented retention intention 0.510 0.464 0.374 0.435 0.645 0.410 0.475 0.880
Construct (HTMT) 1 2 3 4 5 6 7 8
1. Automation services
2. Downsizing 0.597
3. Restructuring 0.590 0.810
4. Health and safety 0.503 0.769 0.552
5. Training support 0.574 0.788 0.553 0.785
6. Talented satisfaction with response strategies 0.446 0.512 0.521 0.486 0.483
7. Job insecurity 0.596 0.541 0.453 0.470 0.495 0.482
8. Talented retention intention 0.593 0.510 0.532 0.385 0.373 0.4395 0.89
Note: Values on the diagonal (bold) are square root of the average variance extracted.
Note: HTMT ratios are good if < 0.90, best if < 0.85.
Structural model, hypotheses testing, mediation, and moderation analysis
To estimate the model fit, Standardised Root Mean Square Residual (SRMR) was employed (Henseler et al., 2015). A SRMR value of 0 would indicate an ideal fit, and generally, a SRMR value of ≤0.1 is rated as satisfactory for PLS models (Kock, 2020). In this study, an SRMR of 0.086 resulted in a satisfactory model fit. As shown in Table 4, hypothesised relationships were supported, except for H2 (β = 0.074, p-value = 0.08) and H3 (β = −0.043, p-value = 0.20). Furthermore, as depicted in Figure 2, R2 shows the effect of the exogenous constructs on endogenous constructs and tests the predictive accuracy of the model. Values below 0.25 show a weak accuracy, those lower than 0.50 indicate a moderate accuracy, and values below 0.75 imply a solid predictive accuracy. Talented satisfaction with response strategies and retention intention explained 58% and 53% (respectively) of the variance in crisis response strategies.Table 4. Hypothesis-testing summary.
Hypotheses Overall sample (n = 310)
Β p-value Results
H1: AUT TSRS 0.140 0.004** Supported
H2: DOW TSRS 0.074 0.080 Not supported
H3: RES TSRS 0.043 0.206 Not supported
H4: HES TSRS 0.486 <0.001** Supported
H5: TRN TSRS 0.171 <0.001** Supported
H6: TSRS RET 0.581 <0.001** Supported
H7: JINS mod TSRS and RET −0.151* 0.002** Supported
AUT: Automation, DOW: Downsizing, RES: Restructuring, HES: Health and safety, TRN: Training, TSRS: Talented satisfaction with response strategies, JIS:job insecurity, TRET: Talented retention intention.
**Critical p-value for two-tailed tests: p < 0.01.
Figure 2. Structural model.
The mediation effects were employed to examine the significance of the indirect effects. Table 5 exhibits the mediation analysis results. Indirect effects were assessed to uncover the mediating role of talented satisfaction with response strategies in the association between automation, downsizing, restructuring, health and safety, and training and retention intention. The findings show that talented employee satisfaction with response strategies partially mediates the link between automation, health and safety, and training and retention intention, but does not mediate the relationship between downsizing and restructuring and retention intention. Thus, H8.1, H8.4, and H8.5 are accepted, but H8.2 and H8.3 are not supported.Table 5. Mediation analysis.
Paths Path a Path b IEF DE BCI Decision Results
95% LL 95% UL
M1 = H8.1 TSRS med AUT and TRET 0.140 0.581 0.081 0.088 0.009 0.154 Partial mediation Supported
M2 = H8.2 TSRS med DOW and TRET 0.074 0.581 0.043 0.084 −0.030 0.116 No-mediation Not supported
M3 = H8.3 TSRS med RES and TRET 0.043 0.581 0.025 0.027 −0.048 0.098 No-mediation Not supported
M4 = H8.4 TSRS med HES and TRET 0.486 0.581 0.099 0.355 0.027 0.172 Partial mediation Supported
M5 = H8.5 TSRS med TRN and TRET 0.171 0.581 0.282 0.111 0.212 0.353 Partial mediation Supported
M: Mediator, med: mediates, IEF: Indirect effect, DE: Direct effect, BCI: Bootstrapped Confidence Interval, SE: standard error, LL: lower level, UL: upper level.
The moderation effects of using the two-stage approach were also measured to indicate if the relationship between some variables is strengthened or dampened. The formula proposed by Kock (2020) was used to evaluate the variations in path coefficient. To examine the probability of the moderating effect, talent satisfaction with response strategies as a predictor and job insecurity as a moderator were multiplied to generate an interaction construct to predict talent retention intention. The projected standardised path coefficients for the effect of the moderator on predicting talent retention intention (β = −0.151; p = 0.002) were significant (see Table 4). Therefore, H7 has supported that job insecurity dampens the positive relationship between talent satisfaction with response strategies and talent retention intention (see Figure 3).Figure 3. Moderating role of job insecurity.
Discussion
Based on the talent management literature in the hospitality sector and drawing on multiple theoretical perspectives such as conservation of resources theory, organisational support theory, and self-determination theory, we proposed a model that aims to identify the most effective pandemic response strategies related to talents’ attitudes of satisfaction, job insecurity, and retention intention. In this vein, three pandemic response strategies, namely, automation services, health and safety, and training support, were found to positively affect talents' satisfaction with response strategies. In addition, these three essential pandemic response strategies were found to favourably enhance talents' retention intentions through the mediating role of talents' satisfaction. These findings confirm prior research that established the link between organisational support and employee satisfaction and retention intention (Ashton, 2017; Koo et al., 2021). For instance, Moncarz et al. (2009) argue that successful training is necessary to retain individuals with long-term goals.
However, the mediated link of talent satisfaction between pandemic response strategies and talent retention intentions was negatively moderated by job insecurity. This implies that high job insecurity would challenge the positive impact of talent satisfaction on their retention intention. This finding is consistent with past research that has shown that job instability has a significant impact on turnover intentions (Koo et al., 2021). In addition, previous studies reported that job insecurity can have a detrimental effect on staff psychological, mental, and physical health-e.g. anxiety and frustration (Chen and Eyoun, 2021; Jung et al., 2021; Koo et al., 2021). On the other hand, this study’s results do not support our hypotheses that suggest that downsizing and restructuring strategies are significantly related to talented satisfaction and retention intention with response strategies. The insignificant impact of the restructuring response strategy on talented satisfaction and retention intention could be due to the fact that restructuring strategies such as reducing costs or eliminating inefficiency are better than quitting the company, since practically every industry was affected by COVID-19, making it difficult for staff to find a new job (Bajrami et al., 2021). In addition, the non-significant association between downsizing strategy and talents' satisfaction and retention intention could indicate that organisations in their downsizing strategy are not affecting the talented staff due to their outstanding qualities and potential contribution to the competitive advantage of the organisation (Fuentes, 2021; Mai and Thuy, 2021). The following part goes into the theoretical and practical implications.
Theoretical implications
Theoretically, this study demonstrates that automation services, health and safety, and training support as pandemic response strategies have a significant direct influence on talent satisfaction. This implies that top management’s reaction to the crisis, in this case, the pandemic, can positively influence their employees’ satisfaction. In this vein, this study adds to the body of knowledge by identifying the firms' pandemic response plans as major drivers of talent satisfaction. Furthermore, the findings of this study suggest that talent retention is influenced by pandemic response tactics and talent satisfaction. In other words, this research provides a better understanding and emphasis on the applicability of automation services, health and safety, and training support strategies in improving talents' satisfaction and their subsequent positive role in enhancing talented retention intention. This was confirmed by Ashton (2017), who stated that organisational support, such as a small investment in employee training programs, would result in a significant increase in employee satisfaction, which would have a significant impact on their intention to stay. For example, they revealed that a 1% improvement in work satisfaction can raise the intention to stay by more than 50%. In sum, the study provides an enhanced understanding of talent retention intentions by involving pandemic response strategies (i.e. automation services, health and safety, and training support strategies) and talent satisfaction and job insecurity.
Practical implications
The study results provide useful insights for employers in the hospitality industry on the factors that influence talent satisfaction and retention intention during the crisis. In this vein, the results suggest that in order to satisfy and retain talented employees, policymakers and managers in the hospitality industry need to adopt effective response strategies. For example, hotel authorities are urged to take advantage of automated services during a crisis, such as using artificial intelligence, robots, and digital/automated services. In this regard, services such as Hospitality 5.0, which comprises innovative smart technology including automation, robots, mobile phone technology, augmented reality, and virtual reality, are expected to make hotel operations more efficient and organised and reduce costs, which in turn can enhance talented employee satisfaction. Furthermore, based on the positive association between training support and talented satisfaction and retention intention, hospitality companies are encouraged to offer extensive training programmes to minimise the negative impact of the crisis on talented satisfaction and retention intention. For instance, providing training related to crisis management, minimises damage and safeguards the health of staff and consumers. Additionally, hospitality businesses are also urged to provide guidelines, rules, advice, information, processes, and training to keep personnel healthy and disease-free during a crisis, particularly in the event of COVID-19.
Limitation and future research
There are limitations to every research study, and this one is no different. The generalizability of this study’s findings to Egypt’s 4- and 5-star hotel sector setting is limited because the target sample for this study is Egyptian employees of 4- and 5-star hotels. The utilised data collection tool (i.e. questionnaire) in this study might be further implemented to better grasp the talented employee satisfaction with response methods during COVID-19 in various regional contexts as a plan to compensate for this case and concern for future research. The target sample being limited to 4- and 5-star hotels is another drawback of this study. Similarly, only 4- and 5-star hotels in Egypt are included in the study’s conclusions, leaving out all other accommodation and hospitality establishments. Additionally, rather than conducting this study during a crisis, it might be done under more typical conditions. Longitudinal research is essential in this situation. Last but not least, this study was carried out in Egypt, a developing nation. Greater insights regarding keeping skilled workers during the COVID-19 epidemic may come from a multi-group analysis contrasting developed and less developed countries.
Author Biographies
Islam Elbayoumi Salem holds a PhD in Hotel Management from Alexandria University and is an associate professor at the Faculty of Tourism and Hotels, Alexandria University, Egypt, and at the University of Technology and Applied Sciences - Salalah, Oman. His research interests are related to hospitality leadership, hospitality marketing, hospitality technology, hotels' outsourcing, and Human resources. He has published more papers in scholarly journals, such as Tourism Management, International Journal of Hospitality Management, International Journal of Contemporary Hospitality Management, Sustainable Development, Tourism Management Perspectives, International Journal of Tourism Research, and Journal of Hospitality and Tourism Technology.
Hassan Aided holds a PhD in Tourism and Hospitality from Bournemouth University, UK. He is an Assistant Professor at the University of Technology and Applied Sciences, Salalah, Oman. His research interests are event management, tourism marketing, and hospitality management.
Nasser Alhamar Alkathiri holds a PhD in Business administration from Plymouth University, UK. He is an Assistant Professor at the University of Technology and Applied Sciences, Salalah, Oman. His research interests are Knowledge Management, Knowledge transfer, International, Business, Entrepreneurship, Staff localization. He has published multiple papers in reputed journals indexed in WoS and Scopus (e.g., https://www.springer.com/journal/10916/ Journal of Knowledge Management; International Journal of Finance & Economics).
Karam Mansour Ghazi holds a PhD in hospitality crisis management practices from Alexandria University, Egypt. He is an Associate Professor at the High Institute of Tourism and Hotels in Alexandria (EGOTH), Egypt. His research interests are hospitality crisis management, hotel safety and security, hospitality marketing, hospitality trends, and hospitality online reviews. He has a book and many articles published in hospitality crisis management.
ORCID iDs
Islam E Salem https://orcid.org/0000-0002-9483-3550
Karam M Ghazi http://orcid.org/0000-0002-8836-8939
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding: The author(s) received no financial support for the research, authorship, and/or publication of this article.
==== Refs
References
Abuelnasr A (2020) Reactions to Covid-19: the impact of job insecurity on Survivors’ attitudes in five-star hotels. Journal of Association of Arab Universities for Tourism and Hospitality 19 (2 ): 166–188.
Ashton AS (2017) How human resources management best practice influence employee satisfaction and job retention in the Thai hotel industry. Journal of Human Resources in Hospitality & Tourism 17 (2 ): 175–199.
Bajrami DD Terzić A Petrović MD , et al. (2021) Will we have the same employees in hospitality after all? The impact of COVID-19 on employees’ work attitudes and turnover intentions. International Journal of Hospitality Management 94 : 102754.34785836
Brislin RW (1986) The wording and translation of research instrument. In Lonner WJ Berry JW (Eds)Field methods in cross-cultural research (pp. 137-164). Beverly Hills, CA: Sage.
Baum T Mooney SKK Robinson RNS , et al. (2020) COVID-19’s impact on the hospitality workforce? International Journal of Contemporary Hospitality Management 23 (9 ): 2813–2829.
Biron M De Cieri H Fulmer I , et al. (2021) Structuring for innovative responses to human resource challenges: A skunk works approach. Human Resource Management Review 31 (2 ): 100768.
Chen H Eyoun K (2021) Do mindfulness and perceived organizational support work? Fear of COVID-19 on restaurant frontline employees’ job insecurity and emotional exhaustion. International Journal of Hospitality Management 94 : 102850.34785844
Chee LC Ed. (2017) Case Studies in Management and Business. (Volume 3 ) UUM Press.
Chin WW Thatcher JB Wright RT (2012) Assessing common method bias: Problems with the ULMC technique. MIS quarterly 36 : 1003–1019.
Collings DG Mellahi K (2009) Strategic talent management: A review and research agenda. Human Resource Management Review 19 (4 ): 304–313.
Creswell JW (2014) Research Design: Qualitative, Quantitative, and Mixed Methods Approaches. 4th edition. Thousand Oaks, CA: SAGE Publications.
Deery M Jago L (2015). Revisiting talent management, work-life balance and retention strategies. International Journal of Contemporary Hospitality Management, 20 (7 ): 792–806.
Egyptian Hotel Association (2020). Egyptian hotel Guide (online), Retrieved from http:// www.egyptianhotels.org/Statistics.aspx. (Accessed 22 April 2022).
Eisenberger R Huntington R Hutchison S , et al. (1986). Perceived organizational support. Journal of Applied Psychology, 71 (3 ): 500.
Elkhwesky Z Elkhwesky EFY (2022). A systematic and critical review of Internet of Things in contemporary hospitality: a roadmap and avenues for future research. International Journal of Contemporary Hospitality Management, (ahead-of-print).
Elbaz AM Salem IE Onjewu AK Shaaban MN (2022). Hearing employee voice and handling grievance: Views from frontline hotel and travel agency employees. International Journal of Hospitality Management, 107 :103311.
Elkhwesky Z Manzani El Elbayoumi Salem I (2022). Driving hospitality and tourism to foster sustainable innovation: A systematic review of COVID-19-related studies and practical implications in the digital era. Tourism and Hospitality Research, 14673584221126792.
Font X Lynes J (2018) Corporate social responsibility in tourism and hospitality. Journal of Sustainable Tourism 26 (7 ): 1027–1042.
Fornell C Larcker DF (1981) Structural equation models with unobservable variables and measurement error: Algebra and statistics. Journal of Marketing Research 18 (3 ): 382–388.
Fuentes Z (2021). Talent Trends in the Hospitality Industry. Powered by Hosco, July 20, 2021.
Fusté-Forné F Jamal T (2021) Co-creating new directions for service robots in hospitality and tourism. Tourism and Hospitality 2 (1 ): 43–61.
García JAC Pino JMR Elkhwesky Z Salem IE (2022) Identifying core “responsible leadership” practices for SME restaurants. International Journal of Contemporary Hospitality Management, (ahead-of-print).
Glavas A (2016) Corporate social responsibility and employee engagement: Enabling employees to employ more of their whole selves at work. Frontiers in Psychology 7 : 796.27303352
Hair JFJ Hult GTM Ringle CM , et al. (2014) Partial Least Squares Structural Equation Modeling. Los Angeles: SAGE Publications, Inc. DOI: 10.1007/978-3-319-71691-6 10.1007/978-3-319-71691-6.
Hair JF Jr Hult GTM Ringle CM , et al. (2021) A Primer on Partial Least Squares Structural Equation Modeling (PLS-SEM). LA: Sage publications.
Henseler J Ringle CM Sarstedt M (2015) A new criterion for assessing discriminant validity in variance-based structural equation modeling. Journal of the academy of marketing science 43 (1 ): 115–135.
Hobfoll SE (1989). Conservation of resources: a new attempt at conceptualizing stress. American Psychologist, 44 (3 ), 513.2648906
Ivanov S Webster C Stoilova E , et al. (2020) Biosecurity, automation technologies and economic resilience of travel, tourism and hospitality companies. Sorarxiv Pap: 1–34.
Jooss S McDonnell A Burbach R , et al. (2019) Conceptualising talent in multinational hotel corporations. International Journal of Contemporary Hospitality Management 31 (10 ): 3879–3898.
Jooss S Lenz J Burbach R (2022) Beyond competing for talent: an integrative framework for coopetition in talent management in SMEs. International Journal of Contemporary Hospitality Management: 1–17. in press.
Kravariti F Voutsina K Tasoulis K , et al. (2022) Talent management in hospitality and tourism: a systematic literature review and research agenda. International Journal of Contemporary Hospitality Management 34 (1 ): 321–360.
Jung HS Jung YS Yoon HH (2021) COVID-19: the effects of job insecurity on the job engagement and turnover intent of deluxe hotel employees and the moderating role of generational characteristics. International Journal of Hospitality Management 92 : 102703 .33041428
Kaewsaeng-On R (2016) Talent Management: A Critical Investigation in the Thai Hospitality Industry. Doctoral dissertation, University of Salford. Salford.
Kichuk A (2017) Understanding Talent Management in the Hotel Sector: Employees’ Narratives of Personal Career Development. Doctoral dissertation. Bournemouth, England: Bournemouth University.
Kim EJ Pomirleanu N (2021) Effective redesign strategies for tourism management in a crisis context: A theory-in-use approach. Tourism Management 87 : 104359.
Kock N (2020) WarpPLS User Manual: Version 7.0. Laredo, TX, USA: ScriptWarp Systems.
Kock N Lynn G (2012) Lateral collinearity and misleading results in variance-based SEM: An illustration and recommendations. Journal of the Association for Information Systems 13 (7 ): ■■■.
Koo B Curtis C Ryan B (2021) Examining the impact of artificial intelligence on hotel employees through job insecurity perspectives. International Journal of Hospitality Management 95 : 102763.
Lukanova G Ilieva G (2019) Robots, Artificial Intelligence and Service Automation in Hotels. In: Ivanov S Webster C (eds) Robots, Artificial Intelligence, and Service Automation in Travel, Tourism, and Hospitality. Bingley: Emerald Publishing Limited, pp. 157–183.
Mai NN Thuy VHN (2021) Talent Retention at Commercial Banks in Vietnam. International Review of Management and Marketing 11 (1 ): 48.
McDevitt R Giapponi C Houston DM (2013) Organizational downsizing during an economic crisis: Survivors' and victims' perspectives. Organization Management Journal 10 (4 ): 227–239.
Moncarz E Zhao J Kay C (2009) An exploratory study of US lodging properties' organizational practices on employee turnover and retention. International Journal of Contemporary Hospitality Management.
Maslow AH (1943) Preface to motivation theory. Psychosomatic Medicine 5 : 85–92.
Patton MQ (2015) Qualitative Research and Evaluation Methods. 4th edition. Thousand Oaks, CA: SAGE Publications.
Pillai SG Haldorai K Seo WS , et al. (2021) COVID-19 and hospitality 5.0: Redefining hospitality operations. International Journal of Hospitality Management 94 : 102869.34785847
Ritchie BW Jiang Y (2019) A review of research on tourism risk, crisis and disaster management. Annals of Tourism Research 79 : 102812 .
Robina-Ramírez R Medina-Merodio JA Moreno-Luna L , et al. (2021). Safety and health measures for COVID-19 transition period in the hotel industry in Spain. International Journal of Environmental Research and Public Health, 18 (2 ), 718.33467614
Ryan RM Deci EL (2000) Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist 55 (1 ): 68.11392867
Salem IE Elkhwesky Z Ramkissoon H (2022) A content analysis for government’s and hotels’ response to COVID-19 pandemic in Egypt. Tourism and Hospitality Research 22 (1 ): 42–59.
Salem IE Elbaz AM Elkhwesky Z , et al. (2021) The COVID-19 pandemic: the mitigating role of government and hotel support of hotel employees in Egypt. Tourism Management 85 : 104305.
Salem IE Elbaz AM Al-Alawy A Alkathiri NA Elkhwesky Z (2022) Is eco-label hotel engagement the pathway to sustainability practices via entrepreneurial resilience and orientation in Oman? Findings from PLS-SEM and fsQCA. International Journal of Contemporary Hospitality Management. (ahead-of-print).
Santana M Valle R Galan JL (2017) Turnaround strategies for companies in crisis: watch out the causes of decline before firing people. BRQ Business Research Quarterly 20 (3 ): 206–211 .
Scott ME (2016) Strategies for Retaining Employees in the Hospitality Industry. Doctoral dissertation. Minneapolis, Minnesota, USA: Walden University.
Sheehan M Grant K Garavan T (2018) Strategic talent management: A macro and micro analysis of current issues in hospitality and tourism. Worldwide Hospitality and Tourism Themes 10 : 28–41.
Silverman D (2017) Doing Qualitative Research. 5th edition. Thousand Oaks, CA: SAGE Publications.
Tansley C Foster C Harris L , et al. (2007) Talent: Strategy, Management and Measurement. London, England: Chartered Institute of Personnel and Development (CIPD).
Thumiki VRR Jovancai-Stakić A Al BRSS (2019) Resultant effect of crisis-driven HR strategies applied during current economic crisis in Oman: an HR manager's perspective. The European Journal of Applied Economics 16 (1 ): 77–98. .
Vasquez D (2014) Employee retention for economic stabilization: A qualitative phenomenological study in the hospitality sector. International Journal of Management, Economics and Social Sciences (IJMESS) 3 (1 ): 1–17 .
Vo-Thanh T Vu TV Nguyen NP , et al. (2021) How does hotel employees’ satisfaction with the organization’s COVID-19 responses affect job insecurity and job performance? Journal of Sustainable Tourism 29 (6 ): 907–925 .
Wong AKF Kim SS Kim J , et al. (2021) How the COVID-19 pandemic affected hotel Employee stress: Employee perceptions of occupational stressors and their consequences. International Journal of Hospitality Management 93 : 102798.
WTTC (2015) Global Talent Trends and Issues for the Travel and Tourism Sector. A report prepared by Oxford Economics for the World Travel & Tourism CouncilRetrieved from http://www.wttc.org/-/media/382bb1e90c374262bc951226a6618201.ashx.
Zainee IA Puteh F (2020) Corporate social responsibility impact on talent retention among Generation Y. Revista de Gestão 27 (4 ): 369–392.
Zeng Z Chen PJ Lew AA (2020) From high-touch to high-tech: COVID-19 drives robotics adoption. Tourism geographies 22 (3 ): 724–734.
Zhang J Ahammad MF Tarba S , et al. (2014) The effect of leadership style on talent retention during merger and acquisition integration: Evidence from China. The International Journal of Human Resource Management 26 (7 ): 1021–1050.
| 0 | PMC9703016 | NO-CC CODE | 2022-11-29 23:21:05 | no | 2022 Nov 24;:14673584221141294 | utf-8 | null | null | null | oa_other |
==== Front
Am Behav Sci
Am Behav Sci
ABS
spabs
The American Behavioral Scientist
0002-7642
1552-3381
SAGE Publications Sage CA: Los Angeles, CA
10.1177/00027642221138274
10.1177_00027642221138274
Article
Using Health Behavior Theory to Address COVID-19 Vaccine Hesitancy: A Scoping Review of Communication and Messaging Interventions
Orr Caroline A. 1
Gordon Ruthanna 1
1 Applied Research Laboratory for Intelligence and Security (ARLIS), University of Maryland, College Park, MD, USA
Caroline A. Orr, Applied Research Laboratory for Intelligence and Security (ARLIS), University of Maryland, 7005 52nd Ave, College Park, MD 20742, USA. Email: [email protected]
24 11 2022
24 11 2022
00027642221138274© 2022 SAGE Publications
2022
SAGE Publications
This article is made available via the PMC Open Access Subset for unrestricted re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the COVID-19 pandemic or until permissions are revoked in writing. Upon expiration of these permissions, PMC is granted a perpetual license to make this article available via PMC and Europe PMC, consistent with existing copyright protections.
Vaccine hesitancy has been among the most vexing challenges during the COVID-19 pandemic, ultimately leading to maladaptive health behaviors such as vaccine delay and refusal. A variety of approaches have been employed to address this problem, including communication and messaging interventions targeting the underlying determinants of vaccine hesitancy. However, there exists no published evidence synthesis examining how such interventions are using health behavior theory to address COVID-19 vaccine hesitancy. The purpose of this study was to conduct a scoping review of health communication and messaging interventions aimed at addressing COVID-19 vaccine hesitancy, and to systematically evaluate the use of health behavior theory in the design of these interventions. The review followed a five-step iterative framework proposed by Levac and colleagues. Comprehensive searches using an exhaustive list of keyword combinations were used across three online databases to identify articles to screen for inclusion. A structured, validated coding scheme was then applied to assess the use of health behavior theory. Additional study data were extracted using a separate structured form. A total of 36 articles published between January 2020 and February 2022 met inclusion criteria and were included in the review. Ten studies (27.7%) did not mention or use health behavior theory at all. Most studies (n = 26) at least mentioned theory or theory-relevant constructs, with 26 different theories and 52 different theoretical constructs represented in the sample. Although theory and theoretical determinants of vaccination behavior were often mentioned, few studies used theory to specify and target causal pathways of behavior change, and only one study targeted misinformation as a determinant of vaccine hesitancy. The findings from this review provide critical insight into the state of theory-based intervention design and point to significant gaps in the literature to prioritize in future research.
COVID
health behavior theory
vaccine hesitancy
edited-statecorrected-proof
typesetterts1
==== Body
pmcIntroduction
Vaccination is considered to be one of the top 10 public health achievements of the 20th century (Centers for Disease Control and Prevention, 1999), and widespread uptake of the COVID-19 vaccine has been identified as a key step toward ending the coronavirus pandemic. However, large gaps in COVID-19 vaccination coverage persist in the United States and globally (New York Times, 2022). While the reasons for low vaccination uptake are varied, socio-psychological factors such as vaccine hesitancy are key contributors (Aw et al., 2021; Bogart et al., 2021; Karlsson et al., 2021; Xiao & Wong, 2020).
Health communication interventions may be ideally suited to address the challenge of vaccine hesitancy, but there remain many unanswered questions about the most effective communication strategies, delivery formats, messengers, timing, and more (Dubé et al., 2015; Jarrett et al., 2015). Furthermore, a review of reviews of interventions in this area found “no strong evidence to recommend any specific intervention to address vaccine hesitancy/refusal” (Dubé et al., 2015), and a systematic review of such approaches found “limited evidence” on how to address the problem (Jarrett et al., 2015). During the coronavirus pandemic, these challenges have been further exacerbated by the existence of an infodemic, defined as “an overabundance of information – some accurate and some not—that occurs during an epidemic” (World Health Organization [WHO], 2020a; WHO, 2020b), resulting in an erosion of trust and persistent fear, anxiety, and vaccine hesitancy (Bullock et al., 2022; Loomba et al., 2021).
A promising way to address these challenges and advance cumulative knowledge is through the use of social and behavioral science theory in intervention design and research, which can guide the selection of intervention techniques, assist in identifying key determinants of behavior, and inform decisions about message tailoring, participant selection, and measurement, as well as provide a framework for synthesizing evidence. To date, however, there is a lack of research on theory-based communication interventions in this area (Dubé et al., 2015; Kenzig & Mumford, 2022). In the current study, we sought to conduct the first scoping review focused on systematically evaluating the use of theory in the design of health communication interventions promoting COVID-19 vaccination, and identifying promising directions for applying theory to create more effective future interventions.
Determinants of Vaccine Uptake and Refusal
Many social and behavioral science theories have been used to explain and predict vaccination behaviors. Some of the most widely used theories include the Theory of Planned Behavior (TPB)/Theory of Reasoned Action (Ajzen, 1999), the Health Belief Model (HBM) (Rosenstock, 1974), the Extended Parallel Process Model (Witte, 1992), Protection Motivation Theory (PMT) (Rogers, 1975), and the “3 Cs” model of vaccine hesitancy (WHO, 2014). Drawing on these theories, a variety of social and psychological constructs have been identified as key determinants of vaccine acceptance and uptake, including knowledge, attitudes, health beliefs, self-efficacy, and perceived behavioral control (Chu & Liu, 2021; Schmid et al., 2017; Xiao & Wong, 2020). In the context of COVID-19, specifically, lower levels of perceived risk and perceived severity related to the virus, as well as safety concerns about the vaccine, have been found to be associated with increased vaccine hesitancy, as have beliefs that COVID-19 is not severe or is not a real disease, or that the threat of COVID-19 has been exaggerated (Aw et al., 2021). Similarly, determinants of COVID-19 vaccine acceptance include perceived severity of COVID-19, perceived benefits of the vaccine, and cues to action, all of which have been shown to be positive correlates of vaccine acceptance, while perceived access barriers and harm were found to be negative correlates (Wong et al., 2021).
Use of Theory in Intervention Design
In the field of social and behavioral sciences, the term “theory” generally refers to “a set of interrelated concepts, definitions, and propositions that present a systematic view of events or situations by specifying relations among variables, in order to explain and predict the events or situations” (Glanz et al., 2008, p. 26). Evidence suggests that theory-based interventions, or those that target theoretical mechanisms of behavior change, are more likely to be effective than non-theory-based interventions (Michie et al., 2008; Noar & Zimmerman, 2005). This has been found across a variety of health conditions and behaviors, including physical activity (Taylor et al., 2012), nutrition habits (Lara et al., 2014), and cancer (Albada et al., 2009), as well as Internet-based health behavior interventions (Webb et al., 2010). However, a significant proportion of published interventions still make no reference to a theoretical basis (Albarracín et al., 2005; Davies et al., 2010; Hardeman, et al., 2002; Painter et al., 2008). Furthermore, even when theory is applied, it is rarely used to its full potential, and significant discrepancies exist between reported theory-use and actual application of theory (Dombrowski et al., 2007; Gardner et al., 2010; Orr, 2020; Painter et al., 2008). For example, in a meta-analysis of prenatal behavior change interventions, Orr (2020) found that theory-use was almost exclusively limited to descriptive purposes (as opposed to explanatory or predictive). Similarly, Dombrowski et al. (2007) found that although 44% of the trials in their review reported a theoretical basis for intervention development, none of the studies actually explained how theory was used to develop the intervention. Furthermore, studies ostensibly drawing on the same theory may leverage different constructs and concepts from the theory, or interpret them in different and sometimes contradictory ways.
To our knowledge, there exist no systematic evaluations of theory-use in the design of health communication interventions aimed at promoting the COVID-19 vaccine. Thus, the objective of this study was to conduct a scoping review of the literature in this area, with a focus on assessing how social and behavioral science theories are being used to inform intervention design. We were particularly interested in how health communication interventions are addressing misinformation as an important determinant of vaccination behavior, given its key role as a driver of vaccine hesitancy during the COVID-19 pandemic.
Methods
The review followed a five-step iterative framework proposed by Levac et al. (2010): (1) identification of the research questions; (2) literature search and identification of relevant articles; (3) selection of studies for inclusion; (4) charting the data (including application of Theory Coding Scheme [TCS]); and (5) collating, summarizing, and reporting the results.
The objectives of the review included the following:
1. To systematically evaluate the use of theory in the design and evaluation of health communication interventions aimed at promoting COVID-19 vaccination intentions and/or behaviors. Specific research questions included:
a. How is theory being used? Is it being used optimally?
b. What are the most frequently used theories and theoretical constructs/predictors?
c. How is theory being used to address vaccine-related misinformation and its effects, such as vaccine hesitancy?
2. To identify promising theoretical models and frameworks, as well as discrepancies and/or inconsistent findings, limitations, gaps in the literature, and questions for future research.
a. Is use of theory associated with intervention success?
b. What are the most consistent findings and/or promising uses of theory?
c. Are descriptions of theories and theoretical constructs/predictors consistent? Is there consensus on conceptual definitions and measurement?
d. What gaps in application of theory suggest potential new research directions?
Search Strategy
A literature search was conducted to identify published studies evaluating health communication interventions aimed at increasing COVID-19 vaccination intentions and/or behaviors. The pandemic provided a fixed timeframe for our literature search, which was limited to studies published since the start of the global pandemic in January 2020 until February 2022. To identify articles, an exhaustive list of keyword pairings (available in Appendix 2) were used across three online databases (PubMed, Google Scholar, and ScienceDirect). MESH terms were utilized for the searches.
Study Eligibility
Screening Process: To screen articles for eligibility, we used three levels of review: Level one screening reviewed the titles and dates of all search results. Level two screening reviewed the abstracts of all articles that the search terms returned. Level three screening reviewed the full text of articles to determine their eligibility. After selecting eligible articles, we reviewed the full articles’ reference lists and identified additional articles for possible inclusion.
Inclusion/Exclusion Criteria: To be included in the review, articles must have reported the results of experimental research focused on using health communication strategies to increase COVID-19 vaccine intentions and/or behavior. We included both intentions and behaviors as outcomes because the COVID-19 vaccine is still relatively new, and much of the published research was carried out before the first vaccines were administered. Articles must have been published in English between January 2020 and February 2022. We did not place any geographic limits on the location where the research was conducted or published. Peer-reviewed full-text articles, dissertations, masters theses, scientific reports, and working papers were all considered for inclusion.
Data Extraction and Synthesis
A structured data extraction form was used to manually collate data from the included studies. The extracted data included: article type, publication date, study aims, study design and methods, location of study, sample size and participant characteristics, primary and secondary outcomes, mediators/moderators, and key findings. Additional data were extracted during the process of coding for use of theory (described below). Characteristics of studies were summarized individually in tables and described collectively in narrative form. Results from the TCS were analyzed in the same way.
Data Analysis
The TCS (Michie & Prestwich, 2010) was used to code for reported theory-use in the design of message-testing experiments and related communication-focused interventions. Items on the TCS are coded categorically (Yes/No/Don’t Know) and demonstrated substantial agreement during initial development and validation (kappa > 0.70 for 18/19 items; kappa = 0.64 for item 19d) (Michie & Prestwich, 2010). Full TCS scoring instructions, as well as information on how it was adapted for use in this review, are included in Appendix 1.
Coding
To establish interrater reliability, two trained coders independently applied the TCS to a set of articles that were not part of the final sample for this review, but which described health communication/messaging interventions targeting flu and/or HPV vaccination intentions. Cohen’s kappa (k) coefficient was used to assess agreement between coders (Cohen, 1960). Once interrater reliability was achieved on this set of non-included articles (k = 0.66), the two coders then independently coded approximately 10% of the articles included in the review to ensure interrater reliability within this sample (k = 0.72). Disagreements between coders were resolved through discussion and further examination of the studies and item content. The remaining articles were each coded by a single coder.
Scoring
Items on the TCS can be treated individually as well as grouped together to form composite measures reflecting the extent and function of their use. In this analysis, items were analyzed individually and, in some cases, composite measures were created to reflect specific uses of theory. Six composite measures were created based on the scoring criteria developed by Prestwich et al. (2014). The measures reflect the following:
1) “Mention of Theory” was calculated on a scale of 0–4 by summing the scores of four items assessing whether theory and/or theoretical predictors were explicitly mentioned.
2) “Intervention Techniques Linked to Theoretical Constructs” was calculated on a scale of 0–3 by summing the scores of five items assessing whether intervention techniques were linked to relevant theoretical constructs and/or predictors.
3) “Tailored Intervention Techniques” was calculated on a scale of 0–2 by summing the scores of two items assessing whether theory was used to select participants and/or tailor intervention techniques to specific participant characteristics.
4) “Theoretical Constructs Measured” was calculated as the score of a single 1-point item assessing whether the targeted theoretical constructs were measured pre-/post-intervention or post-intervention.
5) “Tested or Refined Theory” was calculated on a scale of 0–4 by summing the scores of four items assessing the extent and nature of theory testing and refinement based on the results.
6) “Overall Use of Theory” was calculated on a scale of 0–14 by summing the totals of the composite measures.
Results
Search Results
A total of 8,293 articles were initially identified across the three database searches. After removing duplicates, 3,191 titles and abstracts were screened, with 205 full-text articles proceeding to further review. From those 205 articles, a total of 31 were retained for inclusion in the review. Articles were excluded if they did not report an outcome of COVID-19 vaccine intentions or behavior, if they were not experimental in design, if they did not report the results of a health communication intervention, if they were published in a language other than English, or if they were published prior to 2020 or after February 2022. After examining the reference lists of these 31 articles, an additional five articles were determined to be eligible for the review. A total of 36 studies were retained for inclusion in the review based on the criteria described above, including 32 peer-reviewed journal articles, two theses or dissertations, and two working papers or reports. Half of these studies (n = 18) were conducted in the United States and/or online with an American audience. The remainder were conducted in China (n = 5), the UK (n = 4), Nigeria (n = 1), Japan (n = 1), Germany (n = 1), Italy (n = 1), France (n = 1), Switzerland and Sweden (n = 1), Saudi Arabia (n = 1), and Latin America (Argentina, Brazil, Chile, Colombia, México, and Peru) (n = 1). Of these studies, 34 measured COVID-19 vaccine intentions as the primary outcome, one measured intentions and behavior (vaccination), and one measured behavior only (scheduling a vaccine appointment). Nearly all (n = 34) were delivered in an online setting.
RQ1: How is Theory Being Used? Is it Being Used Optimally?
Total scores on the 14-point TCS ranged from 0 to 10 (M = 3.81, SD = 3.30), as seen in Figure 1. Ten articles were given a score of zero, while just one achieved a score of 10. Scores on the first subscale (“Was theory mentioned?”) ranged from 0 to 4 (M = 1.72, SD = 1.39). Only three studies (8.3%) had a score of 4.0, while 10 articles (27.7%) scored a zero on this subscale, indicating that the studies did not explicitly mention theory or theoretical determinants of vaccination behavior. Scores on the second subscale (“Were relevant theoretical constructs targeted?”) ranged from 0 to 3 (M = 1.39, SD = 1.32), with 15 studies (69.4%) scoring a zero and 8 studies (22.2%) achieving a score of 3.0. None of the studies in the review used any theories in their entirety (by targeting all of the constructs within a specified theory). On the third subscale (“Was theory used to select participants or tailor interventions?”), all studies in the review were given a score of zero, indicating that they did not use theory or theoretical predictors to select participants or to tailor the treatment/experimental content. On the fourth subscale (“Were theory-relevant constructs/predictors measured?”), 15 studies (69.4%) scored a 1.0, indicating that they measured at least one theory-relevant construct or predictor post-intervention or pre- and post-intervention, while 23 studies (63.9%) scored a zero. Scores on the fifth subscale (“Is theory tested or refined?”) ranged from 0 to 3 (M = 0.63, SD = 0.91). More than half of the articles in the review (n = 21; 58.3%) scored a zero on this 4-point subscale, indicating that the studies did not use evidence to support or refute theorized pathways of change and/or relationships between theoretical constructs, and did not conduct mediational analyses of constructs or predictors. Related, explanations for intervention outcomes were usually not situated in the context of theory. Only one study (Borah et al., 2021) made an effort to refine theory by discussing the results in terms of the theoretical basis of the intervention.
Figure 1. Mean scores on TCS subscales and total scale.
Note. TCS = Theory Coding Scheme.
In all, 25 studies (69.4%) mentioned a theory or model of behavior in the introduction or methods section, even if the intervention was not guided by it. Most of these studies (n = 21; 58.3%) targeted at least one theory-relevant construct or predictor, and provided evidence from the literature of its relationship with behavior. Only four of these studies (11.1%) were guided by a single theory, while the remainder drew on multiple theories or a combination of theoretical constructs and predictors from multiple theories and/or that were both linked to theory and not linked to theory (Borah, 2022; Borah et al., 2021; Chen et al., 2022; Gong et al., 2021). Nine of the studies in the review were at least partially “theory based.” Theory-based interventions are distinguished from “theory-guided,” “theory-influenced,” or “theory-inspired” interventions by their specification of an explicit causal pathway(s) underlying behavior change (Michie & Abraham, 2004; Michie et al., 2008).
RQ2: What Are the Most Promising Theories and Frameworks? Is Use of Theory Associated with Intervention Success?
Previous evidence suggests that behavior change interventions based on a single theory may be more effective than those based on multiple theories or a combination of theory and predictors (Prestwich et al., 2014), but we did not find that to be the case in our review. Of the four studies guided by a single theory—all of which tested the effects of gain versus loss framing—only one study reported that the message condition had a significant effect on intentions to vaccinate (Gong et al., 2021). Conversely, of the 17 studies that drew on multiple theories or combinations of theories and predictors, nine reported significant results for intentions. More broadly, the relationship between theory-use and intervention success was unclear, in large part due to substantial variation in methodology and reporting practices, which made it difficult to compare across all studies and outcomes.
RQ3: What Are the Most Frequently Used Theories and Theoretical Constructs/Predictors?
A total of 26 different theories were represented in the sample, as seen in Table 1. Only six of these theories were mentioned in more than one article: Prospect Theory (n = 9), TPB (n = 4), HBM (n = 3), Framing Theory/Framing Effects Theory (n = 3), Behavioral Economics (n = 2), and PMT (n = 2). As seen in Table 2, the most common theoretical constructs used to guide intervention design were self-efficacy/efficacy beliefs (n = 5), attitudes (n = 5), and risk perceptions (n = 5). These theories have in common a focus on psychological phenomena relevant to health attitudes and behavior, including decision-making under uncertainty, risk assessment, and assessment of likely outcomes of potential choices.
Table 1. Specific Uses of Theory in Interventions Targeting Vaccine Hesitancy.
Uses of theory in intervention design Yes, n (%) No, n (%)
Did the study mention theory or theoretical determinants of vaccination behavior? 25 (69.5%) 11 (30.5%)
Did the intervention target at least one theoretical determinant of vaccination behavior? 21 (58.3%) 15 (41.6%)
Did the study use theory to select participants or tailor intervention techniques to participants? 0 (0%) 36 (100%)
Did the study measure at least one theoretical determinant of vaccination behavior? 15 (41.6%) 21 (58.3%)
Did the study test or refine theory? 1 (2.8%) 35 (97.2%)
Table 2. Social and Behavioral Science Theories Used as Conceptual Frameworks and/or to Guide Intervention Design.
Theories Citations n
Prospect theory Borah (2022)
Borah et al. (2021)
Baumgartner (2020)
Chen et al. (2022)
Hong and Hashimoto (2021)
Huang and Liu (2021)
Motta et al. (2021)
Reinhardt and Rossmann (2021)
Sasaki et al. (2022) 9
TPB Baumgartner (2020)
Capasso et al. (2021)
Reinhardt and Rossmann (2021)
Ugwuoke et al. (2021) 4
Framing theory/framing effect theory Abdel-Raheem and Alkhammash (2021)
Gong et al. (2021)
Gursoy et al. (2022)
Ye et al. (2021) 4
HBM Giampaolo (2021)
Kerr et al. (2021)
Ye et al. (2021) 3
PMT Gursoy et al. (2022)
Yang (2022) 2
Behavioral economics Sasaki et al. (2022)
Strickland (2022) 2
Exemplification theory Ye et al. (2021) 1
Elaboration likelihood model Ye et al. (2021) 1
Cultural cognition theory Yuan and Chu (2022) 1
Social cognitive theory Ugwuoke et al. (2021) 1
Social identity theory Sinclair (2021) 1
Competence hypothesis for dealing with ambiguity Simonovic and Taber (2022) 1
Socioemotional selectivity theory Reinhardt and Rossmann (2021) 1
Terror management theory Motta et al. (2021) 1
Collective action theory James et al. (2021) 1
Theories of cooperation and prosocial behavior James et al. (2021) 1
Prevention-detection framework Hong and Hashimoto (2021) 1
Micro–macro framework Giampaolo (2021) 1
3C’s model of vaccine hesitancy Giampaolo (2021) 1
5C’s model Giampaolo (2021) 1
Trust and confidence model Giampaolo (2021) 1
Construal-level theory Huang and Liu (2021) 1
Secondary risk theory Gursoy et al. (2022) 1
Uncertainty management theory Huang and Liu (2021) 1
Extended parallel processing model Yang (2022) 1
Affect as information theory Yang (2022) 1
Table 3. Constructs, Predictors, and Mechanisms of Action From Social and Behavioral Science Theories Used by Studies in the Review.
Theoretical constructs Study authors n
Perceived efficacy Baumgartner (2020)
Ye et al. (2021)
Palm (2021) 5
Perceived effectiveness Peng et al. (2021)
Efficacy beliefs Kerr et al. (2021)
Risk perceptions (severity and susceptibility) Gursoy et al. (2022)
Simonovic and Taber (2022)
Yang (2022)
Ye et al. (2021) 5
Perceived susceptibility (but labeled “perceived risk”) Hong and Hashimoto (2021)
Attitudes (about COVID vaccination) Altay et al. (2021)
Borah et al. (2021)
Chen et al. (2022)
Hong and Hashimoto (2021)
Reinhardt and Rossmann (2021) 5
Emotional induction
Emotion invocation (embarrassment, guilt, anger) Yang (2022)
James et al. (2021) 2
Perceived benefits Borah et al. (2021)
Ye et al. (2021) 2
Social norms Palm (2021)
Sinclair (2021) 2
Self-efficacy Simonovic and Taber (2022)
Ugwuoke et al. (2021) 2
Issue involvement Baumgartner (2020)
Reinhardt and Rossmann (2021) 2
Vaccine hesitancy Freeman et al. (2021)
Giampaolo (2021) 2
Perceived safety of COVID vaccine Baumgartner (2020)
Palm (2021) 2
Perceived costs (barriers) Ye et al. (2021) 1
Cultural cognition constructs (hierarchy-egalitarianism and individualism-communitarianism) Yuan and Chu (2022) 1
Individualism-collectivism Borah (2022) 1
Task efficacy Ugwuoke et al. (2021) 1
Outcome expectancies Ugwuoke et al. (2021) 1
Outcome certainty/uncertainty Chen et al. (2022) 1
Anticipated positive affective reactions to COVID-19 vaccination Capasso et al. (2021) 1
Anticipated negative affective reactions to COVID-19 vaccination Capasso et al. (2021) 1
Health cognitions Simonovic and Taber (2022) 1
Ambiguity tolerance Simonovic and Taber (2022) 1
Dispositional optimism Simonovic and Taber (2022) 1
Cognitive attitudes toward vaccinating against COVID-19 Capasso et al. (2021) 1
Attitudes (about vaccines in general) Reinhardt and Rossmann (2021) 1
Psychological reactance Reinhardt and Rossmann (2021) 1
Recognition accuracy Reinhardt and Rossmann (2021) 1
Individual awareness (“people’s understanding of COVID-19 and their awareness of observing the government’s epidemic prevention and control measures”) Peng et al. (2021) 1
Social relationship factors (norms, support; % of family and friends who are vaccinated) Peng et al. (2021) 1
Worry (about vaccine side effects) Li et al. (2022) 1
Concern over side effects (vaccine beliefs, perceived risk) Kerr et al. (2021) 1
Concern over safety/regulatory timeline (vaccine beliefs, perceived risk) Kerr et al. (2021) 1
Decisional conflict Kerr et al. (2021) 1
Vaccine hesitancy (COVID vaccine attitudes) Kerr et al. (2021) 1
Personal health risk (of not vaccinating) Motta et al. (2021) 1
Community health risk (from not vaccinating) Motta et al. (2021) 1
Self-interest James et al. (2021) 1
Altruism Gong et al. (2021) 1
Personal benefits of getting vaccinated Freeman et al. (2021) 1
Community interest James et al. (2021) 1
Collective benefits of vaccination Freeman et al. (2021) 1
Cognitive elaboration Hong and Hashimoto (2021) 1
Numeracy skills Chen et al. (2022) 1
Psychological ownership Dai et al. (2021) 1
COVID-19 vaccine confidence and complacency beliefs (measures collective importance, efficacy, side effects, and speed of development) Freeman et al. (2021) 1
Institutional and medical trust Giampaolo (2021) 1
Psychological uncertainty Huang and Liu (2021) 1
Behavioral beliefs Huang and Liu (2021) 1
Perceived threat (to freedom) Huang and Liu (2021) 1
Personal freedom James et al. (2021) 1
Among the articles we examined, theory was used primarily to guide the design of COVID-19 messaging strategies, either by informing the framing of the message(s) such as in the case of gain/loss framing or by guiding the selection of the content or target of the message such as in the case of promoting pro-vaccine social norms or reducing uncertainty. In most instances, however, the theoretical construct(s) targeted by intervention content were not measured separately from outcomes, so it was not possible to determine whether the intervention led to changes in those constructs, nor whether those constructs were associated with vaccine intentions or behavior. For example, Sinclair and Agerström (2021) designed messages to reduce vaccine hesitancy and strengthen COVID-19 vaccine intentions by communicating strong (vs. weak) descriptive norms (information about what most other people are doing) and manipulating the reference group. These messages did not result in increased vaccine intentions compared to standard information, and manipulating the reference group did not produce any significant changes. However, since the study did not measure perceived norms, it is not possible to say whether the lack of significant results reflects a lack of relationship between perceived norms and COVID-19 vaccine intentions, or whether these messages in particular did not influence perceived norms. Similarly, Argote et al. (2021) provided information about vaccine efficacy and social approval but did not measure corresponding constructs such as outcome expectancies, perceived benefits, efficacy beliefs, or perceived norms.
RQ4: What Are the Most Consistent Findings and/or Promising Uses of Theory?
The most consistent finding reported across studies was that loss framing (sometimes described as “negative” or “consequences” framing) outperformed gain framing (sometimes described as “positive” or “benefits”), and that the effects of message framing were often moderated by health beliefs. In total, 12 studies performed some type of gain versus loss message testing. Nine of these studies presented evidence that loss framing outperformed gain framing in terms of its impact on intentions to vaccinate. However, the effects of loss framing varied according to factors such as age and health beliefs. Regarding age, Reinhardt and Rossmann (2021) reported that loss framing led to stronger intentions to vaccinate, but only among younger participants. Regarding health beliefs, Ye et al. (2021) found that the effects of loss framing were moderated by perceived severity of COVID-19, perceived benefits of vaccination, and perceived costs of not vaccinating. In addition, Hong et al. (2021) reported that loss framing led to stronger vaccination intentions, but only for those with low perceived risk of COVID-19, while Gursoy et al. (2022) found that perceived vaccination risk fully mediated the effect of gain/loss framing on changes in intentions. In addition, Borah et al. (2021) found that the effects of loss framing were only present for those with higher perceived benefits, while Ye et al. (2021) found that perceived benefits mediated the effects of narrative framing on vaccination intentions. Beyond the implications for message framing, these results also provide support for the explanatory power of HBM constructs, which functioned as mediators in six studies. Also of note, Abdel-Raheem and Alkhammash (2021) reported that framing effects may vary depending on the type of media in which the vaccine-related message is embedded. Specifically, participants were susceptible to framing effects when they read news articles containing vaccine-related messaging, but not when the messaging was embedded within a cartoon.
RQ5: How is Theory Being Used to Address Vaccine-Related Misinformation and its Effects, such as Vaccine Hesitancy?
Very few studies made explicit reference to misinformation or offered strategies to address its effects. An exception was Motta et al.’s (2021) study, which examined the impact of including “prebunking” information as part of a pro-vaccine message. However, the inclusion of such information did not have a significant impact on vaccine intentions. In Chang et al.’s (2021) study of behavioral nudges, one of the message frames was designed to counter misinformation related to vaccination among children, though belief in/exposure to misinformation was not measured. Finally, Huang and Liu (2021) referenced misinformation as a rationale for their study design and objectives, stating that their aim was to help health agencies communicate more effectively on social media by breaking through the noise of misinformation and uncertainty. However, the study did not include any messaging or techniques specifically designed to address or counter misinformation, nor were any misinformation-related variables measured.
However, while the connection was not made explicit, many of the mediators of loss framing described above reflect common themes in COVID misinformation. For example, perceived risk of vaccine side effects seems to reflect the frequency of messaging that exaggerates the commonality and severity of those effects (Nguyen & Catalan, 2020).
RQ6: Are Descriptions of Theories and Theoretical Constructs/Predictors Consistent? Is There Consensus on Conceptual Definitions and Measurement?
Inconsistent terminology and lack of consensus on defining and measuring theory-related concepts were significant challenges that made it difficult to compare and synthesize results. In several instances, studies used different terminology to describe conceptually similar phenomena. For example, Chang et al. (2021) designed messages emphasizing either vaccine safety or the consequences of not vaccinating, labeling these groups “Vaccine Safety” and “Health Consequences.” Conceptually, these are quite similar to loss/gain frames, but the study made no mention of Prospect Theory or loss/gain framing, so it was not coded as such. At other times, theory-relevant constructs appeared to be present but were not labeled as such and were not referenced in the context of theory, so they were not coded as a theoretical construct or predictor. For example, Peng et al. (2021) found that participants in the gain frame condition whose family members and/or friends were vaccinated reported significantly greater intentions to get vaccinated than those who did not have friends or family who were vaccinated. This is a descriptive norm, but it was not labeled or measured as such, so it was not coded as a theoretical construct or predictor.
This seems to reflect the degree to which theoretical concepts in the social sciences often come into common understanding and use within the field, without retaining a consistent connection to the original theory. Many theory-relevant studies miss opportunities both to strengthen intervention design via more systematic reference to a full portfolio of interrelated theoretical constructs and to actively test and refine theories that could benefit the field as a whole.
RQ7: What Gaps in Application of Theory Suggest Potential New Research Directions?
Given the unprecedented and ongoing role of misinformation in promoting vaccine hesitancy, it is surprising that more studies did not explicitly target misinformation and/or its effects in the process of intervention design. One promising avenue for future research is to use theory to explore the pathways and mechanisms of action through which misinformation influences health behavior, and then use the findings to identify promising targets for public health messaging. Future research could also explore how theories such as the HBM could be applied to increase public understanding of, and resilience to, susceptibility to misinformation. This could be done as part of a broader effort to encourage people to think more critically about the relationships between misinformation and health, as proposed by Houlden et al. (2021). Furthermore, future research should seek to operationalize and measure misinformation as the multidimensional construct it is, capturing dynamics such as exposure to, engagement with, and belief in various forms of misinformation.
It is also surprising that we did not find any health communication interventions specifically targeting COVID-19 vaccination intentions and behaviors among pregnant persons. Research suggests that pregnant persons have significant concerns about the safety of the COVID-19 vaccine and, as a result, are less likely to get vaccinated (Geoghegan et al., 2021; Levy et al., 2021; Razzaghi et al., 2021). In addition, pregnant persons may be particularly susceptible to misinformation about the COVID-19 vaccine (Sajjadi et al., 2021). As such, there is a pressing need for research aimed at developing effective communication techniques to promote vaccination during pregnancy. Theoretical frameworks such as Self-Determination Theory and Health Self-Empowerment Theory may be particularly suitable for use in designing messaging targeted toward pregnant persons (Olander et al., 2018; Rockliffe et al., 2021). Relatedly, none of the studies in the review used theory to tailor messages or select participants, suggesting a possible unmet need for communication and messaging campaigns aimed at specific subgroups of the population, such as those who may be particularly susceptible to misinformation and/or who may respond to specific message content, messengers, and/or channels.
In addition, most studies in the review targeted individual-level determinants of behavior and were unimodal in nature. Given the complexity of vaccination and the influence of organizational, cultural, and societal factors as determinants of vaccine hesitancy and acceptance, individual-level interventions may be limited in their impact. In future studies, theories such as the Social Ecological Model could be used to identify intervention targets beyond the individual to situate health communication campaigns within a broader, community-based approach to vaccine promotion.
Finally, there is a pressing need for more rigorous application of behavior change theory in intervention design and evaluation, including measuring key constructs and testing and/or refining theory. Without this kind of measurement of mediating causal relationships, all we can say about an intervention is whether or not it worked, but not why or how. Improving and standardizing measurement would also facilitate the synthesis and accumulation of evidence in reviews and meta-analyses.
Discussion
The purpose of this review was to evaluate the use of theory in the design of health communication interventions aimed at promoting COVID-19 vaccination. Overall, we found that most studies mentioned theory and targeted at least one theoretical construct or predictor of behavior, but few used theory optimally, none used a theoretical framework in its entirety, none used theory to select participants or tailor message content, and only one made an effort to refine theory. However, nine studies specified one or more pathways through which change was hypothesized to take place, making them at least partially theory based. Finally, only three studies mentioned misinformation, of which only two included intervention content addressing some aspect of misinformation, and only one measured the impact of this intervention content.
Several notable findings emerged from the literature regarding specific uses of theory and theory-relevant constructs and predictors. Prospect Theory was found to be the most widely used theoretical framework, and loss framing consistently outperformed gain-framing in terms of its impact on COVID-19 vaccination intentions. However, these effects were moderated and/or mediated by a variety of demographic and socio-psychological variables, including perceived risk, perceived benefits, attitudes, and age. Evidence from mediation and moderation tests supported the explanatory power of constructs from the HBM, particularly perceived benefits and perceived risk.
We were unable to clearly assess whether use of theory was associated with the success of the interventions, in large part because variation in methodology and outcomes made it difficult to make comparisons of this nature. In addition, some of the models and theories used by studies in the review were not change theories and thus did not specify mechanisms of action or causal pathways through which behavior change is hypothesized to take place.
The results of this review should be interpreted with several limitations in mind. First, in line with the purpose of a scoping review, we did not conduct a quality assessment of the studies. There is significant variation in the methodological rigor of the included studies, which may impact the reliability and validity of the underlying findings. Second, since we used broad inclusion criteria to get a comprehensive view of the literature, there is also significant variation in intervention content and design which limited our ability to make direct comparisons between the studies. Third, 34 of the 36 studies in this review measured vaccine intentions, rather than vaccination. While intentions have been shown to be among the strongest predictors of behavior (Sheeran, 2002), it cannot be assumed that strong intentions will necessarily lead to behavior, particularly with a novel vaccine. Importantly, the one study in our review that measured both intentions and behavior reported that intentions were not strongly associated with vaccination (Chang et al., 2021). As more studies are released which post-date the launch of COVID-19 vaccination campaigns, rather than anticipating them, more evidence on this key relationship will hopefully become available.
To our knowledge, this review provides the first systematic evaluation of the use of theory in health communication interventions aimed at promoting COVID-19 vaccination. The findings from this review provide critical insight into the state of theory-based intervention design and point to significant gaps in the literature to prioritize in future research. We hope this study will serve as a solid foundation to build from as the field of health communication faces the dual challenges of mitigating an infodemic and ending the largest pandemic of our lifetime.
Author Biographies
Dr. Caroline Orr is a postdoctoral research associate at the Applied Research Laboratory for Intelligence and Security (ARLIS) at the University of Maryland (UMD), where she studies misinformation, disinformation, and malinformation related to COVID and vaccination, as well as conducting research in the domain of cognitive security.
Dr. Ruthanna Gordon is an Associate Research Scientist at the Applied Research Laboratory for Intelligence and Security, a University-Affiliated Research Center of the United States Department of Defense supporting the Intelligence Community. She is on the leadership team for the Cognitive Security mission area, with a focus on influence detection, characterization, impact, and defense. She recently completed the US COVID Response project, exploring how the coronavirus pandemic has affected the information environment, and contributes to independent evaluation for the DARPA INCAS program. Prior to ARLIS, Gordon worked with Booz Allen Hamilton as a subject matter expert supporting the Intelligence Advanced Research Projects Activity (IARPA), and as a AAAS Science and Technology Policy fellow at the US Environmental Protection Agency. She received her PhD in Experimental Psychology from Stony Brook University in 2003.
Appendix 1: TCS Scoring Instructions (Adapted From Michie and Prestwich (2010))
Items on the TCS can be treated individually as well as grouped together to form composite measures reflecting the extent and function of their use. In this analysis, items were analyzed individually and, in some cases, composite measures were created to reflect specific uses of theory. Six composite measures were created based on the scoring criteria developed by Prestwich et al. (2014). The measures reflect the following:
1) Was theory mentioned?
Four items on the TCS reflect whether theory and/or theoretical predictors of behavior were explicitly mentioned. Item 1 assessed whether the study mentioned a theory, even if theory was not used to inform the intervention. Item 2 assessed whether theoretical predictors of vaccination behavior were explicitly mentioned (and also targeted). Item 3 assessed whether the intervention was based on a single theory (rather than multiple theories or a combination of theoretical predictors). Item 16 assessed whether the results were discussed in relation to theory. A total score was calculated by summing the scores of these four items, where “yes” = 1 and “no” = 0. Thus, total scores for this category ranged from 0 (no mention of theory or theoretical predictors) to 4 (optimal use of theory).
2) Were relevant theoretical constructs targeted?
Five items on the TCS reflect whether relevant theoretical constructs were targeted in the intervention. Item 5 assessed whether intervention techniques were based on a theory, theoretical predictor, or combination of theories and/or predictors. Items 7–10 examined the extent to which the intervention targeted specific theory-relevant constructs. Items 7 and 9 reflect optimal use of theory, indicating that all intervention techniques are linked to a theory-relevant predictor (item 7) and all theory-relevant predictors mentioned in the article text are associated with a specific intervention technique (item 9). Items 8 and 10 reflect less optimal use of theory, indicating an indirect link between intervention techniques and theoretical constructs/predictors (and vice versa). A total score was calculated by summing the scores on item 5 (“yes” = 1; “no” = 0), items 7 and 9, and items 10 and 11. Studies coded “yes” on items 5, 7, and 9 were given a score of 1. Studies coded “yes” on items 8 and 10 were given a score of 0.5. Thus, total scores for this subscale ranged from 0 (no theory-use) to 3 (optimal use of theory).
3) Was theory used to select participants or tailor interventions?
Two items assessed the use of theory to select participants and/or tailor intervention techniques for individual participants. Item 4 assessed whether theory was used to select participants based on their scores or levels on a particular theoretical construct or predictor. Item 6 assessed whether theory was used to tailor the intervention to the needs of individual participants. A total score was calculated by summing the scores on items 4 and 6, where “yes” = 1 and “no” = 0. Thus, total scores ranged from 0 (no use of theory) to 2 (optimal use of theory).
4) Were relevant theoretical constructs measured?
One item (11) assessed whether the targeted theoretical constructs were measured. If at least one of the targeted constructs/predictors was measured pre/post intervention or post-intervention, the item was coded as “yes.” If the construct/predictor was not measured or if it was only measured pre-intervention, the item was coded as “no.” Thus, total scores for this measure ranged from 0 (no theoretical constructs were measured) to 1 (at least one theoretical construct was measured pre–post or post-intervention).
5) Is theory tested or refined?
Four items on the TCS reflect the extent and nature of theory testing. Item 14 assessed whether the intervention led to significant changes in at least one targeted theoretical construct, and items 15, 17, and 18 assessed whether these changes explained the intervention effect. Item 15 assessed whether the study provided evidence that changes in the theoretical construct led to changes in behavior through mediational analysis. Item 17 assessed whether the results provide appropriate evidence to support or refute the theory. Item 18 assessed whether the results were used to refine theory by either adding or removing constructs, or specifying changes that should be made to the interrelationships between theoretical constructs. A total score for was calculated by summing the scores of items 14, 15, 17, and 18, where “yes” = 1 and “no” = 0. Thus, total scores ranged from 0 (no theory testing or refinement) to 4 (optimal theory testing and refinement).
6) Overall use of theory.
A total theory score was calculated by summing the totals of the composite measures, where a score of zero reflected minimum (inadequate) use of theory, and a score of 14 reflected maximum (optimal) use of theory.
TCS Modifications
We made several specifications and minor adaptations to the TCS to modify it for the purposes of this review. First, we agreed to use definitions of “intervention” and “intervention techniques” that focused on communication strategies and message testing. Second, for items that assessed “all theory-relevant constructs or predictors,” such as Item 9 (“All theory-relevant constructs/predictors are explicitly linked to at least one intervention technique”), we interpreted this to mean “all theory-relevant constructs or predictors that were explicitly mentioned in the article,” rather than all possible constructs associated with a given theory. A third modification was made to eliminate one item that assesses whether groups of intervention techniques are linked to groups of theoretical constructs/predictors. Given the nature of the articles in our review, we were unlikely to have studies with multiple intervention techniques, and thus the item was not applicable to our sample. Furthermore, the items assessing “at least one” and “all” intervention techniques were deemed to be sufficient without an additional category. A final modification was made for ease of scoring, and involved assigning a score of “0.5” to two items reflecting less optimal theory-use. The original rubric calls for scoring those items as “1” and increasing the score of another set of items to “2,” but we chose to keep the scale consistent so that a score of “1” indicated optimal theory-use across all items.
Appendix 2
Keywords for Literature Search.
Communication/messaging search terms Vaccine search terms
Messaging intervention COVID vaccine
Messaging campaign COVID vaccination
Mass media intervention Coronavirus vaccine
Mass media campaign Coronavirus vaccination
Social media intervention Pfizer vaccine
Social media campaign Moderna vaccine
Social media messaging AstraZeneca vaccine
Mobile technology intervention
Mobile technology campaign
Social marketing
Media campaign
Fear appeals
Persuasive messaging
Gain/loss framing
Message framing
Communication intervention
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding: The authors received no financial support for the research, authorship, and/or publication of this article.
==== Refs
References
Abdel-Raheem A. Alkhammash R. (2021). ‘To get or not to get vaccinated against COVID-19’: Saudi women, vaccine hesitancy, and framing effects. Discourse & Communication, 16 (1 ), 21–36. 10.1177/17504813211043724
Ajzen I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50 (2 ), 179–211. 10.1016/0749-5978(91)90020-T
Albada A. Ausems M. G. Bensing J. M. van Dulmen S. (2009). Tailored information about cancer risk and screening: A systematic review. Patient Education and Counseling, 77 (2 ), 155–171.19376676
Albarracín D. Gillette J. C. Earl A. N. Glasman L. R. Durantini M. R. Ho M. H. (2005). A test of major assumptions about behavior change: A comprehensive look at the effects of passive and active HIV-prevention interventions since the beginning of the epidemic. Psychological Bulletin, 131 (6 ), 856–897.16351327
Altay S. Hacquin A. S. Chevallier C. Mercier H. (2021). Information delivered by a chatbot has a positive impact on COVID-19 vaccines attitudes and intentions. Journal of Experimental Psychology: Applied. Advance online publication. 10.1037/xap0000400
Argote P. Barham E. Daly S. Z. Gerez J. E. Marshall J. Pocasangre O. (2021). The shot, the message, and the messenger: COVID-19 vaccine acceptance in Latin America. NPJ Vaccines, 6 (1 ), 118.34593822
Ashworth M. Thunström L. Cherry T. L. Newbold S. C. Finnoff D. C. (2021). Emphasize personal health benefits to boost COVID-19 vaccination rates. Proceedings of the National Academy of Sciences, 118 (32 ), e2108225118. 10.1073/pnas.2108225118
Aw J. Seng J. Seah S. Low L. L. (2021). COVID-19 vaccine hesitancy-a scoping review of literature in high-income countries. Vaccines, 9 (8 ), 900. 10.3390/vaccines9080900 34452026
Bandura A. (1986). Social foundations of thought and action: A social cognitive theory. Prentice-Hall.
Baumgartner J. (2020). Framing the COVID-19 vaccine: An experimental investigation of the influence of gain and loss frames on intention to receive COVID-19 vaccination [Master’s thesis, Lund University]. https://lup.lub.lu.se/luur/download?func=downloadFile&recordOId=9059803&fileOId=9059804
Bogart L. M. Ojikutu B. O. Tyagi K. Klein D. J. Mutchler M. G. Dong L. Lawrence S. J. Thomas D. R. Kellman S. (2021). COVID-19 related medical mistrust, health impacts, and potential vaccine hesitancy among Black Americans living with HIV. Journal of Acquired Immune Deficiency Syndromes, 86 (2 ), 200–207.33196555
Borah P. (2022). Message framing and COVID-19 vaccination intention: Moderating roles of partisan media use and pre-attitudes about vaccination. Current Psychology. Advance online publication. 10.1007/s12144-022-02851-3
Borah P. Hwang J. Hsu Y. C. (2021). COVID-19 vaccination attitudes and intention: Message framing and the moderating role of perceived vaccine benefits. Journal of Health Communication, 26 (8 ), 523–533.34424140
Bullock J. Lane J. E. Shults F. L. (2022). What causes COVID-19 vaccine hesitancy? Ignorance and the lack of bliss in the United Kingdom. Humanities and Social Science Communication, 9 (1 ), 87. 10.1057/s41599-022-01092-w
Cane J. O’Connor D. Michie S. (2012). Validation of the theoretical domains framework for use in behaviour change and implementation research. Implementation Science, 7 (1 ), 37. 10.1186/1748-5908-7-37 22530986
Capasso M. Caso D. Conner M. (2021). Anticipating pride or regret? Effects of anticipated affect focused persuasive messages on intention to get vaccinated against COVID-19. Social Science & Medicine, 289 , 114416. 10.1016/j.socscimed.2021.114416 34562773
Centers for Disease Control and Prevention (CDC). (1999). Ten great public health achievements — United States, 1900–1999. https://www.cdc.gov/mmwr/pdf/wk/mm4812.pdf
Chang T. Jacobson M. Shah M. Pramanik R. Shah S. B. (2021). Financial incentives and other nudges do not increase COVID-19 vaccinations among the vaccine hesitant (Working Paper No. 29403). National Bureau of Economic Research. http://www.nber.org/papers/w29403
Chen T. Dai M. Xia S. Zhou Y. (2022). Do messages matter? Investigating the combined effects of framing, outcome uncertainty, and number format on COVID-19 vaccination attitudes and intention. Health Communication, 37 (8 ), 944–951.33499677
Chu H. Liu S. (2021). Integrating health behavior theories to predict American’s intention to receive a COVID-19 vaccine. Patient Education and Counseling, 104 (8 ), 1878–1886.33632632
Cohen J. (1960). A coefficient of agreement for nominal scales. Educational and Psychological Measurement, 20 (1 ), 37–46.
Dai H. Saccardo S. Han M. A. Roh L. Raja N. Vangala S. Modi H. Pandya S. Sloyan M. Croymans D. M. (2021). Behavioral nudges increase COVID-19 vaccinations. Nature, 597 (7876 ), 404–409.34340242
Davies P. Walker A. E. Grimshaw J. M. (2010). A systematic review of the use of theory in the design of guideline dissemination and implementation strategies and interpretation of the results of rigorous evaluations. Implementation Science, 5 (1 ), 14.20181130
Diament S. M. Kaya A. Magenheim E. B. (2022). Frames that matter: Increasing the willingness to get the COVID-19 vaccines. Social Science & Medicine, 292 , 114562. 10.1016/j.socscimed.2021.114562 34810030
Dombrowski S. U. Sniehotta F. F. Avenell A. Coyne J. C. (2007). Towards a cumulative science of behavior change: Do current conduct and reporting of behavioral interventions fall short of best practice. Psychological Health, 22 (8 ), 869–874.
Dubé E. Gagnon D. MacDonald N. E. , SAGE Working Group on Vaccine Hesitancy. (2015). Strategies intended to address vaccine hesitancy: Review of published reviews. Vaccine, 33 (34 ), 4191–203.
Freeman D. Loe B. S. Yu L. M. Freeman J. Chadwick A. Vaccari C. Shanyinde M. Harris V. Waite F. Rosebrock L. Petit A. Vanderslott S. Lewandowsky S. Larkin M. Innocenti S. Pollard A. J. McShane H. Lambe S. (2021). Effects of different types of written vaccination information on COVID-19 vaccine hesitancy in the UK (OCEANS-III): A single-blind, parallel-group, randomized controlled trial. The Lancet Public Health, 6 (6 ), e416–e427.
Gardner B. Whittington C. McAteer J. Eccles M. P. Michie S. (2010). Using theory to synthesise evidence from behaviour change interventions: The example of audit and feedback. Social Science & Medicine, 70 (10 ), 1618–1625.20207464
Geoghegan S. Stephens L. C. Feemster K. A. Drew R. J. Eogan M. Butler K. M. (2021). “This choice does not just affect me.” Attitudes of pregnant women toward COVID-19 vaccines: A mixed-methods study. Human Vaccines & Immunotherapeutics, 17 (10 ), 3371–3376.34009096
Giampaolo N. (2021). Regretting vaccine indecision: Solving COVID-19 vaccine hesitancy through anticipated regret-nudging [Master's thesis, University of Twente]. http://essay.utwente.nl/88815/
Glanz K. Rimer B. K. Viswanath K. (2008). Health behavior and health education: Theory, research, and practice. John Wiley & Sons.
Golos A. M. Hopkins D. J. Bhanot S. P. Buttenheim A. M. (2022). Partisanship, messaging, and the COVID-19 vaccine: Evidence from survey experiments. American Journal of Health Promotion 36 (4 ), 602–611. 10.1177/08901171211049241 35232232
Gong Z. Tang Z. Li J. (2021). What strategy is better for promoting COVID-19 vaccination? A comparison between gain-framed, loss-framed, and altruistic messages. Annals of Behavioral Medicine, 56 (4 ), 325–331.
Gursoy D. Ekinci Y. Can A. S. Murray J. C. (2022). Effectiveness of message framing in changing COVID-19 vaccination intentions: Moderating role of travel desire. Tourism Management, 90 , 104468. 10.1016/j.tourman.2021.104468 34898791
Hardeman W. Johnston M. Johnston D. Bonetti D. Wareham N. Kinmonth A. L. (2002). Application of the theory of planned behavior in behavior change interventions: A systematic review. Psychology and Health, 17 (2 ), 123–158.
Hong Y. Hashimoto M. (2021). I will get myself vaccinated for others: The interplay of message frame, reference point, and perceived risk on intention for COVID-19 vaccine. Health Communication. Advance online publication. 10.1080/10410236.2021.1978668
Houlden S. Hodson J. Veletsianos G. Reid D. Thompson-Wagner C. (2021). The health belief model: How public health can address the misinformation crisis beyond COVID-19. Public Health in Practice (Oxford, England), 2 , 100151. 10.1016/j.puhip.2021.100151 34494011
Huang Y. Liu W. (2021). Promoting COVID-19 vaccination: The interplay of message framing, psychological uncertainty, and public agency as a message source. Science Communication, 44 (1 ), 3–29. https://doi.org/10755470211048192
James E. K. Bokemper S. E. Gerber A. S. Omer S. B. Huber G. A. (2021). Persuasive messaging to increase COVID-19 vaccine uptake intentions. Vaccine, 39 (49 ), 7158–7165.34774363
Jarrett C. Wilson R. O’Leary M. Eckersberger E. Larson H. J. (2015). Strategies for addressing vaccine hesitancy–a systematic review. Vaccine, 33 (34 ), 4180–4190.25896377
Karlsson L. C. Soveri A. Lewandowsky S. Karlsson L. Karlsson H. Nolvi S. Karukivi M. Lindfelt M. Antfolk J. (2021). Fearing the disease or the vaccine: The case of COVID-19. Personality and Individual Differences, 172 , 110590.33518869
Kenzig M. J. Mumford N. S. (2022). Theoretical considerations for communication campaigns to address vaccine hesitancy. Health Promotion Practice, 23 (1 ), 46–50.34672826
Kerr J. R. Freeman A. L. Marteau T. M. van der Linden S. (2021). Effect of information about COVID-19 vaccine effectiveness and side effects on behavioral intentions: Two online experiments. Vaccines, 9 (4 ), 379.33924542
Lara J. Evans E. H. O’Brien N. Moynihan P. J. Meyer T. D. Adamson A. J. Errington L. Sniehotta F. F. White M. Mathers J. C. (2014). Association of behavior change techniques with effectiveness of dietary interventions among adults of retirement age: A systematic review and meta-analysis of randomized controlled trials. BMC Medicine, 12 (1 ), 177.25288375
Levac D. Colquhoun H. O’Brien K. K. (2010). Scoping studies: Advancing the methodology. Implementation Science, 5 (1 ), 69.20854677
Levy A. T. Singh S. Riley L. E. Prabhu M. (2021). Acceptance of COVID-19 vaccination in pregnancy: A survey study. American Journal of Obstetrics & Gynecology MFM, 3 (5 ), 100399. 10.1016/j.ajogmf.2021.100399 34020098
Li J. Tang Z. Gong Z. (2022). Does the message frame promote people’s willingness to vaccinate when they worry about side effects from the COVID-19 vaccine? Evidence from an online survey experiment in China. Health Communication. Advance online publication. 10.1080/10410236.2022.2028469
Loomba S. de Figueiredo A. Piatek S. J. de Graaf K. Larson H. J. (2021). Measuring the impact of COVID-19 vaccine misinformation on vaccination intent in the UK and USA. Nature Human Behavior, 5 (3 ), 337–348.
Michie S. Abraham C. (2004). Interventions to change health behaviors: Evidence-based or evidence-inspired? Psychology & Health, 19 (1 ), 29–49.
Michie S. Johnston M. Abraham C. Lawton R. Parker D. Walker A. (2005). Making psychological theory useful for implementing evidence based practice: A consensus approach. BMJ Quality & Safety, 14 (1 ), 26–33.
Michie S. Johnston M. Francis J. Hardeman W. Eccles M. (2008). From theory to intervention: Mapping theoretically derived behavioral determinants to behavior change techniques. Applied Psychology, 57 (4 ), 660–680.
Michie S. Prestwich A. (2010). Are interventions theory-based? Development of a Theory Coding Scheme. Health Psychology, 29 (1 ), 1–8.20063930
Moola S. Gudi N. Nambiar D. Dumka N. Ahmed T. Sonawane I. R. Kotwal A. (2021). A rapid review of evidence on the determinants of and strategies for COVID-19 vaccine acceptance in low- and middle-income countries. Journal of Global Health, 11 , 05027. 10.7189/jogh.11.05027 34912550
Motta M. Sylvester S. Callaghan T. Lunz-Trujillo K. (2021). Encouraging COVID-19 vaccine uptake through effective health communication. Frontiers in Political Science, 3 , 630133. 10.3389/fpos.2021.630133
New York Times. (2022). Tracking coronavirus vaccinations around the world. https://www.nytimes.com/interactive/2021/world/covid-vaccinations-tracker.html
Nguyen A. Catalan D. (2020). Digital mis/disinformation and public engagement with health and science controversies: Fresh perspectives from COVID-19. Media and Communication, 8 (2 ), 323–328.
Noar S. M. Zimmerman R. S. (2005). Health behavior theory and cumulative knowledge regarding health behaviors: Are we moving in the right direction? Health Education Research, 20 (3 ), 275–290.15632099
Olander E. K. Smith D. M. Darwin Z. (2018). Health behavior and pregnancy: A time for change. Journal of Reproductive and Infant Psychology, 36 (1 ), 1–3.29517295
Olson O. Berry C. Kumar N. (2020). Addressing parental vaccine hesitancy towards childhood vaccines in the United States: A systematic literature review of communication interventions and strategies. Vaccines, 8 (4 ), 590.33049956
Orr C. A. (2020). Behavior change techniques to promote smoking cessation during pregnancy: A theory-based meta-analysis [Doctoral dissertation, Virginia Commonwealth University]. 10.25772/12SC-QM19
Painter J. E. Borba C. P. Hynes M. Mays D. Glanz K. (2008). The use of theory in health behavior research from 2000 to 2005: A systematic review. Annals of Behavioral Medicine, 35 (3 ), 358–362.18633685
Palm R. Bolsen T. Kingsland J. T. (2021). The effect of frames on COVID-19 vaccine resistance. Frontiers in Political Science, 3 , 661257. 10.3389/fpos.2021.661257
Peng L. Guo Y. Hu D. (2021). Information framing effect on public’s intention to receive the COVID-19 vaccination in China. Vaccines, 9 (9 ), 995.34579232
Penţa M. A. Băban A. (2018). Message framing in vaccine communication: A systematic review of published literature. Health Communication, 33 (3 ), 299–314.28059557
Prestwich A. Sniehotta F. F. Whittington C. Dombrowski S. U. Rogers L. Michie S. (2014). Does theory influence the effectiveness of health behavior interventions? Meta-analysis. Health Psychology, 33 (5 ), 465.23730717
Rabb N. Swindal M. Glick D. Bowers J. Tomasulo A. Oyelami Z. Wilson K.H. Yokum D. (2021). The limits of nudges: Results of a statewide vaccination RCT. Brown University Policy Lab. https://osf.io/srckw/
Razzaghi H. Meghani M. Pingali C. Crane B. Naleway A. Weintraub E. Kenigsberg T. A. Lamias M. J. Irving S. A. Kauffman T. L. Vesco K. K. Daley M. F. DeSilva M. Donahue J. Getahun D. Glenn S. Hambidge S. J. Jackson L. Lipkind H. S. . . . Patel S. A. (2021). COVID-19 vaccination coverage among pregnant women during pregnancy — eight integrated health care organizations, United States, December 14, 2020–May 8, 2021. MMWR Morbidity and Mortality Weekly Reports, 70 (24 ), 895–899. 10.15585/mmwr.mm7024e2
Reinhardt A. Rossmann C. (2021). Age-related framing effects: Why vaccination against COVID-19 should be promoted differently in younger and older adults. Journal of Experimental Psychology: Applied, 27 (4 ), 669–678. 10.1037/xap0000378 34292048
Reñosa M. D. C. Landicho J. Wachinger J. Dalglish S. L. Bärnighausen K. Bärnighausen T. McMahon S. A. (2021). Nudging toward vaccination: A systematic review. BMJ Global Health, 6 (9 ), e006237.
Rockliffe L. Peters S. Heazell A. E. Smith D. M. (2021). Factors influencing health behavior change during pregnancy: A systematic review and meta-synthesis. Health Psychology Review, 15 (4 ), 613–632.34092185
Rogers R. W. (1975). A protection motivation theory of fear appeals and attitude change. Journal of Psychology, 91 (1 ), 93–114. 10.1080/00223980.1975.9915803 28136248
Rosenstock I. M. (1974). The health belief model and preventative behavior. Health Education Monographs, 2 (4 ), 27–59.
Sajjadi N. B. Nowlin W. Nowlin R. Wenger D. Beal J. M. Vassar M. Hartwell M. (2021). United States internet searches for “infertility” following COVID-19 vaccine misinformation. Journal of Osteopathic Medicine, 121 (6 ), 583–587.33838086
Santos H. C. Goren A. Chabris C. F. Meyer M. N. (2021). Effect of targeted behavioral science messages on COVID-19 vaccination registration among employees of a large health system: A randomized trial. JAMA Network Open, 4 (7 ), e2118702–e2118702.
Sasaki S. Saito T. Ohtake F. (2022). Nudges for COVID-19 voluntary vaccination: How to explain peer information? Social Science and Medicine, 292 , 114561. 10.1016/j.socscimed.2021.114561 34823128
Schmid P. Rauber D. Betsch C. Lidolt G. Denker M. L. (2017). Barriers of influenza vaccination intention and behavior–a systematic review of influenza vaccine hesitancy, 2005–2016. PLoS One, 12 (1 ), e0170550.
Seo K. W. (2008). Effects of visual images in health message framing [Master’s thesis]. https://etda.libraries.psu.edu/files/final_submissions/3668
Sheeran P. (2002). Intention—behavior relations: A conceptual and empirical review. European Review of Social Psychology, 12 (1 ), 1–36.
Simonovic N. Taber J. M. (2022). Psychological impact of ambiguous health messages about COVID-19. Journal of Behavioral Medicine, 45 (2 ), 159–171.34811623
Sinclair S. Agerström J. (2021). Do social norms influence young people’s willingness to take the COVID-19 vaccine? Health Communication. Advance online publication. 10.1080/10410236.2021.1937832
Strickland J. C. Reed D. D. Hursh S. R. Schwartz L. P. Foster R. N. Gelino B. W. LeComte R. S. Oda F. S. Salzer A. R. Schneider T. D. Dayton L. Latkin C. Johnson M. W. (2022). Behavioral economic methods to inform infectious disease response: Prevention, testing, and vaccination in the COVID-19 pandemic. PLoS One, 17 (1 ), e0258828.
Taylor N. Conner M. Lawton R. (2012). The impact of theory on the effectiveness of worksite physical activity interventions: A meta-analysis and meta-regression. Health Psychology Review, 6 (1 ), 33–73.
Ugwuoke J. C. Talabi F. O. Adelabu O. Sanusi B. O. Gever V. C. Onuora C. (2021). Expanding the boundaries of vaccine discourse: Impact of visual illustrations communication intervention on intention towards COVID-19 vaccination among victims of insecurity in Nigeria. Human Vaccines & Immunotherapeutics, 17 (10 ), 3450–3456.34128754
Webb T. Joseph J. Yardley L. Michie S. (2010). Using the internet to promote health behavior change: A systematic review and meta-analysis of the impact of theoretical basis, use of behavior change techniques, and mode of delivery on efficacy. Journal of Medical Internet Research, 12 (1 ), e1376.
WHO. (2014). Report of the SAGE working group on vaccine hesitancy. https://www.who.int/immunization/sage/meetings/2014/october/1_Report_WORKING_GROUP_vaccine_hesitancy_final.pdf
WHO. (2020a). 1st WHO infodemiology conference. https://www.who.int/news-room/events/detail/2020/06/30/default-calendar/1st-who-infodemiology-conference
WHO. (2020b). Infodemic management—infodemiology. https://www.who.int/teams/risk-communication/infodemic-management
Witte K. (1992). Putting the fear back into fear appeals: The extended parallel process model. Communication Monographs, 59 (4 ), 329–349. 10.1080/03637759209376276
Wong M. C. Wong E. L. Huang J. Cheung A. W. Law K. Chong M. K. … Chan P. K. (2021). Acceptance of the COVID-19 vaccine based on the health belief model: A population-based survey in Hong Kong. Vaccine, 39 (7 ), 1148–1156.
Xiao X. Wong R. (2020). Vaccine hesitancy and perceived behavioral control: A meta-analysis. Vaccine, 38 (33 ), 5131–5138. 10.1016/j.vaccine.2020.04.076 32409135
Yang C. (2022): Exploring communication strategies to encourage COVID-19 vaccination: Motivation-based message appeals, incidental emotions, and risk perception. Health Communication. Advance online publication. 10.1080/10410236.2022.2028481
Ye W. Li Q. Yu S. (2021). Persuasive effects of message framing and narrative format on promoting COVID-19 vaccination: A study on Chinese college students. International Journal of Environmental Research and Public Health, 18 (18 ), 9485.34574406
Yuan S. Chu H. (2022). Vaccine for yourself, your community, or your country? Examining audiences’ response to distance framing of COVID-19 vaccine messages. Patient Education and Counseling, 105 (2 ), 284–289.34479746
| 0 | PMC9703017 | NO-CC CODE | 2022-11-29 23:21:05 | no | Am Behav Sci. 2022 Nov 24;:00027642221138274 | utf-8 | Am Behav Sci | 2,022 | 10.1177/00027642221138274 | oa_other |
==== Front
Sociol Res Online
Sociol Res Online
SRO
spsro
Sociological Research Online
1360-7804
SAGE Publications Sage UK: London, England
10.1177/13607804221115433
10.1177_13607804221115433
Article
The Mode of Reflexive Practice among Young Indonesian Creative Workers in the Time of COVID-19
Sutopo Oki Rahadianto Universitas Gadjah Mada, Indonesia
Wibawanto Gregorius Ragil Universitas Gadjah Mada, Indonesia
Utomo Ariane The University of Melbourne, Australia
Beta Annisa R The University of Melbourne, Australia
Kurnia Novi Universitas Gadjah Mada, Indonesia
Oki Rahadianto Sutopo, Youth Studies Centre, Faculty of Social and Political Sciences, Universitas Gadjah Mada, Sosio-Justicia No. 2 Bulaksumur, Yogyakarta 55281, Indonesia. Email: [email protected]
25 11 2022
25 11 2022
1360780422111543322 10 2021
4 7 2022
© The Author(s) 2022
2022
SAGE Publications and the British Sociological Association
This article is made available via the PMC Open Access Subset for unrestricted re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the COVID-19 pandemic or until permissions are revoked in writing. Upon expiration of these permissions, PMC is granted a perpetual license to make this article available via PMC and Europe PMC, consistent with existing copyright protections.
This article examines reflexive practice among young creative workers in Yogyakarta, Indonesia, during COVID-19. Since March 2020, the COVID-19 pandemic has unleashed a series of relentless and overlapping crises across the Indonesian archipelago. In urban centres across Indonesia, the arts and creative sectors are among the key economic sectors severely afflicted by the pandemic. COVID-19 implies a lot more than the loss of income and livelihoods. Mobility restrictions, gig cancellations, venue closures, all entail the loss of connections, opportunities, and creative outlets. Yet despite such uncertain conditions, young creative workers remain reflexively creative in order to survive in everyday life. Building upon interviews and focus-group discussions with young creative workers in Yogyakarta, we found three modes of temporality-based reflexive practice: waiting, doing something and re-learning, which represent young creative workers’ active responses manifested in the practical and contradictory relationship to the diverse possibilities within hierarchical and heterogenous cultural fields in a pandemic era characterised by regular ruptures. The analysis of the data below contributes to the literature on reflexivity and habitus among young creative workers in a time of pandemic.
creativity
Indonesia
habitus
pandemic
reflexivity
work
youth
Australia-Indonesia Centre edited-statecorrected-proof
typesetterts1
==== Body
pmcIntroduction
Since March 2020, the COVID-19 pandemic has unleashed a series of relentless and overlapping crises across the Indonesian archipelago. The general uncertainties associated with the peaks and troughs of a constantly evolving pandemic – combined with a lack of coordinated and effective government response (Setijadi, 2021) – have disrupted the livelihoods of many Indonesians. By mid-July 2021, with daily reported new cases of more than 50,000 and COVID-19 mortality in excess of 900 deaths per day, Indonesia became the global epicentre of the pandemic (Australian Broadcasting Corporation, 2021).
In urban centres across Indonesia, the arts and creative sectors are among the key economic sectors severely afflicted by the pandemic. As in many other parts of the world, venue closures, show cancellations, and the general calls to ‘stay home’ during lengthy mobility restrictions have amplified the precarious nature of the sector. Put simply, COVID-19 has paralysed artists and creative workers everywhere. But, consistent with how emerging societal issues have been typically represented in academia, much of the recent literature on the impacts of the COVID-19 pandemic on the creative economy to date has over-emphasised the experience of the Global North (see, for example, Comunian and England (2020) on the UK; Florida and Seman (2020) on the US; Betzler et al. (2021) for some European countries; Flew and Kirkwood (2021) and (Flore et al., 2021) on regional Australia; (Howard et al., 2021) on the comparison between Australia, England and Portugal).
Research on COVID-19 and creative workers in the context of lower middle-income economies remain relatively limited (see Joffe, 2021). In places like Indonesia, workers in the creative sector share many job characteristics with the large number of non-agricultural workers working in the informal sector of the labour market. These workers often lack union representation, lack job security, and are working in precarious conditions without adequate health and safety protection (Brata, 2010; Dartanto et al., 2015; Izzati et al., 2021). Moreover, given the general lack of welfare support in the country’s tax transfer system, creative workers – and workers working in the informal sector more generally – are confronted with the prospect of having limited/no access to State-sponsored social safety nets during mobility restrictions.
For young people working in the creative sector, COVID-19 implies a lot more than loss of income and livelihoods. As shown by earlier studies elsewhere, the impact of the COVID-19 crisis extends beyond the realm of financial hardship (Spiro et al., 2021). Mobility restrictions, gig cancellations, venue closures, all entail the loss of connections, opportunities, and creative outlets. All these have an important bearing on the health and well-being of creative workers (Spiro et al., 2021). To document and examine such challenges, this article focuses on the reflexive practice of young creative workers during the COVID-19 pandemic in Yogyakarta. Yogyakarta has been dubbed the city of culture in Indonesia. The city is home to 172,000 creative workers (Wicaksono, 2019), and the creative economy generated US$238 million in 2016 (Badan Ekonomi Kreatif (Bekraf) dan Badan Pusat Statistik (BPS), 2016). Young Yogyakartans in the creative industry have a diverse pool of talents with unequal distribution of valuable and relevant forms of cultural capital within and across their sub-fields. Our paper highlights their varied experience and reflexive practice to navigate COVID-19 pandemic-related challenges. Looking at the local fields of cultural production, we discovered three modes of reflexive practice, ‘waiting’, ‘doing something’, and ‘re-learning’, which show the complexities of interlinkage between temporality, reflexivity and habitus under conditions of crisis and disruption.
In the first section, the article maps previous studies on reflexivity and habitus. The next section examines Yogyakarta as a festive and reflexive space. In the end, we show plural narratives of survival among young creative workers in the time of pandemic, which manifest in the three modes of reflexive practice, namely ‘waiting’, ‘doing something’, and ‘re-learning’.
Reflexivity and habitus in a time of pandemic
Previous studies on reflexivity and habitus in the context of late modernity by youth studies scholars and sociologists, on the one hand, have highlighted the problems of being free from structural constrained agency and, on the other hand, too focused on deterministic and structuralist reading in explaining rapid social changes (see Adams, 2006; Coffey and Farrugia, 2014; Kelly, 2016; Sweetman, 2003; Threadgold, 2011; Woodman, 2009). In addition, some scholars have brought reflexivity and habitus together in order to understand the complexities of moving towards the reflexive modernisation era, as proposed by major sociological theorists such as Giddens (1991) and Beck (1992).
Notably, McNay (1999) explains the emergence of reflexivity provoked by the conflict and tensions of social forces operating within and across specific fields, thus highlighting the embeddedness of the subject within differing sets of power relations (p. 110). Sweetman (2003) argues that since the disjuncture between habitus and various fields of struggle has become increasingly ‘taken for granted’, flexible or reflexive habitus may become common and significant in the era of late modernity, as he claims that ‘reflexivity ceases to reflect a temporary lack of fit between habitus and field but itself becomes habitual’ (p. 541). However, being able to apply reflexive habitus in everyday life does not necessarily lead social agents to become ‘reflexivity winners’. This particular situation leads to discussion about what comes ‘after’ the moment of reflexive awareness, in which choices are resourced or otherwise (Adams, 2006, p. 523). Threadgold and Nilan (2009), building on ‘post-reflexive choice’ as suggested by Adams (2006), propose reflexivity as a form of cultural capital. Thus, they underscore the importance of socio-economic backgrounds in explaining the reproduction of social inequalities.
Mouzelis (2007) takes the discussion about reflexivity and habitus further using figuration theory. He proposes that in order to effectively account for social practices, habitus has to be connected with interactive and figurational structures (p. 4). Moreover, instead of viewing habitus as deterministic, Hilgers (2009) put forward freedom as an essential and necessary component in his analysis. For him, habitus determines practice but is also determined by it. Habitus is thus being in a state of permanent mutation (p. 731). Farrugia (2013) argues for a theory of reflexivity as actualising a practical intelligibility shaped by the dispositions of the habitus (p. 283). So, instead of viewing reflexivity as disembodied rationality, reflexive practices are embedded within the logic of fields and describe creative responses to local structural conditions (p. 296). In addition, Farrugia (2015) constructs reflexivity as social practice, which reflects the contradictions and insecurities intrinsic to modern social structures. Consequently, reflexivity is a concept that combines the macro and the micro, the structural, and the personal (p. 872).
The research in this article continues the previous discussions about reflexivity and habitus mentioned above, while differing in several aspects. First, the temporal context of this research is under conditions of disruption and crisis, specifically during the COVID-19 pandemic. It can be argued that the extraordinariness of this crisis is its unpredictable consequences at multiple levels of everyday life, both temporally and spatially, and ranging from local to global (see Connell, 2020; Matthewman and Huppatz, 2020). In short, the pandemic is ‘a monstrous threat’ (see Zinn, 2020). Moreover, we define temporality from a Bourdieusian perspective, which means time is understood in radically historicist terms as engendered through social being (McNay, 1999: 101). Time is what practical activity produces in the very act whereby it produces itself; in short, a non-ethical reference to a future inscribed in the immediacy of the present (Bourdieu and Wacquant, 1992: 138). Second, we critically apply the concept of ‘hysteresis’ (Bourdieu, 2000) not only to understand the rupture between habitus and field but also as theoretical bridging at the meso-level. Thus, analytically, ‘hysteresis’ mediates the interlinkage between crisis and disruption at the macro level and social agents’ reflexivity at the micro level. Third, we eclectically combine ‘hysteresis’ (Bourdieu, 2000) and reflexivity as social practice (Farrugia, 2015) in a time of pandemic to help us make an abstraction from empirical data about young creative workers’ strategies which resulted in three modes of reflexive practice, namely, ‘waiting’, ‘doing something’, and ‘re-learning’. For us, reflexive practice shows the complex interrelation between embodied dispositions and overlap yet disrupted fields of struggle.
Methods
The aim of our research is to understand the reflexive practice of young creative workers at the time of the COVID-19 pandemic in Yogyakarta, Indonesia, a cosmopolitan city with a sturdy atmosphere of activism, education and leisure. In particular, our participants are musicians, dancers, and theatre artists who predominantly rely on physical and non-digital activities to produce their artistic works. Due to mobility restrictions, data were gathered through online Focus Group Discussions (FGDs) and in-depth interviews using Zoom as well as other platforms such as WhatsApp and Google Meet. In total, we conducted two Focus Group Discussions with 12 informants, and we did in-depth interviews with 18 informants. Participants were recruited through our durable research networks with local music and art communities. We also approached the participants through Instagram and gained further insights of their artistic activities through their social media platforms. However, to give more weight to the voices of young creative workers, in-depth interviews rather than FGDs were selected as the principal data collection method. The in-depth interviews were conducted using a mix of Javanese, the local language, and Bahasa Indonesia, the national language. The participants were interviewed online for more or less 1.5 hours each.
During the process of data gathering, we aimed to build an equal and reciprocal relationship as well as sharing trust with the participants; therefore, they could share their subjective experiences freely, resulting in the voices of young creative workers being key to this research. These interviews and FGDs were held in October 2020, thus the research captures the situation in Yogyakarta after approximately 8 months of the pandemic. Analysis of the interviews and FGD data proceeded in various ways. First, the interviews and FGDs were transcribed and then translated from Javanese and Indonesian into English. Second, the transcripts were subjected to thematic analysis to create key themes, and third, the chosen quotations were assembled by key argument, and analysed using selected conceptual frameworks. Ethics approval for this project was granted by the University of Melbourne Faculty of Arts Human Ethics Advisory Group – Ethics ID 2057775.1.
Yogyakarta: a festive and reflexive space
Yogyakarta is globally known as a central space of arts and creative display. The city has hosted numerous popular annual arts festivals such as ArtJog, Jogja Biennale, Papermoon Puppet Festival, NgayogJazz Music Festival, and the Jogja-Netpac Asian Film Festival (JAFF) to name a few. Although each of the events has its own specific arts form to display, they often provide spaces for public artistic collaborations. Musicians perform in a film festival, dancers appear as an opening act for an art exhibition, and contemporary theatre performances are often seen in a dance festival. This cross-‘artistic’ and public interaction has helped the city become a hallmark of arts festivals, thus creating a distinctive identity for its city façade (Irawanto, 2010). This practice has also opened up global-scale collaboration between young creative workers from Indonesia and artists from abroad.
Aside from its arts festivals, the popular status of the city as a tourist destination has also contributed to its distinct features. Many cafés and hotels offer live music as entertainment (Suharyanto et al., 2021). Each of these urban entertainment sites has its own distinct music genre. Jazz is often played live in hotels; popular tunes are heard in cafés; and traditional repertoires are performed at cultural sites as well as at local festivals (see Sutopo et al., 2020). Some tourist destinations manage series of scheduled attractions such as court dance performances (Rindrasih and Witte, 2021). Similar art displays also take place in neighbouring suburbs, such as Ramayana Shows at the Prambanan Temple Complex (Sedana and Foley, 2020). These various spaces have become places of production, exploration, creation, and labour for young creative workers in Yogyakarta. Not only do they provide sites for display and presentation, these particular venues have also become physical fields of socialisation, learning, and embodiment of arts practice among young creative workers, thus, nurturing the accumulation of durable and strategic social capital (Sutopo et al., 2017) and embodied cultural capital (see Bourdieu, 1986; Wacquant, 2014).
In navigating those particular practices through various media and spaces, young creative workers in Yogyakarta are not necessarily institutionalised or dependent upon arts agencies or professional event organisers. Collaborative practice among artists has become common in production and learning processes. Tsui (2015) called this sort of art practice ‘the Jogja way’. This value contains a prevailing sense of community spirit embedded among artists and other arts practitioners in the creative and dynamic pursuit of resources to support DIY artistic endeavours (Tsui, 2015: 540, 542). This particular value can be seen across art fields in Yogyakarta. In the theatre community, the idea of communal liberation as a mode of creative production has become a collective value in the city since the New Order era in Indonesia (Bodden, 2007). In the music scene, the practice of nongkrong (hanging out) among musicians has provided a space ‘to accumulate social and cultural capital relevant to enhance their future career’ (Sutopo, 2019: 80). In this particular social activity, young musicians acquire knowledge of music production together with networks that are significant in their arts endeavours. A similar pattern of space-related activity as the main feature in Jogja is also apparent in other arts practices, as expressed by one of our informants, who is a dancer herself: For me, Jogja is a space [ruang]. I see a lot of arts places or spaces where dancers can work on their pieces either in a collective manner in which a network is allowed to use the place or commercial in its character. So, yes, Jogja has many places that we can make as a stage. It is not only in an art gallery, even somebody’s house can be transformed into one (stage). It depends on the creator’s creativity to modify their place to become an exhibition arena or a stage for performance. My point is that Jogja has many places to be transformed into an art space, given the creativity of the artists possess.
(Interview with Sekar – a dancer, 3 October 2020)
Sekar has been managing her own sanggar as a dancing laboratory. In reflecting deeper on her collective arts arena, she adds that ‘Sanggar Seni Kinanti Sekar is a place for us to learn collectively. We welcome any activities either to learn new experiences or just to watch a rehearsal. Just come. We provide the space for people because dancing is fun!’ A similar view is also expressed by Meyda – the founder of an independent theatre company in Yogyakarta. She tells us that Jogja is a very comfortable space to develop creative ideas. Because here, you have wide range of artists that you can build your idea with. If I compare this city with Lombok [East Indonesia], then it would feel very different [. . .] here, you can find many festivals and everyone can join if they find it interesting.
(Interview with Meyda – founder of Theatre Company, 10 October 2020)
Reflecting on Sekar and Meyda’s views, a place as a physical melting-pot seems to have become an important feature for young creative workers in Yogyakarta to learn and explore their fields as well as gaining social and cultural capital for their future artistic endeavours. During the early phase of the COVID-19 pandemic in Indonesia, all these places suddenly became less relevant. Social restrictions meant the dancers, musicians, and theatre workers in Yogyakarta – who were strongly attached to physical places – were struggling to find an adaptive mode of practice. Amid the crisis, the Indonesian government has so far directed the dominant policy narratives towards the digital economy. This rapid response is a continuation of the national development agenda that contains a vision of Making Indonesia 4.0 launched in 2018 (Ministry of Industry, 2018) and coincides with the creative economy development that has been fostered since 2014 (Fahmi et al., 2016, 2017). This particular state-promoted vision is not necessarily attuned to the creative and arts scene in the local context, including Yogyakarta.
As such, the adoption of digital economy practice among young creative workers requires continual engagement with a specific cultural-digital capital (Ignatow and Robinson, 2017). Young creative workers in Yogyakarta – who have long been attached to physical space in their activities – often experience challenges in adapting digital platforms into their practices. The enduring ‘Jogja way’ as a particular value does not always contain the digital element, yet it plays a prevailing role in how the community is dealing with the pandemic. Thus, structurally, we argued that pandemic demonstrates moments of extreme uncertainties both temporally and spatially which force young creative workers to be reflexive. According to Bourdieu (2000), multiple forms of reflexivity can emerge under conditions of ‘lack of fit’ between habitus and the field of struggle, in particular, reflexivity which remains oriented towards practice. Next, we discuss the reflexive practice employed by musicians, dancers, and theatre workers to face the rapid changes caused by the COVID-19 pandemic amid the dominant narratives of a state-promoted digital and creative vision. Building upon empirical data, we explore three modes of temporality-based reflexive practice to deal with the pandemic, namely waiting, doing something and re-learning. We notice that the field of cultural production is hierarchical and heterogenous; thus, valuable forms of capital are not distributed equally (see Bourdieu, 1993; Robbins, 2000).
Waiting
Most of the informants involved in this research were struggling during the early phase of the COVID-19 pandemic in Yogyakarta. Their places of production were closed, including arts spaces, music studios, and performance stages. In these risky and uncertain conditions, young creative workers’ option is chiefly to step back from their usual routine while observing the changing rules of the game in the field of cultural production. We call this reflexive practice waiting. Nisa, an early career musician, shared her story with us: Everyone must have been shocked. I just started a regular job in a café, so excited to begin my new job. And, suddenly! The opportunity was gone! Just like that! From March to May, I literally did nothing. I was just in my room all the time, trying to live. You know, anak kos. Luckily, I still have some saving that can make me survive (only) for three months. I couldn’t go anywhere though and had to survive anyway. It was before Ramadhan that I was in a total micro lockdown. I was so terrified and I couldn’t do anything. I only went out to buy groceries. Life felt very difficult.
(Interview with Nisa – a musician, 14 October 2020)
Similarly, Sekar – a young dancer and a mother of two children – took the crisis as a decisive moment. Instead of keeping her dance laboratory running, she chose to use the moment to spend more time with family. She recalls that: This is the time to rest, there is no need to force myself. I could not even imagine if I was being hard to myself [. . .] my husband also supported me to step back a little bit and more focused on the family, and not being so ambitious.
(Interview with Sekar, 3 October 2020)
A similar story with rather different details was told by Meyda. She is a young mother who has been running a small puppet theatre company with her husband who is also a DIY puppet theatre director. In the early phase of the pandemic, she kept her job as a programme manager in a training-sector start-up company to survive the crisis. However, she found it difficult to navigate between doing the job, looking after her child, and thinking about creative arts production during this time. She eventually gave up her job after 2 months of juggling, as she pointed out: It was too much and because all of the meetings were online, everyone did not have a sense of time. It stressed me out [. . .] and it is pandemic, so I asked myself if it is worth it? I felt I was not personally achieving any improvements, so I decided to stop and focus on my child.
Theoretically, these three young creative workers experienced hysteresis (Bourdieu, 2000) in times of rapid societal transformation, crisis and disruption (see Heaphy, 2007; Lupton and Lewis, 2021). It is a complex relationship of a particular situation ‘when a field undergoes a major crisis and its regularity is profoundly changed’ (Bourdieu, 2000, p. 160). Moreover, Bourdieu (2000) explains that ‘the perfect coincidence between structure and habitus is increasingly lost, and briefly alludes to the large-scale social processes involved in this transformation’ (p. 276). It can also be argued that pandemic becomes a precondition that opens up a possible structural change which produces contestation between existing doxa and indefinite forms of heterodoxa in the field of cultural production (see Bourdieu, 1990; Bourdieu, 1993). Thus, for young creative workers, the loss of a major production site as a main playground could mean an extreme game changing situation. The social restrictions have primarily shifted the labour process among the arts and creative production. Young creative workers have excessively lost their field of production to convert their on-hand stocks of capital and exchange their labour, as described by Rizki – an early career musician and audio engineer: It was very impactful. We could not even talk about job risk, as we did not have job anymore, you know? Every gig was cancelled. My engineering job was about the same. I was scheduled to perform in the first ever Soul Music Festival, but it never happened.
(FGD with Rizki – a musician and audio engineer, 17 October 2020)
Galuh – an early career dancer – shared the same story. She is a member of a dance group that performed regularly in the Ramayana Show at the Prambanan Temple. The COVID-19 pandemic made it impossible for the show to remain on schedule. As a consequence, Galuh lost her dancing activities, as she recalls: I lost all the performances, especially the Ramayana Show. It was usually scheduled in the first three months of the year –January, February and March. We did not have the show in March because of the pandemic. The temple called off the event, and that was also the beginning of my no-performances day.
(Interview with Galuh, 24 October 2020)
Galuh then went back home to South Sumatera to wait until the show is back in business. Further, Pambo – a puppeteer in the Papermoon Puppet Theatre – shared a similar story, he said that all the projects were either cancelled or postponed, so basically, we have not yet written any new story [. . .] we cut all the production expenses.
(Interview with Pambo, 22 October 2020)
According to Sweetman (2003), habitus allows one to respond to the current state of play, while simultaneously limiting one’s responses, and as habitus operates in relation to field, it also ensures that removal from the field – or entry into a new game – will generate a different set of responses dependent upon one’s ‘feel’ for the game with which one is now confronted (p. 534). Thus, in over 3 months of disruption and crisis, young creative workers were faced with unpredictable conditions both in everyday life and in the cultural field. The irrelevance of past practices has put them in hysteresis that left them with limited options. With uncertainty playing as the background yet an active response being necessary, they were forced to step back and wait while observing and navigating through the available possibilities. Therefore, it can be argued that habitus at the individual level must respond to disruption and crisis in the existing field, however, such an active response needs time. Young creative workers are reflexively making calculations based upon available yet unsettled individual dispositions and on-hand stocks of capital. Nevertheless, this does not guarantee they will be able to adapt in the present and in the future. In the next phase, we elaborate on a second mode of reflexive practice, namely doing something, employed by young creative workers who start to make series of attempts to do something.
Doing something
I turned into something else; I have started a new business with my friend at home; raising goats. It is better if I try to do something than just waiting. That is kind of our initiative, we also make use of village by planting papaya. We also use that piece of land to keep our goat feed.
Eko, Musician
The above excerpt reflects how young creative workers have had to begin rapidly responding to the uncertain situation. Based on our interviews and FGD sessions, we found two modes of reflexive response to find another available alternative. First, a response that is available outside of their cultural field. Second, a response that is available inside of their field. Both responses reflect the practical manifestation of reflexivity as an embodied form of cultural capital and availability of relevant on-hand stocks of capital at an individual level embedded in the overlapping domains of ‘transition’ and ‘culture’ (Furlong et al., 2011; Woodman and Bennett, 2015) under conditions of crisis and disruption. For Eko – who took the first response category – doing something regardless of the activity is more important since he is married and has a family to look after. Besides, Eko had tried to survive with a small number of gigs in the beginning but as most cafés gradually stopped their operation, he was left with very limited options. Thus, it is arguable that Eko’s reflexive practice is formed by a practical relationship to the possibilities available in a given social environment (Farrugia, 2013: 296).
A similar route has also been taken by Adrian. He is a professional session musician who usually played gigs with orchestral groups and famous bands in Jakarta, although his family is in Yogyakarta. He preferred to be with his family rather than waiting for calls about projects in Jakarta, since all of his past contracts were cancelled. Faced with limited networks in Yogyakarta, he tried to generate income through entrepreneurial activity outside of the music scene, as he described: I sell frozen food now and also do something with my friends who run their own business. Some of them produce masks, the others make food. We are just like: “Ok, what do you need from me? What can you do to help me?” That kind of relationship.
(Interview with Adrian – a Session Musician, 7 October 2020)
Agnes – an independent theatre director – had a similar response. She had been managing her clothing business even before the pandemic and it has been her main income generating strategy while maintaining her artistic activity, as she recalls satirically: My clothing business has provided me food on the table for some years or so. It is way more promising than what I have got from arts [laughing].
(Interview with Agnes, 27 October 2020)
Even so, she tried to remain in the art field during the early phase of the pandemic by creating a collective form of online distribution for recorded pieces of theatrical performance.
According to Hardy (2008), when hysteresis happens, new possibilities are invented by modified field structure. During the pandemic, the new structure of the fields disruptively shifted into the digital arena. In this case, reflexive practice was also adopted by other young creative workers who were trying to engage with digital-online platforms by seeking available alternatives. Sekar managed a collaborative production with UNESCO to create a series of dance tutorial videos and documentation of her sanggar, as she told us: with UNESCO and a collective project named “Kita Muda Creative” [We Are Creative-Youth] we produced a streaming dance video for the public. We also created an online class and tutorial video as this sanggar has slowly back to business.
(Interview with Sekar – a dancer, 3 October 2020)
As a young creative worker who runs her own place, she has shown reflexive practice in taking an alternative opportunity that is available inside the field, without doing something outside of the dance scene. Presumably, having been born and raised in the family of a reputable national pantomime artist and painter helped her embody the disposition of art as a way of life throughout her life trajectory. This particular reflexive capacity to grasp available opportunity does not necessarily coincide with her relatively dominant position in the field, as shown by Nisa. Although she is in the relative position of an early career musician, she was able to convert her durable and strategic social capital (see Sutopo et al., 2017) to take a job offered by her fellow musicians in an existing digital based collective in the music field, as she told us: Fortunately, one of my partners offered me a job as a content creator for his talent management. In this organisation, I am both an employee as well as a talent member whose record will be distributed and managed.
(Interview with Nisa – a musician, 14 October 2020)
A similar opportunity was also available to Galuh as an early career dancer. She had been teaching before the pandemic, and she kept teaching with a few online adjustments, as she described: Fortunately, I have been part time teaching in this school, and it starts to use virtual platforms such as Zoom for the class.
(Interview with Galuh, 24 October 2020)
Even though they have managed to take job opportunities inside the field, both Sekar and Galuh continue to struggle when asked about the involvement of digital platforms for their individual display medium. Accordingly, social agents are actively responding to situations of crisis or sudden change, however, they often have difficulty in holding together the dispositions associated with the different stages of the given field, and adjusting to the newly established order (Bourdieu, 2000).
The digital field seems to be part of the new rules of the game within the field of cultural production (Bourdieu, 1993) as the hysteresis caused by the COVID-19 pandemic has brought about changing requirements to survive in the field. Further, social agents are compelled to be reflexive in order to negotiate fields that are breaking apart and reforming in ways that are unfamiliar (Farrugia, 2013: 882) in the domains of ‘transition’ and ‘culture’ (Woodman and Bennett, 2015). In this case, the linear assumption of unified and coherent single habitus no longer exists. Thus, the narratives above show how young creative workers are reflexively adapting yet experience continuous habitus disruption at an individual level, since for a time at least, field struggles are taking place in the context of an unknown future (Hardy, 2008: 148). The contingent nature of pandemic has exaggerated the disrupted yet unsettled embodied artistic dispositions of young creative workers. Apparently, doing something as a mode shows reflexive practice as agile responses to multi-layered distortion yet plural (Lahire, 2003) and conflicting structural logics in the time of the pandemic. In the next section, we will explore the strategy employed by young creative workers to re-learn the incoming rules of the game, dominated by the presence of digital media.
Re-learning
I have my own YouTube Channel, but it is not yet ready to monetise
and also, my record has not generated significant income yet.
Nisa, Musician
As stated by Nisa, not all young creative workers are individually and generationally equipped with sufficient and relevant forms of capital to embark on the digital platform as a space of display and distribution. Some of them have certainly tried to experiment with digital technology, including Nisa herself, as she told us that I’ve seen that my friends are quite clever dealing with the digital. They have massive followers and once they go streaming, they will put their bank number in the display. So, audience can transfer some money (nyawer). It’s like, there are many alternatives available. People are getting more creative. That one is just a simple example. But it is of course it’s not sustainable and you can’t gain significant income from that.
(Interview with Nisa – a musician, 14 October 2020)
Such a short-term strategy is not suitable for the platform as it requires a solid understanding of online behaviour, the working of various platforms, and engagement with algorithmic-based creative distribution (Arriagada and Concha, 2020; Ignatow and Robinson, 2017). This particular requirement needs to be reflexively re-mapped and re-learned by young creative workers in order to survive in the digital arena. Pambo shared his collaborative learning experience with his puppet group in dealing with the new rules of the game. He said that We have to be really smart in maintaining our audiences, we always put our updated info through our channel [. . .] in fact, we always respond and answer all of the comments and questions from our audiences and, regularly, we discuss what our audiences have said both in their own channel and in our own channel, this helps us to explore the closest topic that relate to our audiences.
He continued discussing the technical aspects of the medium as explained below: Talking about digital performance, we have to touch on its strength and weaknesses. We understand that the mediated show is a whole different world. It detaches audiences with the performers yet allow them to enjoy every detailed visual material captured by camera. Because of that, we have to offer a detailed depiction of our performance, which is off sight during live performance. Once we comprehend those visual aspects, we can then shoot tiny detail that audience can see during live performance.
(Interview with Pambo – a puppeteer, 22 October 2020)
Appearing on camera is seen as one of the main important skills that has to be re-learnt by young creative workers, as also experienced by Galuh that: Dancing in front of camera is different. In live performance, I receive a lot of energy from audience and from other dancers as well. The atmosphere is totally different. Now, you have to dance in front of the camera, the “feeling” is not there anymore. I don’t know what’s missing, but something has gone.
(Interview with Galuh – a dancer, 24 October 2020)
Agnes highlighted similar technical and embodiment aspects that need to be understood and re-learned by young creative workers, as she recalled that: It is not about being cool, but to make yourself contextual [. . .] Corona forces us to learn editing; musicians have to be able to produce their own drum sound, play their keyboard, sing their own verse. It got to be there at some point.
(Interview with Agnes, 27 October 2020)
This coincides with the requirement for new skills pointed out by Rizki, as he said that: In the near future, I need to master some new skills, for example in the recording I used to do a mere operating job, while now, I start to learn producing my own audio-marketing and scoring a movie since such jobs do not entail crowd and you can do it on your own.
(FGD with Rizki – a musician and audio engineer, 17 October 2020)
A similar reflexive strategy is also employed by Adrian, who said he wanted to produce [his] own piece without depending upon a project from other group.
(Interview with Adrian, a Session Musician, 7 October 2020)
Hence, it can be argued that their habitus is not only open to regular adjustment, but also perpetually differs in reaction to disrupted field structures and field positions at macro and micro levels.
Although young creative workers have seemingly gravitated into mediated production, some remain in the existing field without moving much of their practice onto digital platforms. This strategy has been made possible as show business has begun to adjust its event management. Some cafés have started back in operation with certain health protocols; other public events such as weddings and gatherings are starting to occur. With adjustments, some young creative workers are able to gradually work in their field, as experienced by Eko where he has to adjust to the situation, I can play a gig with health protocol such as wearing a mask all the time, do the test screening before the show, and bring a small format-band [. . .] also now in the wedding event, the guest can no longer make a request to sing with the band and we mostly play instrumental songs with a small portion of repertoire played with a singer.
(Interview with Eko, Musician, 15 October 2020)
Having been a session player throughout his music career, Eko’s reflexive practice shows a contradiction between being able to adapt to the new rules of the game, while at the same time, struggling to erase his ‘nostalgia’ for reintegration (see Bourdieu, 1998; Friedman, 2016) into pre-pandemic forms of playing live music. Nevertheless, a ‘feel for the game’ predicated on relationships and regularities of the game and field structures as they were in the past, pre-pandemic, is no longer fit for purpose (Graham, 2020: 451).
Based on the narratives of the young creative workers above, it can be argued that, during conditions of multi-level and multi-temporal dynamics of hysteresis (Graham, 2020) characterised by disruption and crisis (Lupton and Willis, 2021), the multiple schemas of embodied dispositions are reworked in order to fit with the disrupted, irreconcilable yet situational new rules of the game. In addition, new rules of the game in the hierarchical and heterogeneous fields of cultural production are rapidly introduced and learned improvisationally, while concomitantly an enduring practice of an established doxa remains running in the field with few, yet significant, contextual adjustments. Thus, the re-learning strategy applied by young creative workers shows two different types of practice. First, young creative workers who observe other possibilities of engaging with new platforms and alternatives which require some previously unlearned skills. In this line of reflexive strategy, they re-map their field and re-learn the embodied cultural capital and other forms of capital required to pursue the available options. Second, young creative workers who try to adjust their enduring practice into a new situation, in particular, the strict health protocols applied during live shows and performances. Both strategies represent how pandemic as a form of contingent crisis and disruption at macro level are managed fragmentedly at an individual level which is oriented towards re-navigating their future creative careers. Neither reflexive strategy guarantees that young creative workers will be able to sustain their future creative careers; yet they still have to do it in order to survive. In sum, three modes of reflexive practice at an individual level represent active responses manifested in the practical and contradictory relationship to the diverse possibilities within the hierarchical and heterogenous cultural fields in the pandemic era characterised by regularity of ruptures.
Conclusion
Based on the data analysis above, we found three types of temporality-based modes of reflexive practice in the time of pandemic in Yogyakarta, Indonesia: ‘waiting’, ‘doing something’, and ‘re-learning’. Under conditions of crisis and disruption, young creative workers have to wait in order to cope with the experiences of hysteresis. Their ‘pre-pandemic habitus’ and relevant existing forms of capital need to be reflexively re-examined and re-calculated to prepare for the unpredictable yet regular ruptures ahead. ‘Doing something’ indicates the interlinkages between temporal conditions, multi-layered structural/cultural forces in the overlapping domains of transition and culture, and continuous disruption of individual habitus which put pressure on young creative workers to do something under undisclosed future conditions. ‘Re-learning’ reveals how young creative workers reworked their multiple embodied dispositions and other forms of capital in a manner that fit with the disrupted, incompatible yet circumstantial new rules of the game. Being able to re-learn does not by any means guarantee the sustainability of their future creative careers; nonetheless, young creative workers still have to undertake it to survive. Theoretically, pandemic as an extraordinary moment has highlighted the complex relationship with reflexivity and habitus as conceptual tools. In particular, we argued that the contingent nature of a pandemic becomes a prerequisite of a ‘chaotic’ mechanism of adaptation among social agents in the hierarchical and heterogenous fields of struggle under conditions of extreme uncertainties. The intertwined relations between reflexivity and habitus are in a state of constant revisions, yet the precarious nature of human being as cumulative labour remains. Thus, this article has made a contribution to the literature on reflexivity and habitus among young creative workers at a time of pandemic.
The authors would like to thank all the young creative workers in Yogyakarta who participated in this project.
Author biographies
Oki Rahadianto Sutopo is Associate Professor of Sociology and Director of Youth Studies Centre at the Faculty of Social and Political Sciences, Universitas Gadjah Mada. His research interests include youth studies, cultural sociology and sociology of knowledge. He has published his work in Sociological Research Online, Journal of Youth Studies, Crime Media Culture, Continuum, Perfect Beat, and Asian Music.
Gregorius Ragil Wibawanto is a lecturer at the Department of Sociology and a researcher with the Youth Studies Centre, Faculty of Social and Political Sciences Universitas Gadjah Mada. He received his BA (Hons) in Sociology from the same department in 2015 and his master’s degree in Asian and Pacific Studies from The Australian National University in 2019. He has published his work in Geoforum, Journal of Applied Youth Studies, Environmental Education Research, Perfect Beat, and Jurnal Studi Pemuda.
Ariane Utomo is Senior Lecturer in Demography and Population Geography at the School of Geography, Earth and Atmospherics Sciences, the University of Melbourne. Ariane’s core research outputs explore how development and social change relates to attitudes to gender roles; transition to adulthood; women’s employment; marriage; fertility and family patterns; and the nature of inequalities and social stratification in contemporary Indonesia. More recently, she has been involved in a series of collaborative research on ageing, health, and migration in Indonesia and in Australia.
Annisa R Beta is currently a lecturer in Cultural Studies at the School of Culture and Communication, the University of Melbourne, Australia. Her research is broadly concerned with youth, new media and political subjectivity in Indonesia and Southeast Asia. She has published her work in New Media & Society, Feminist Media Studies, Continuum, International Communication Gazette, Asiascape: Digital Asia, Inter-Asia Cultural Studies, International Journal of Communication, and Media and Communication. She has also published her writings with the Conversation, South China Morning Post, and the Jakarta Post.
Novi Kurnia is a senior lecturer at the Department of Communication Science, Faculty of Social and Political Sciences, Universitas Gadjah Mada. Her main interest in research and publication: Indonesian cinema, gender and media, digital literacy and Indonesian youth. She won WhatsApp Misinformation and Social Science Research Award on WhatsApp Group and Digital Literacy Among Indonesian Women published in a book in 2020 with similar title. Her latest book both as co-editor and author is about youth and digital literacy empowerment in East Indonesia.
Funding: The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: this work was supported by the Australia-Indonesia Centre (AIC).
==== Refs
References
Adams M (2006) Hybridizing habitus and reflexivity: Towards an understanding of contemporary identity? Sociology 40 (3 ): 511–528.
Arriagada A Concha P (2020) Cultural intermediaries in the making of branded music events: Digital cultural capital in tension. Journal of Cultural Economy 13 (1 ): 42–53.
Australian Broadcasting Corporation (2021) Indonesia’s coronavirus cases hit a record 54,500 in a single day, overtaking India to become epicenter. Available at: https://www.abc.net.au/news/2021-07-15/indonesia-record-daily-covid-infections-india-most-new-cases/100294708
Badan Ekonomi Kreatif (Bekraf) dan Badan Pusat Statistik (BPS) (2016) Data Statistik Ekonomi Kreatif. Jakarta, Indonesia: Kementerian BUMN Indonesia.
Beck U (1992) Risk Society: Towards a New Modernity. London: SAGE.
Betzler DE Loots M Prokůpek L , et al . (2021) COVID-19 and the arts and cultural sectors: Investigating countries’ contextual factors and early policy measures. International Journal of Cultural Policy 27 (6 ): 796–814.
Bodden MH (2007) ‘Tradition’, ‘modernism’, and the struggle for cultural hegemony in Indonesian national art theatre. Indonesia and the Malay World 35 (101 ): 63–91.
Bourdieu P (1986) The forms of capital. In: Richardson J (ed.) Handbook of Theory and Research for the Sociology of Education. New York: Greenwood, pp. 81–93.
Bourdieu P (1990) The Logic of Practice. Stanford, CA: Stanford University Press.
Bourdieu P (1993) The Field of Cultural Production: Essays on Art and Literature. New York: Columbia University Press.
Bourdieu P (1998) The State Nobility: Elite Schools in the Field of Power. Stanford, CA: Stanford University Press.
Bourdieu P (2000) Pascalian Meditations. Stanford, CA: Stanford University Press.
Bourdieu P Wacquant L (1992) An Invitation to Reflexive Sociology. Chicago: The University of Chicago Press.
Brata AG (2010) Vulnerability of urban informal sector: Street vendors in Yogyakarta, Indonesia. Theoretical and Empirical Researches in Urban Management 5 (5 ): 47–58.
Coffey J Farrugia D (2014) Unpacking the black box: The problem of agency in the sociology of youth. Journal of Youth Studies 17 (4 ): 461–474.
Comunian R England L (2020) Creative and cultural work without filters: Covid-19 and exposed precarity in the creative economy. Cultural Trends 29 (2 ): 112–128.
Connell R (2020) Covid-19/sociology. Journal of Sociology 56 (4 ): 741–751.
Dartanto TJF Rezki CH Siregar H , et al . (2015) Expanding Universal Health Coverage in the Presence of Informality in Indonesia: Challenges and Policy Implications (No. 201511). Jakarta, Indonesia: LPEM, University of Indonesia.
Fahmi FZ McCann P Koster S (2017) Creative economy policy in developing countries: The case of Indonesia. Urban Studies 54 (6 ): 1367–1384.
Fahmi FZ Koster S van Dijk J (2016) The location of creative industries in a developing country: The case of Indonesia. Cities 59 : 66–79.
Farrugia D (2013) The reflexive subject: Towards a theory of reflexivity as practical intelligibility. Current Sociology 61 (3 ): 283–300.
Farrugia D (2015) Addressing the problem of reflexivity in theories of reflexive modernization: Subjectivity and structural contradiction. Journal of Sociology 51 (4 ): 872–886.
Flew T Kirkwood K (2021) The impact of COVID-19 on cultural tourism: Art, culture and communication in four regional sites of Queensland, Australia. Media International Australia 178 (1 ): 16–20.
Flore J Hendry NA Gaylor A (2021) Creative arts workers during the Covid-19 pandemic: Social imaginaries in lockdown. Journal of Sociology. Epub ahead of print 2 August. DOI: 10.1177/14407833211036757.
Florida R Seman M (2020) Measuring COVID-19’s devastating impact on America’s creative economy. Available at: https://www.libarts.colostate.edu/redi/wp-content/uploads/sites/50/2020/08/Florida_Seman_COVID_Brookings.pdf
Friedman S (2016) Habitus clive and the emotional imprint of social mobility. Sociological Review 64 : 129–147.
Furlong A Woodman D Wyn J (2011) Changing times, changing perspectives: Reconciling transition and cultural perspectives on youth and young adulthood. Journal of Sociology 47 (4 ): 355–370.
Giddens A (1991) Modernity and Self-Identity. Cambridge: Polity Press.
Graham H (2020) Hysteresis and the sociological perspective in a time of crisis. Acta Sociologica 63 (4 ): 450–452.
Hardy C (2008) Hysteresis. In: Grenfell M (ed.) Pierre Bourdieu: Key Concepts. Durham, NC: Acumen, pp. 131–150.
Heaphy B (2007) Late Modernity and Social Change: Reconstructing Social and Personal Life. London: Routledge.
Hilgers M (2009) Habitus, freedom and reflexivity. Theory & Psychology 19 (6 ): 728–755.
Howard FA Bennett B Green P , et al . (2021) It’s turned me from a professional to a bedroom DJ once again: Covid-19 and new forms of inequality for young music makers. Young: Nordic Journal of Youth Research 29 (4 ): 417–432.
Ignatow G Robinson L (2017) Pierre Bourdieu: Theorizing the digital. Information, Communication & Society 20 (7 ): 950–966.
Irawanto B (2010) Contemporary Indonesian independent documentaries in the Yogyakarta documentary film festival: Notes from the Juror’s seat. Asian Cinema 21 (2 ): 150–162.
Izzati FFRS Larasati BKC Laksana R , et al . (2021) Pekerja Industri Kreatif Indonesia: Flexploitation, Kerentanan Dan Sulitnya Berserikat. Jakarta, Indonesia: Sindikasi x FNV Mondiaal.
Joffe A (2021) Covid-19 and the African cultural economy: An opportunity to reimagine and reinvigorate? Cultural Trends 30 (1 ): 28–39.
Kelly P (2016) Youth studies and the problem of structure and agency: Foucault vs Marx, Tait vs Sercombe, Beck vs Bourdieu, Woodman vs Threadgold vs Roberts. In: Baker S Robards B Buttigieg B (eds) Youth Cultures and Subcultures: Australian Perspectives. London: Routledge, pp. 21–30.
Lahire B (2003) From the Habitus to an individual heritage of dispositions: Towards a sociology at the level of Individual. Poetics 31 (5–6 ): 329–355.
Lupton D Willis K (eds) (2021) The Covid-19 Crisis: Social Perspective. London: Routledge.
Lupton D Lewis S (2021) The day everything changed: Australians’ Covid-19 risk narratives. Journal of Risk Research. Epub ahead of print 29 July. DOI: 10.1080/13669877.2021.1958045.
McNay L (1999) Gender, habitus and the field: Pierre Bourdieu and the limits of reflexivity. Theory, Culture & Society 16 (1 ): 95–117.
Matthewman S Huppatz K (2020) A sociology of Covid-19. Journal of Sociology 56 (4 ): 675–683.
Mouzelis N (2007) Habitus and reflexivity: Restructuring Bourdieu’s theory of practice. Sociological Research Online 12 (6 ): 1–6.
Rindrasih E Witte P (2021) Reinventing the post-disaster cultural landscape of heritage tourism in Kotagede, Yogyakarta, Indonesia. Journal of Heritage Tourism 16 (2 ): 136–150.
Robbins D (2000) Bourdieu and Culture. London: SAGE.
Sedana IN Foley K (2020) Report: Indonesian Ramayana festival at Prambanan (2012). Asian Theatre Journal 37 (1 ): 228–245.
Setijadi C (2021) The pandemic as political opportunity: Jokowi’s Indonesia in the time of Covid. Bulletin of Indonesian Economic Studies 57 (3 ): 297–320.
Spiro NR Perkins S Kaye U , et al . (2021) The effects of COVID-19 lockdown 1.0 on working patterns, income, and wellbeing among performing arts professionals in the United Kingdom (April–June 2020). Frontiers in Psychology 11 : 4105.
Suharyanto II van Aalst I van Liempt A , et al . (2021) More than Jedug-Jedug: Dynamics of discontent with tourist activity in Prawirotaman, Yogyakarta. Tourism Geographies. Epub ahead of print 15 January. DOI: 10.1080/14616688.2020.1861080.
Sutopo OR (2019) Learning by doing: Young Indonesian musicians, capital and night life. In: Stahl G Botta G (eds) Nocturnes: Popular Music and the Night. London: Palgrave Macmillan, pp. 79–94.
Sutopo OR Wibawanto GR Lukisworo AA (2020) Resist or perish! Understanding the mode of resistance among young DIY Indonesian musicians. Perfect Beat 20 (2 ): 116–133.
Sutopo OR Threadgold S Nilan P (2017) Young Indonesian musicians, strategic social capital, reflexivity and timing. Sociological Research Online 22 (3 ): 186–203.
Sweetman P (2003) Twenty-first century dis-ease? Habitual reflexivity or the reflexive habitus. The Sociological Review 51 (4 ): 528–549.
Threadgold S (2011) Should I pitch my tent in the middle ground? On ‘middling’ tendency, beck and inequality in youth sociology. Journal of Youth Studies 14 (4 ): 381–393.
Threadgold S Nilan P (2009) Reflexivity of contemporary youth, risk and cultural capital. Current Sociology 57 (1 ): 47–68.
Tsui D (2015) A grassroots perspective on Yogyakarta’s art world. Journal of Southeast Asian Studies 46 (3 ): 537–545.
Wacquant L (2014) Putting habitus in its place: Rejoinder to the symposium. Body & Society 20 (2 ): 118–139.
Wicaksono P (2019) Faktor Penyebab Pertumbuhan Ekonomi Kreatif di Indonesia Tinggi. Tempo. Available at: https://bisnis.tempo.co/read/1221592/faktor-penyebab-pertumbuhan-industri-kreatif-di-yogya-tinggi
Woodman D (2009) The mysterious case of the pervasive choice biography: Ulrich beck, structure/agency and the middling state of theory in the sociology of youth. Journal of Youth Studies 12 (3 ): 243–256.
Woodman D Bennett A (2015) Cultures, transitions and generations: The case for a new youth studies. In: Woodman D Bennett A (eds) Youth Cultures, Transitions and Generations: Bridging the Gap in Youth Research. London: Palgrave Macmillan, pp. 1–15.
Zinn JO (2020) A monstrous threat: How state exception turns into a new normal. Journal of Risk Research 23 (7–8 ): 1083–1091.
| 0 | PMC9703019 | NO-CC CODE | 2022-11-29 23:21:05 | no | Sociol Res Online. 2022 Nov 25;:13607804221115433 | utf-8 | Sociol Res Online | 2,022 | 10.1177/13607804221115433 | oa_other |
==== Front
Omega (Westport)
Omega (Westport)
spome
OME
Omega
0030-2228
1541-3764
SAGE Publications Sage CA: Los Angeles, CA
36423236
10.1177_00302228221141937
10.1177/00302228221141937
Original Manuscript
Growing in Suffering: The Curvilinear Relationship Between Prolonged Grief and Post-traumatic Growth of Recently Bereaved Individual During the COVID-19 Pandemic
Qian Wenli 1
Tang Renzhihui 1
Jiao Keyuan 2
Xu Xin 1
Zou Xinyan 1
https://orcid.org/0000-0001-7331-7525
Wang Jianping 1
1 Beijing Key Laboratory of Applied Experimental Psychology, National Demonstration Center for Experimental Psychology Education (Beijing Normal University), Faculty of Psychology, 47836 Beijing Normal University , China
2 Department of Social Work and Social Administration, 25809 The University of Hong Kong , Hong Kong SAR, China
Jianping Wang, Faculty of Psychology, Beijing Normal University, Xinjiekou 19, Haidian District, Beijing, Beijing 100875, China. Email: [email protected]
24 11 2022
24 11 2022
00302228221141937© The Author(s) 2022
2022
SAGE Publications
This article is made available via the PMC Open Access Subset for unrestricted re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the COVID-19 pandemic or until permissions are revoked in writing. Upon expiration of these permissions, PMC is granted a perpetual license to make this article available via PMC and Europe PMC, consistent with existing copyright protections.
The outbreak of the COVID-19 pandemic brought new challenges to mourning and growth of bereaved. The purpose is to explore the relationship between the prolonged grief (PG) symptoms and the post-traumatic growth (PTG) of recently bereaved people during the COVID-19 period, and the mediating role of meaning making. 305 participants were recruited to complete the Posttraumatic Growth Inventory, Inventory of Complicated Grief, and Integration of Stressful Life Experiences Scale. Hierarchical multiple regression analyses and Medcurve in SPSS were adapted to test the hypotheses. The results revealed that there was a curvilinear relationship between PG and PTG and meaning making had a completely mediating effect on this relationship. Different intervention goals - whether alleviating distress or cultivating growth – should be set up according to the level of grief for recently bereaved individuals during COVID-19. More attention should be paid to the role of meaning making in the future clinical practice.
Prolonged grief
posttraumatic growth
bereaved
meaning making
COVID-19
National Social Science Fund of China https://doi.org/10.13039/501100012456 16ZDA233 edited-statecorrected-proof
typesetterts10
==== Body
pmcIntroduction
Bereavement During the COVID-19 Pandemic
The novel coronavirus disease (COVID-19) has rapidly spread across the entire globe and undeniably caused severe consequences. As of April in 2022, there were over 510 million confirmed cases and over 6 million deaths of individual worldwide (World Health Organization, 2022a). The background in which bereavement occurs is special. Many have died alone because of the requirements for social distancing. Moreover, a larger number of the deaths were caused by other COVID-19-related challenges, such as avoiding visiting hospitals or postponing treatments of their other life-threatening diseases (Eisma et al., 2020). At the beginning of the epidemic, the COVID-19 was explosive and traumatic, which seriously affected psychological health of the bereaved, especially the people in China from the end of 2019 to the 2020. Due to the social distance policies, the governments banned large scale funerals, some cemeteries were closed to the public, people might be unable to say goodbye to their beloved ones in to receive warmth or affection from others (Cardoso et al., 2020; Diolaiuti et al., 2021; Gomez-Salgado et al., 2020; Ingravallo, 2020; Marazziti et al., 2020).
Researchers believe that coping with loss and grief during the COVID-19 pandemic, particularly in the situation of individual susceptibility and related challenges, may influence the grief process and increase the risk of Prolonged Grief Disorder (PGD) (Eisma et al., 2020; Gesi et al., 2020).
Positive and Negative Psychological After Bereavement
When confronted with the death of a loved one, people commonly experience grief and may return to their normal level of functioning after a period of bereavement (Bonanno & Kaltman, 2001). However, it is estimated that one in 10 bereaved of natural causes and one in two of unnatural causes have reported severe and disabling grief reactions that deserve clinical attention (Djelantik et al., 2020; Lundorff et al., 2017). Some of them cannot be alleviated and may develop pathological and maladaptive consequences, such as PGD defined by ICD-11 (World Health Organization, 2022b) and the text-revision of 5th Diagnostical and Statistical Manual of Mental Disorders (DSM-5-TR; APA, 2022). In DSM-TR, the PGD is characterized by yearning for or persistent preoccupation with the deceased, accompanied by intense emotional symptoms which persist for at least 12 months (6 months for children) and lead to impairment of individual functioning. Considering of inability to say “goodbye” to the bereaved and conduct funeral ceremonies, lack of expectation for the death, lower level of support, multiple losses (due to the disease spreading), and continuous realistic stress, the COVID-19 pandemic may increase the risk of PGD (Eisma et al., 2020; Stroebe & Schut, 2021; Wallace et al., 2020; J. Xu et al., 2020a). Previous research also highlighted the prolonged grief (PG) symptoms of recently bereaved individuals as a strong predictor of future pathological grief (Boelen & Lenferink, 2020; Eisma et al., 2021; Goldsmith et al., 2008).
In the meantime, struggling with stressful events, such as bereavement, could also lead to positive psychological changes named posttraumatic growth (PTG) (Eisma et al., 2019; Levi-Belz, 2020; Salloum et al., 2019; Xu et al., 2015; X. Xu et al., 2020b). The PTG includes feeling stronger, feeling closer to others, experiencing new possibilities, more appreciation of life, and spiritual change (Tedeschi & Calhoun, 2004). For example, losing a loved one could increase the compassion for other bereaved people (Eisma et al., 2019). Additionally, as COVID-19 is also considered to be a type of mass trauma (Xie & Kim, 2022), people who lost their relatives within 1 year during the COVID-19 also showed both PG symptoms and PTG (Chen & Tang, 2021). However, the relationship of PG and PTG for recently bereaved during this pandemic is unidentified.
Prolonged grief and Post-Traumatic Growth
The PG symptoms and PTG sound like two sides of a coin, which brought a “less-is-better” view at first: Less grief enhances growth (Engelkemeyer & Marwit, 2008; Feigelman et al., 2009). However, some researchers challenged this view. For example, positive association (Xu et al., 2015) or no significant correlation (Salloum et al., 2019) between the two variables were found. According to the model of growth in grief, a high level of distress means that the assumptive world beliefs were challenged by the bereavement, which promoted individuals to find ways to manage pain, understand the death, reassess the event and finally accept the changed world or gain growth, while people with a very low level of grief adjusted to the death directly because of the beliefs were not being challenged (Calhoun et al., 2010). Besides, too much grief may be too stressful to develop growth (Butler et al., 2005). Drawing from these insights, some researchers seek to provide a more balanced point of view: Moderate-is-better, which means that there was a curvilinear (inverted U-shape) relationship between PG and PTG (Eisma et al., 2019; Levi-Belz, 2020; Yilmaz & Zara, 2016).
The meaning making theory may contribute to understanding the mechanisms of this curvilinear relationship between PG and PTG (Park, 2010). People constructed their global meaning in early life, consisting of beliefs, goals, and subjective feelings. With the occurrence of a potentially stressful event (such as loss), people began the process including assigning meaning to the event (or appraised meaning), determining discrepancies between appraised and global meaning (feeling distress if it was discrepant), meaning making, meanings made (such as perception of growth or positive life change), and adjustment to the event. In other words, PTG, as a commonly accessed type of meaning made, was the product of meaning-making processes. Meaning making (MM) refers to the cognitive processes aimed at understanding and finding significance or benefits of experiences and reflects the degree to which individuals integrated memories of the event into a coherent self-narrative (Holland et al., 2010). Some empirical researches indicate a positive relationship between meaning making and PTG (Boyraz & Efstathiou, 2011; Jin et al., 2014; Williams et al., 2020). Additionally, meaning making is associated with decreased PG symptoms both theoretically (Park, 2010) and empirically (Boyraz & Efstathiou, 2011; Holland et al., 2010; Milman et al., 2019; Pan et al., 2018; Zakarian et al., 2019).
Although prior research has generated a wealth of insights about PG and PTG, this view has yet to be examined among recently bereaved people during the COVID-19, as the relationship between PG and PTG as well as the role of meaning making in this relationship may vary with bereavement characteristics and circumstances (Eisma et al., 2020).
Cross-Culture View of Grief and Growth
Not only do social circumstances influence grief and growth after loss, but the cultural influence is also significant. Chinese people view death and grief from a collective perspective rather than as an individual, while American tend to focus on the individual view (Zhao et al., 2007). For instance, with the ingrained traditional culture of filial piety in China, the death of children means the interruption of continuity of the family line and the loss of future hope for Chinese parents (Zhang & Jia, 2018). Research of Chinese Shidu parents who lost their only child indicated these bereaved parents were more likely to endorse negative beliefs about the self, life, world, and the future (He et al., 2014) and they held culture-related grief beliefs containing filial piety belief, destiny belief and perceived stigma (Shi et al., 2019). When comforting the bereaved person, it is common to hear American people say words about passing and accepting grief like “you will get over this in time” or “he or she is in a better place”. However, in Chinese culture, a common saying heard is “save the tears and follow the flow (jie ai shun bian)”, which implies an avoidance tendency in it (Chow et al., 2007). Cross-culture studies about bereavement showed that Chinese people were discouraged to express grief openly and their acceptance of death still mix with shock and pain (Moats, 2011). They tended to search for meaning and avoiding thinking about the deceased (Pressman & Bonanno, 2016). Given the culture-related cognitions and emotional expression about death, it is not surprising that Chinese people do have special characteristic in grieving. As a result, paying specific attention to the Chinese bereaved is necessary.
The Present Study
The aims of the current study were twofold. Firstly, to examine the relationship of PG and PTG among Chinese recently bereaved people during the COVID-19. We hypothesize that PG and PTG have a curvilinear relationship and the moderate levels of grief yield higher levels of PTG. Secondly, we set out to clarify the role of meaning making in the relationship between PG and PTG. In line with this, we hypothesized that meaning making mediated the relationship between PG and PTG.
Methods
Participants and Procedure
As a part of the project named “Psychological assistance for bereaved persons during COVID-19 pandemic in mainland China”, the present study was an online-research which collected data from May 2020 to January 2021 via internet media, such as WeChat and websites related to bereavement. The details of the project were introduced in another essay (Tang et al., 2021).
A representative sample of 339 bereaved persons who lost their loved one during the COVID-19 epidemic period in mainland China was adopted in our current study. After excluding 8 adolescent data (age <18) and 26 incomplete questionnaires, 305 questionnaires were included in the data analysis finally. The participants contained 219 females (71.80%) and 86 males (28.20%), and their average age was 34.08 years (SD = 11.82). Although most of the individuals had no religious beliefs, it was worth noting that two bereaved people became interested in religious after their relatives died. The demographic information is presented in Table 1.Table 1. Demographic of the sample (N = 305).
Variables n (%)/M (SD) Variables n (%)
Gender Monthly Income (CNY)
Female 219 (71.80%) 0 96 (31.48%)
Age (range) 0–3000 34 (11.15%)
18–79 34.08 (11.82) 3000–5000 64 (20.98%)
Residence 5000–10,000 69 (22.62%)
Rural 40 (13.11%) 10,000–20,000 19 (6.23%)
Town 41 (13.44%) >20,000 9 (2.95%)
Urban 224 (73.44%) Missing 14 (4.59%)
Education level Subjective economic condition
Junior high school 14 (4.59%) Wealthy 3 (0.98%)
Senior high school/technical secondary school 25 (8.20%) Moderately wealthy 68 (22.30%)
College/junior college 197 (64.59%) General 185 (60.66%)
Master 60 (19.67%) Poor 49 (16.07%)
Doctor 9 (2.95%) Religion
Marriage status None 263 (86.23%)
Unmarried 142 (46.56%) Buddhism 23 (7.54%)
Married 117 (38.36%) Christianity 15 (4.92%)
Divorced 10 (3.28%) Taoism 1 (0.33%)
Widowed 36 (11.80%) elsea 3 (0.98%)
aelse include: 1. Catholicism; 2. No specific religious belief but being interested in Buddhism after loss.
About half of them lost their parent (n = 132, 43.28%), and most of the rest lost their other important individual (e.g., spouse, sibling, grandparent). All participants lost their important one within 12 months. Besides, 151 (49.51%) participants had not psychological expectation of the death at all. Among the death, 12.31% of them died of COVID-19 and at least 18.04% was violent death (from suicide or accidental death). Although the COVID-19 was not the direct cause of death, two participants reported that their relatives died of untimely treatment or care because of the COVID-19. The loss characteristics of participants are presented in Table 2.Table 2. The Loss-related Characteristics of the sample (N = 305).
Variables n (%)/M (SD) Variables n (%)
Gender of deceased Relationship to deceased
Male 197 (64.59%) Parent 132 (43.28%)
Age of deceased (range) Spouse 41 (13.44%)
0–99 57.31 (22.97) Child 22 (7.21%)
Time since loss (month range) Sister/brother 30 (9.84%)
0–12 3.99 (2.45) Grandparent 50 (16.39%)
Cause of death Other relatives 15 (4.92%)
COVID-19 37 (12.31%) Friends 4 (1.31%)
Acute illness (<1 month) 86 (28.20%) Else 11 (3.61%)
Chronic illness (≥1 month) 99 (32.46%) Relationship with deceased
Accident 34 (11.15%) Very close 171 (56.07%)
Suicide 21 (6.89%) Quite close 96 (31.48%)
Elsea 28 (9.18%) General 30 (9.84%)
Expectation of death Quite distant 7 (2.30%)
Totally no expectation 151 (49.51%) Very distant 1 (0.33%)
Not enough expectation 70 (22.95%)
General 33 (10.82%)
Quite enough expectation 46 (15.08%)
Totally enough expectation 5 (1.64%)
aelse include: 1. Fell ill during the COVID-19 and could not be treated in time, which delayed the disease and led to death; 2. Quarantine due to the COVID-19 and lack of timely care; 3. Abortion; 4. Natural aging; 5. The persistent chronic disease suddenly worsened; 6. Natural disaster.
Measures
Socio-Demographics Information
A brief self-designed questionnaire was adopted to collect participants’ information about themselves (e.g., gender, age, education) and their loss (e.g., cause of death, time since loss).
Posttraumatic Growth Inventory (PTGI)
Posttraumatic growth (PTG) was accessed by Posttraumatic Growth Inventory (Tedeschi & Calhoun, 1996; Wang et al., 2011). PTGI consisted of 21 items and that can be divided into 5 dimensions: relating to others (7 items, e.g., “A sense of closeness with others.”), new possibilities (5 items, e.g., “I developed new interests.”), personal strength (4 items, e.g., “A feeling of self-reliance.”), spiritual change (2 items, e.g., “A better understanding of spiritual matters.”), and appreciation of life (3 items, e.g., “An appreciation for the value of my own life.”). Items were rated on a 6-Liked scale, which described the extent of experiencing the change as a result of this crisis, ranging from 1 (did not at all) to 6 (a very great degree). Total scores ranged from 21 to 126, with higher scores reflecting greater growth. The PTGI displayed good internal consistency in the present study (α = 0.931).
Inventory of Complicated Grief (ICG)
Inventory of Complicated Grief was adapted to measure the severity of participants’ prolonged grief (PG) symptoms (Prigerson et al., 1995). It showed good reliability and validity in previous research conducted in the Chinese sample (Li & Prigerson, 2016). The ICG has 19 items (e.g., “Memories of the person who died upset me.”) and participants were asked to report the frequency of these grief experiences on a 5-point Likert scale (0 = never, 4 = always). The scale also demonstrated good internal consistency (α = 0.935) in the present sample.
Integration of Stressful Life Experiences Scale (ISLES)
The Integration of Stressful Life Experiences Scale measured the extent to which stressful life experiences (such as bereavement) were adaptively integrated into individuals’ broader life narrative (Holland et al., 2010). Five items assessed comprehensibility (e.g., “This event is incomprehensible to me.”), and 11 items related to footing in the world (e.g., “Since this event happened, I don’t know where to go next in my life.”). A five-point scoring system was used (1 = strongly agree and 5 = strongly disagree) for rating. Item 2 was calculated by reverse scoring, and a higher total score reflected a higher level of meaning making of the death. The Chinese version of ISLES was obtained by a back-translation method. The results of confirmatory factor analysis (CFA) were: RMSEA = 0.104, CFI = 0.917, TLI = 0.904, and SRMR = 0.039. Besides, Cronbach’s alpha was 0.945 for ISLES in this sample. As a result, the psychometric properties of ISLES were acceptable in the current study.
Data Analyses
There was three missing data and sequence average was used to replace them. Before analyzing, Harman’s single-factor test was used to determine whether a common method bias existed. All observation variables, including PTG, PG and meaning making, were included into an exploratory factorial analysis (EFA). The results showed that 8 factors eigenvalues were larger than 1 and the maximum factor explanted 33.44% of the total variance, which was less than 40% (Podsakoff et al., 2003). It indicated that there was no significant common method bias in the present study and further analyses could be conducted.
Statistical analyses were performed by SPSS version 25.0. Descriptive statistics were used to introduce demographic information, means and standard deviation of the main variables. Person correlation was carried out to present the relationship between variables. Additionally, series hierarchical multiple regression (HMR) analyses were conducted to test our hypotheses about the curvilinear relationship between PG and PTG. Lastly, Medcurve in SPSS was adapted to test the mediation model (Hayes & Preacher, 2010).
Results
Preliminary Analyses
Descriptive results and bivariate correlations of variables are presented in Table 3. Expectation of death had significantly positive association with PTG (r = 0.16, p < .01), negative association with PG (r = −0.32, p < .001) and positive association with MM (r = 0.33, p < .001). Additionally, PTG was negatively associated with PG (r = −0.13, p < .05) and positively associated with MM (r = 0.19, p < .01). Moreover, there was a significantly negative correlation between PG and MM (r = −0.80, p < .001).Table 3. Descriptive results and bivariate correlations.
Variables M (SD) PTG PG MM
1 PTG 58.86 (20.74) 1
2 PG 37.66 (16.28) −0.13* 1
3 MM 49.77 (15.95) 0.19** −0.80*** 1
4 Gender / −0.08 0.08 −0.20***
5 Age 34.08 (11.82) 0.00 0.07 −0.12*
6 Relationship to deceased / −0.03 −0.27*** 0.26***
7 Relationship with deceased 1.59 (0.79) 0.05 −0.36*** 0.26***
8 Cause of death / 0.04 0.09 −0.04
9 Time since loss 3.99 (2.45) 0.06 −0.05 0.15**
10 Expectation of death 1.96 (1.17) 0.16** −0.32*** 0.33***
Notes. PTG = Posttraumatic growth; PG = Prolonged grief; MM = Meaning making. *p < .05; **p < .01; ***p < .001.
Table 4 presents the results of hierarchical Multiple Regression (HMR). In Model 3, the addition of PG2 shows significant R2 changes (ΔR2= 0.02, p < .01). It indicates that there was a significant curvilinear relationship between PG and PTG in the current study. As a result, it supports our first hypothesis that there was an inverted U-shape model between PG and growth after controlling the influence of covariate variables (Figure 1). As the PG level increases, the PTG gradually increases, and at very high PG levels the PTG begins to decrease.Table 4. Hierarchical Regression for PTG.
Model 1 DV: PTG Model 2 DV: PTG Model 3 DV: PTG Model 4 DV:MM Model 5 DV: PTG Model 6 DV: PTG
Gender –1.14 –0.99 –0.94 –3.22** –0.55 –0.67
Age –0.29 –0.39 –0.41 –1.27 –0.26 –0.34
Relationship to deceased –0.54 -0.92 –0.62 0.96 –0.74 –1.07
Relationship with deceased 0.41 -0.08 0.28 –1.30 0.43 0.27
Cause of death 0.59 0.71 1.08 0.52 1.02 0.86
Time since loss 0.55 0.52 0.48 2.39* 0.20 –0.19
Expectation of death 2.80** 2.19* 2.36* 2.79** 2.01* 1.63
PG –1.46 2.30* –5.28*** 2.81** 1.03
PG2 –2.77** 0.38 –2.83** –0.86
MM 2.02* 3.08**
MM2 –2.67**
R2 0.04 0.04 0.07 0.67 0.08 0.10
∆R2 0.04 0.00 0.02 0.01 0.02
Notes. DV = Dependent variable; PTG = posttraumatic growth; PG = prolonged grief; MM = meaning making. *p < .05; **p < .01; ***p < .001.
Figure 1. Curvilinear relationship of PG and PTG (y = −0.01x2 + 0.76x + 45.43).
Afterward, we tested the indirect curvilinear effect of PG. Model 4 (Table 4) indicated that the PG was significantly related to meaning making (β = −0.79, SE = 0.15, p < .05). Besides, comparing model 6 to model 5 showed that when included the quadratic effect of meaning making (ΔR2 = 0.02, p < .01), the effect of PG2 became nonsignificant. These support the indirect effect hypothesis. That is, the observed curvilinear relationship between meaning making and PTG explained the overall curvilinear indirect effect of PG on PTG. The figure 2 shows the relationship between PTG and meaning making virtually after controlling the influence of PG and covariate variables.Figure 2. Curvilinear relationship of MM and PTG (y = −0.01m2 + 1.62m + 5.59).
Finally, we tested the significance of the indirect curvilinear effect of PG on PTG by using the “Medcurve” bootstrapping procedure (Hayes & Preacher, 2010). The 1000 bootstrap was adopted and the results showed statistically significant as the 95% confidence interval of the indirect effect did not contain zero (−0.37, −0.01). Thus, the indirect curvilinear effect of PG on PTG through meaning making was further supported.
Discussion
The current research examined the relationship between PG and PTG, and the mediating role of meaning making among recently bereaved people during COVID-19. The results of this study provide support for the proposed hypotheses: an inverted U-shape relationship between PG and PTG was found and meaning making played a complete mediation role in this curvilinear relationship.
In this research, we found a negative correlation between expectation of death and grief-related results, which is consist with previous research (Eisma et al., 2019; He et al., 2013), while a positive correlation between expectation and posttraumatic growth, which is inconsistent with previous studies (McClatchey, 2020; Salloum et al., 2019). A possible reason for this inconsistency is that different from the prior studies which defined unexpected death as death caused by accidental reasons (McClatchey, 2020; Salloum et al., 2019), we directly measured the subjective expectation degree of the bereaved. For the bereaved, a high level of expectation may be conducive to making psychological preparation for the subsequent separation before the death. Besides, those with high expectation may be able form a reasonable explanation for the death through meaning making and then gain growth. Furthermore, time after loss did not significantly correlate with PG and PTG, which is not consistent with some previous results (He et al., 2013), but echoes the findings of McClatchey, 2020; Xu et al., 2015, 2020a. Differences in sample sizes and ranges of time since loss may explain the discrepancy.
Consistent with our hypothesis, there was a curvilinear relationship between PG and PTG among recently bereaved people during the COVID-19. Specifically, those who experienced little grief in response to the bereavement gained certain growth; those who experienced moderate levels of grief developed higher levels of PTG; and the other people experienced the event as too emotionally overwhelming to experience much PTG. The result aligns with Eisma et al. (2019), Levi-Belz (2020) and Yilmaz and Zara (2016) and confirms the “moderate-is-better” idea: experiencing a certain level of distress contributes to getting positive life changes, but too much distress will hinder one’s path to gain benefits. Additionally, this study focused on the individuals who lost their loved ones during the social situation of COVID-19 pandemic, expanding our understanding of the relationship between grief and growth in different environments.
A strong negative relationship between PG-symptoms severity and meaning making has been reported in the previous literature and this study (Pan et al., 2018; Rozalski et al., 2016; Zakarian et al., 2019). That is to say, people with more distress may be more difficult to make sense of and integrate the loss event. The results also showed a curvilinear relationship between meaning making and PTG, which was inconsistent with previous research demonstrating a positive linear relationship (Boyraz & Efstathiou, 2011; Jin et al., 2014; Williams et al., 2020). In the present research, although the level of PTG increased with meaning making increasing from the whole view, individuals with moderate meaning making of bereavement events already perceived a relatively high level of growth, while excessive meaning making would bring an inappreciable increase of growth. Given that a cross-culture qualitative investigation indicated that Chinese bereaved participants mixed shock, absurdity, and pain with the acceptance and making sense of the death (Moats, 2011), we could interpret that probably culture-related factors played an important role in the relationship between meaning making and PTG. The results suggested that, despite the importance of meaning making for promoting growth, it is unnecessary to blindly pursue an adequate understanding or gain complete meaning from death for the bereaved.
Furthermore, after including the square of meaning making in the model of PG and PTG, the link of PG2 and PTG became non-significant, suggesting full mediation. This finding fits the theory that distress evoked by stressful events would develop PTG through cognitive processes involving attempts to understand and quest for meaning (Park, 2010). Our mediated model suggested that, when individuals experience a low level of grief, they might use an excessive meaning making process to avoid their own pain. Their PTG could include fantasy and deception in the Janus Face Model of PTG (Zoellner & Maercker, 2006). When the bereavement event brought great pain, individuals with seriously damaged functions were unable to carry out better meaning making for the loss event and cannot perceive much PTG either. The findings have implications for future clinical work: For individuals with moderate grief response, we could help them achieve a high level of growth by promoting their expression and meaning construction of death events. For those individuals with severe grief symptoms, simple meaning making is not enough to alleviate their pain and help them grow and more comprehensive intervention is needed to guide their adaptation through bereavement.
The current study extends existing knowledge about the relationship between different psychological outcomes of recently bereaved people during COVID-19 and the significance of cognitive processes in it. The results have some strengths and clinical implications for the treatment of grief-related distress. It is noteworthy that several studies have examined the effectiveness of interventions for alleviating the distress of the bereaved, such as PG and posttraumatic stress symptoms (Lund et al., 2010; Shear et al., 2005; Waller et al., 2016) and some authors have demonstrated the effectiveness of interventions to cultivate growth after stress event (Bower & Segerstrom, 2004; Dolbier et al., 2010; Hagenaars & van Minnen, 2010). However, based on the finding of the current investigation about curvilinear relationship between PG and PTG, for the recently bereaved individuals during COVID-19, reducing distress may not equally increase their PTG. As a result, it is necessary to choose the appropriate intervention scheme according to the different conditions of the bereaved, rather than adopting a unified method for all people. With the normalization of the COVID-19, we suggest that a key to future clinical interventions for bereavement is to focus on both alleviating suffering and cultivating growth. When the bereaved are in moderate grief, we can reduce their suffering and actively promote their posttraumatic growth simultaneously. And for people with a high degree of grief, the primary task is to reduce the pain and then pay attention to growth after the grief is reduced. Moreover, the significant role of meaning making in this study suggested that interventions which help people find meaning in the grief experience and integrate it into a reconstructed self-narrative (Neimeyer et al., 2010), such as narrative retelling and therapeutic writing, may contribute to promote grief to growth and could be adopted when supporting the bereaved people during COVID-19.
There are some limitations that should be acknowledged in the present research. Firstly, cross-section design excludes causal inferences regarding PG, PTG, and meaning making. According to the Janus-Face model of PTG (Maercker & Zoellner, 2004), there was two components of self-perceived PTG: the functional side and the illusory side. Longitudinal studies may show whether individuals’ growth will reduce or increase the possibility of mental health problems, that is, whether it is functional or illusory. This study is a prospective part of a whole program and longitudinal research and intervention studies testing the relationship are warranted in the future. Secondly, the bereavement period in our samples was limited within 1 year. Because the whole program was launched shortly after the outbreak of the COVID-19 in order to provide psychological help to the bereaved as soon as possible. More kinds of bereaved people during the COVID-19 should be included in future study. Thirdly, due to a wide range of bereavement types (such as loss of child, parent, and spouse) and the unequal number of different types among our participants, we need to be cautious in the generalization of results. Lastly, this study conducted a self-reported measurement and an important direction for future studies is adding other indicators, such as evaluations by families.
Conclusion
In a whole, the present study demonstrates that concurrent associations between PG and PTG is curvilinear following recent bereavement during COVID-19. Moreover, the mediated effect of meaning making was tested, showing that meaning making plays a complete mediation role in the relationship between PG and PTG. This suggests that for recently bereaved individuals during COVID-19, different intervention goals need to be selected according to different situations: For bereaved persons with low grief level, they could recover without professional intervention; for bereaved people with moderate grief symptoms, attention should be given to both alleviating distress and cultivating growth simultaneously; and primary aim of intervention for that with serious grief response is to alleviate symptoms. As an important element of grief intervention, meaning making could play a positive role in healing the pain and promoting growth, which should be paid attention to in future clinical practice.
Author Biographies
Wenli Qian is a PhD student at Faculty of Psychology, Beijing Normal University. Her research focus is Prolonged Grief Disorder, Post-traumatic Growth and Bereavement.
Renzhihui Tang is a PhD student at Faculty of Psychology, Beijing Normal University. Her research focus is Prolonged Grief Disorder and Family Bereavement.
Keyuan Jiao, PhD, graduated from Department of Social Work and Social Administration, The University of Hong Kong. Her research focus is Family Bereavement and Parent–child relationship in Widowed family.
Xin Xu, PhD, graduated from Faculty of Psychology, Beijing Normal University. Her research focus is Prolonged Grief Disorder, Psychological stress in parents who lost their only-one child, and Grief Counseling.
Xinyan Zou, Master Degree, graduated from Faculty of Psychology, Beijing Normal University. Her research focus is Prolonged Grief Disorder, Post-traumatic Growth and Risk factors for grief.
Jianping Wang is a professor in Faculty of Psychology, Beijing Normal University. Her research focus is Clinical Psychology, Psychopathology, Cognitive Behavior Therapy and Prolonged Grief Disorder.
ORCID iD
Jianping Wang https://orcid.org/0000-0001-7331-7525
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding: The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: This work was supported by the [National Social Science Fund of China] under Grant [number 16ZDA233].
==== Refs
References
American Psychiatric Association (2022). Trauma and stressor-related disorders. https://dsm.psychiatryonline.org/doi/full/10.1176/appi.books.9780890425787.x07_Trauma_and_Stressor_Related_Disorders
Boelen P. A. Lenferink L. I. M. (2020). Symptoms of prolonged grief, posttraumatic stress, and depression in recently bereaved people: Symptom profiles, predictive value, and cognitive behavioural correlates. Social Psychiatry and Psychiatric Epidemiology, 55 (6 ), 765–777. 10.1007/s00127-019-01776-w 31535165
Bonanno G. A. Kaltman S. (2001). The varieties of grief experience. Clinical Psychology Review, 21 (5 ), 705–734. 10.1016/S0272-7358(00)00062-3 11434227
Bower J. E. Segerstrom S. C. (2004). Stress management, finding benefit, and immune function: Positive mechanisms for intervention effects on physiology. Journal of Psychosomatic Research, 56 (1 ), 9–11. 10.1016/s0022-3999(03)00120-x 14987958
Boyraz G. Efstathiou N. (2011). Self-focused attention, meaning, and posttraumatic growth: The mediating role of positive and negative affect for bereaved women. Journal of Loss and Trauma, 16 (1 ), 13–32. 10.1080/15325024.2010.507658
Butler L. D. Blasey C. M. Garlan R. W. McCaslin S. E. Azarow J. Chen X. Desjardins J. C. DiMiceli S. Seagraves D. A. Hastings T. A. Kraemer H. C. Spiegel D. (2005). Posttraumatic growth following the terrorist attacks of september 11, 2001: Cognitive, coping, and trauma symptom predictors in an internet convenience sample. Traumatology, 11 (4 ), 247–267. 10.1177/153476560501100405
Calhoun L. G. Tedeschi R. G. Cann A. Hanks E. A. (2010). Positive outcomes following bereavement: Paths to posttraumatic growth. Psychologica Belgica, 50 (1-2 ), 125–143. 10.5334/pb-50-1-2-125
Cardoso E. A. O. Silva B. Santos J. H. D. Loterio L. D. S. Accoroni A. G. Santos M. A. D. (2020). The effect of suppressing funeral rituals during the COVID-19 pandemic on bereaved families. Revista Latino-Americana de Enfermagem , 28 , Article e3361. 10.1590/1518-8345.4519.3361
Chen C. Tang S. (2021). Profiles of grief, post-traumatic stress, and post-traumatic growth among people bereaved due to COVID-19. European Journal of Psychotraumatology, 12 (1 ), 1947563. 10.1080/20008198.2021.1947563 34367528
Chow A. Y. Chan C. L. Ho S. M. (2007). Social sharing of bereavement experience by Chinese bereaved persons in Hong Kong. Death Studies, 31 (7 ), 601–618. 10.1080/07481180701405105 17847573
Diolaiuti F. Marazziti D. Beatino M. F. Mucci F. Pozza A. (2021). Impact and consequences of COVID-19 pandemic on complicated grief and persistent complex bereavement disorder. Psychiatry Research, 300 (2 ), 113916. 10.1016/j.psychres.2021.113916 33836468
Djelantik A. A. A. Manik J. Smid G. E. Mroz A. Kleber R. J. Boelen P. A. (2020). The prevalence of prolonged grief disorder in bereaved individuals following unnatural losses: Systematic review and meta regression analysis. Journal of Affective Disorders, 265 , 146–156. 10.1016/j.jad.2020.01.034 32090736
Dolbier C. L. Jaggars S. S. Steinhardt M. A. (2010). Stress-related growth: Pre-intervention correlates and change following a resilience intervention. Stress and Health, 26 (2 ), 135–147. 10.1002/smi.1275
Eisma M. C. Boelen P. A. Lenferink L. I. M. (2020). Prolonged grief disorder following the Coronavirus (COVID-19) pandemic. Psychiatry Research, 288 , 113031. 10.1016/j.psychres.2020.113031 32360895
Eisma M. C. Lenferink L. I. M. Stroebe M. S. Boelen P. A. Schut H. A. W. (2019). No pain, no gain: Cross-lagged analyses of posttraumatic growth and anxiety, depression, posttraumatic stress and prolonged grief symptoms after loss. Anxiety Stress Coping, 32 (3 ), 231–243. 10.1080/10615806.2019.1584293 30793949
Eisma M. C. Tamminga A. Smid G. E. Boelen P. A. (2021). Acute grief after deaths due to COVID-19, natural causes and unnatural causes: An empirical comparison. Journal of Affective Disorders, 278 , 54–56. 10.1016/j.jad.2020.09.049 32950843
Engelkemeyer S. M. Marwit S. J. (2008). Posttraumatic growth in bereaved parents. Journal of Traumatic Stress, 21 (3 ), 344–346. 10.1002/jts.20338 18553420
Feigelman W. Jordan J. R. Gorman B. S. (2009). Personal growth after a suicide loss: Cross-sectional findings suggest growth after loss may be associated with better mental health among survivors. Omega: Journal of Death and Dying, 59 (3 ), 181–202. 10.2190/OM.59.3.a 19791516
Gesi C. Carmassi C. Cerveri G. Carpita B. Cremone I. M. Dell’Osso L. (2020). Complicated grief: What to expect after the coronavirus pandemic. Frontiers in Psychiatry, 11 , 489. 10.3389/fpsyt.2020.00489 32574243
Goldsmith B. Morrison R. S. Vanderwerker L. C. Prigerson H. G. (2008). Elevated rates of prolonged grief disorder in African Americans. Death Studies, 32 (4 ), 352–365. 10.1080/07481180801929012 18850684
Gomez-Salgado J. Andres-Villas M. Dominguez-Salas S. Diaz-Milanes D. Ruiz-Frutos C. (2020). Related health factors of psychological distress during the COVID-19 pandemic in Spain. International Journal of Environmental Research and Public Health, 17 (11 ), 3947. 10.3390/ijerph17113947 32498401
Hagenaars M. A. van Minnen A. (2010). Posttraumatic growth in exposure therapy for PTSD. Journal of Traumatic Stress, 23 (4 ), 504–508. 10.1002/jts.20551 20690196
Hayes A. F. Preacher K. J. (2010). Quantifying and testing indirect effects in simple mediation models when the constituent paths are nonlinear. Multivariate Behavioral Research, 45 (4 ), 627–660. 10.1080/00273171.2010.498290 26735713
He L. Tang X. Zhu Z. Wang J. (2014). Great pain: Qualitative research on grief reactions of the parents who lost their single child. Chinese Journal of Clinical Psychology, 22 (5 ), 792–798. 10.16128/j.cnki.1005-3611.2014.05.053
He L. Wang J. Yu W. Xie Q. Tang S. (2013). Grief reactions and correlates among 301 bereaved individuals. Chinese Journal of Clinical Psychology, 21 (6 ), 932–935. 10.16128/j.cnki.1005-3611.2013.06.031
Holland J. M. Currier J. M. Coleman R. A. Neimeyer R. A. (2010). The integration of stressful life experiences scale (ISLES): Development and initial validation of a new measure. International Journal of Stress Management, 17 (4 ), 325–352. 10.1037/a0020892
Ingravallo F. (2020). Death in the era of the COVID-19 pandemic. The Lancet Public Health , 5 (5 ), Article e258. 10.1016/s2468-2667(20)30079-7
Jin Y. Liu D. Li J. (2014). Promoting factors of post-traumatic growth, model and intervene. Advances in Psychological Science, 22 (2 ), 304. 10.3724/sp.J.1042.2014.00304
Levi-Belz Y. (2020). Longitudinal intercorrelations between complicated grief and posttraumatic growth among suicide survivors. Archives of Suicide Research, 26 (2 ), 1–15. 10.1080/13811118.2020.1820412 32669055
Li J. Prigerson H. G. (2016). Assessment and associated features of prolonged grief disorder among Chinese bereaved individuals. Comprehensive Psychiatry, 66 , 9–16. 10.1016/j.comppsych.2015.12.001 26995230
Lund D. Caserta M. Utz R. De Vries B. (2010). Experiences and early coping of bereaved spouses/partners in an intervention based on the dual process model (dpm). Omega: Journal of Death and Dying, 61 (4 ), 291–313. 10.2190/OM.61.4.c 21058611
Lundorff M. Holmgren H. Zachariae R. Farver-Vestergaard I. O’Connor M. (2017). Prevalence of prolonged grief disorder in adult bereavement: A systematic review and meta-analysis. Journal of Affective Disorders, 212 , 138–149. 10.1016/j.jad.2017.01.030 28167398
Maercker A. Zoellner T. (2004). The Janus face of self-perceived growth: Toward a two-component model of posttraumatic growth. Psychological Inquiry, 15 (1 ), 41–48. 10.2307/20447200
Marazziti D. Pozza A. Di Giuseppe M. Conversano C. (2020). The psychosocial impact of COVID-19 pandemic in Italy: A lesson for mental health prevention in the first severely hit European country. Psychological Trauma, 12 (5 ), 531–533. 10.1037/tra0000687 32525387
McClatchey I. S. (2020). Trauma-informed care and posttraumatic growth among bereaved youth: A pilot study. Omega: Journal of Death and Dying, 82 (2 ), 196–213. 10.1177/0030222818804629 30303460
Milman E. Neimeyer R. A. Fitzpatrick M. MacKinnon C. J. Muis K. R. Cohen S. R. (2019). Prolonged grief and the disruption of meaning: Establishing a mediation model. Journal of Counseling Psychology, 66 (6 ), 714–725. 10.1037/cou0000370 31647284
Moats M. E. (2011). Learning through loss: A qualitative study investigating United States and Chinese meaning-making through bereavement. (Publication Number AAI3412541) APA PsycInfo® https://www.proquest.com/dissertations-theses/learning-through-loss-qualitative-study/docview/862788731/se-2?accountid=8554
Neimeyer R. A. Burke L. A. Mackay M. M. van Dyke Stringer J. G. (2010). Grief therapy and the reconstruction of meaning: From principles to practice. Journal of Contemporary Psychotherapy, 40 (2 ), 73–83. 10.1007/s10879-009-9135-3
Pan H. Cheung C.-K. Hu J. (2018). Intimacy and complicated grief among Chinese elders having lost their spouses: Mediating role of meaning making. Journal of Loss and Trauma, 23 (3 ), 244–258. 10.1080/15325024.2018.1435367
Park C. L. (2010). Making sense of the meaning literature: An integrative review of meaning making and its effects on adjustment to stressful life events. Psychological Bulletin, 136 (2 ), 257–301. 10.1037/a0018301 20192563
Podsakoff P. M. MacKenzie S. B. Lee J. Y. Podsakoff N. P. (2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88 (5 ), 879–903. 10.1037/0021-9010.88.5.879 14516251
Pressman D. L. Bonanno G. A. (2016). With whom do we grieve? Social and cultural determinants of grief processing in the United States and China. Journal of Social and Personal Relationships, 24 (5 ), 729–746. 10.1177/0265407507081458
Prigerson H. G. Maciejewski P. K. Reynolds C. F. Bierhals A. J. Newsom J. T. Fasiczkaa A. Frank E. Doman J. Miller M. (1995). Inventory of complicated grief: A scale to measure maladaptive symptoms of loss. Psychiatry Research, 59 (1–2 ), 65–79. 10.1016/0165-1781(95)02757-2 8771222
Rozalski V. Holland J. M. Neimeyer R. A. (2016). Circumstances of death and complicated grief: Indirect associations through meaning made of loss. Journal of Loss and Trauma, 22 (1 ), 11–23. 10.1080/15325024.2016.1161426
Salloum A. Bjoerke A. Johnco C. (2019). The associations of complicated grief, depression, posttraumatic growth, and hope among bereaved youth. Omega: Journal of Death and Dying, 79 (2 ), 157–173. 10.1177/0030222817719805 28705039
Shear K. Frank E. Houck P. R. Reynolds C. F. (2005). Treatment of complicated grief: A randomized controlled trial. JAMA, 293 (21 ), 2601–2608. 10.1001/jama.293.21.2601 15928281
Shi G. Wen J. Xu X. Zhou N. Wang J. Shi Y. Liu H. Stelzer E. M. Stelzer E. M. (2019). Culture-related grief beliefs of Chinese Shidu parents: Development and psychometric properties of a new scale. European Journal of Psychotraumatology, 10 (1 ), 1626075. 10.1080/20008198.2019.1626075 31303970
Stroebe M. Schut H. (2021). Bereavement in times of COVID-19: A review and theoretical framework. Omega: Journal of Death and Dying, 82 (3 ), 500–522. 10.1177/0030222820966928 33086903
Tang R. Xie T. Jiao K. Xu X. Zou X. Qian W. Wang J. (2021). Grief reactions and grief counseling among bereaved Chinese individuals during COVID-19 pandemic: Study protocol for a randomized controlled trial combined with a longitudinal study. International Journal of Environmental Research and Public Health, 18 (17 ), 9061. 10.3390/ijerph18179061 34501650
Tedeschi R. G. Calhoun L. G. (1996). The posttraumatic growth inventory: Measuring the positive legacy of trauma. Journal of Traumatic Stress, 9 (3 ), 455–471. 10.1002/jts.2490090305 8827649
Tedeschi R. G. Calhoun L. G. (2004). Target article: “Posttraumatic growth: Conceptual foundations and empirical evidence”. Psychological Inquiry, 15 (1 ), 1–18. 10.1207/s15327965pli1501_01
Wallace C. L. Wladkowski S. P. Gibson A. White P. (2020). Grief during the COVID-19 pandemic: Considerations for palliative care providers. Journal of Pain and Symptom Management , 60 (1 ), Article e70–e76. 10.1016/j.jpainsymman.2020.04.012
Waller A. Turon H. Mansfield E. Clark K. Hobden B. Sanson-Fisher R. (2016). Assisting the bereaved: A systematic review of the evidence for grief counselling. Palliative Medicine, 30 (2 ), 132–148. 10.1177/0269216315588728 26415735
Wang J. Chen Y. Wang Y. Liu X. (2011). Revision of the posttraumatic growth inventory and testing its reliability and validity. Journal of Nursing Science, 26 (14 ), 26–28. 10.3870/hlxzz.2011.14.026
Williams H. Skalisky J. Erickson T. M. Thoburn J. (2020). Posttraumatic growth in the context of grief: Testing the mindfulness-to-meaning theory. Journal of Loss and Trauma, 26 (7 ), 611–623. 10.1080/15325024.2020.1855048
World Health Organization (2022a). Coronavirus (COVID-19). https://who.sprinklr.com/. (accessed 31th March 2022).
World Health Organization (2022b). ICD-11 for mortality and morbidity statistics. 6B42 prolonged grief disorder. https://icd.who.int/browse11/l-m/en#/http%3A%2F%2Fid.who.int%2Ficd%2Fentity%2F118383
Xie C. S. Kim Y. (2022). Post-traumatic growth during COVID-19: The role of perceived social support, personality, and coping strategies. Healthcare (Basel), 10 (2 ), 224. 10.3390/healthcare10020224 35206839
Xu W. Fu Z. He L. Schoebi D. Wang J. (2015). Growing in times of grief: Attachment modulates bereaved adults’ posttraumatic growth after losing a family member to cancer. Psychiatry Research, 230 (1 ), 108–115. 10.1016/j.psychres.2015.08.035 26341874
Xu J. He L. Wang J. (2020a). The bereaved one’s grief reaction characteristics and suggesstions for grief counseling under the COVID-19 epidemic. Chinese Mental Health Journal, 34 (4 ), 382–386. 10.3969/j.issn.1000-6729.2020.4.016
Xu X. Wen J. Zhou N. Shi G. Tang R. Wang J. Skritskaya N. A. (2020b). Grief and posttraumatic growth among Chinese bereaved parents who lost their only child: The moderating role of interpersonal loss. Frontiers in Psychology , 11 , 558313. 10.3389/fpsyg.2020.558313 33162907
Yilmaz M. Zara A. (2016). Traumatic loss and posttraumatic growth: The effect of traumatic loss related factors on posttraumatic growth. Anatolian Journal of Psychiatry, 17 (1 ), 5. 10.5455/apd.188311
Zakarian R. J. McDevitt-Murphy M. E. Bellet B. W. Neimeyer R. A. Burke L. A. (2019). Relations among meaning making, PTSD, and complicated grief following homicide loss. Journal of Loss and Trauma, 24 (3 ), 279–291. 10.1080/15325024.2019.1565111
Zhang Y. Jia X. (2018). A qualitative study on the grief of people who lose their only child: From the perspective of familism culture. Frontiers in Psychiatry, 9 , 869. 10.3389/fpsyg.2018.00869
Zhao G. Li X. Fang X. Zhao J. Yang H. Stanton B. (2007). Care arrangements, grief and psychological problems among children orphaned by AIDS in China. AIDS Care, 19 (9 ), 1075–1082. 10.1080/09540120701335220 18058390
Zoellner T. Maercker A. (2006). Posttraumatic growth in clinical psychology — a critical review and introduction of a two component model. Clinical Psychology Review, 26 (5 ), 626–653. 10.1016/j.cpr.2006.01.008 16515831
| 36423236 | PMC9703020 | NO-CC CODE | 2022-11-29 23:21:06 | no | Omega (Westport). 2022 Nov 24;:00302228221141937 | utf-8 | Omega (Westport) | 2,022 | 10.1177/00302228221141937 | oa_other |
==== Front
Exp Biol Med (Maywood)
Exp Biol Med (Maywood)
EBM
spebm
Experimental Biology and Medicine
1535-3702
1535-3699
SAGE Publications Sage UK: London, England
36426683
10.1177/15353702221140406
10.1177_15353702221140406
Introduction
Applied artificial intelligence in healthcare: Listening to the winds of change in a post-COVID-19 world
https://orcid.org/0000-0003-2047-4759
Shaban-Nejad Arash 1
Michalowski Martin 2
Bianco Simone 3
Brownstein John S. 4
Buckeridge David L 5
Davis Robert L 1
1 UTHSC-ORNL Center for Biomedical Informatics, Department of Pediatrics, College of Medicine, University of Tennessee Health Science Center, Memphis, TN 38103, USA
2 School of Nursing, University of Minnesota Twin Cities, Minneapolis, MN 55455, USA
3 Altos Labs – Bay Area Institute of Science, Redwood City, CA 94065, USA
4 Boston Children’s Hospital, Harvard University, Boston, MA 02115, USA
5 Department of Epidemiology, Biostatistics, and Occupational Health, McGill University, Montreal, QC H3A 1G1, Canada
Arash Shaban-Nejad. Email: [email protected]
25 11 2022
25 11 2022
15353702221140406© 2022 by the Society for Experimental Biology and Medicine
2022
The Society for Experimental Biology and Medicine
This article is made available via the PMC Open Access Subset for unrestricted re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the COVID-19 pandemic or until permissions are revoked in writing. Upon expiration of these permissions, PMC is granted a perpetual license to make this article available via PMC and Europe PMC, consistent with existing copyright protections.
This editorial article aims to highlight advances in artificial intelligence (AI) technologies in five areas: Collaborative AI, Multimodal AI, Human-Centered AI, Equitable AI, and Ethical and Value-based AI in order to cope with future complex socioeconomic and public health issues.
Health AI
artificial intelligence
machine learning
AI governance
multimodal AI
human-centered AI
ethical AI
COVID-19
edited-statecorrected-proof
typesetterts1
==== Body
pmcIntroduction
The COVID-19 pandemic impacted almost every sector of our modern world and created unprecedented change and disruption in the way we live, work, communicate, commute, socialize, learn, entertain ourselves, and do business. From the onset of the COVID-19 pandemic, artificial intelligence (AI) tools and technologies have been used to improve disease surveillance, screening, diagnostics, case detection, prediction, risk stratification, drug and vaccine development, resource allocation, and socioeconomic interventions. Despite their great potential, these AI tools have had little, if any, impact on the response to this devastating pandemic. Many of published prediction models were inadequately reported and most of them had low accuracy, weak predicative power, high risk of bias,1 and methodological flaws with limited potential for medical and clinical use.2 Part of the problem might have originated from the lack of access to high-quality COVID-19 data sets,3 insufficient historical data, and inaccurate training data, which may cause researchers to rely on heterogeneous and noisy data collected at low temporal and geographic resolutions. Furthermore, critical issues stem from the frequent suboptimal implementation and use of AI technologies including the ways they are shared, evaluated, governed, and regulated.4,5
The road ahead
Regardless of the above-mentioned challenges, the experiences gained from the COVID-19 pandemic can accelerate innovations in AI technology to better prepare societies to respond to future crises. According to the World Bank,6 to accelerate AI development at the country level, policymakers are advised to focus on AI research, talent development, supporting entrepreneurship, ethical or trustworthy AI, increasing access to quality data, adoption of AI for public service, strategic sectoral targeting of AI, and strengthening AI governance. Furthermore, to cope with complex socioeconomic and public health issues, we anticipate AI technologies to advance in the following five areas:
1. Collaborative AI: During the emergencies imposed by the pandemic, researchers tried to create their own solutions, which often led to many isolated, standalone, and redundant models with similar limitations and biases. Future AI must foster opportunities to promote collaboration (cooperation, competition, or coordination)7 from multiple stakeholders (human and machines) to maximize a common goal while balancing each entity’s individual interests.
2. Multimodal AI: Multimodal technologies8 enable users to access, integrate, and process ever-increasing multimodal and complex medical data sets and interact with a system in different modalities at the same time. Multimodal AI particularly attempts to process, manage, and understand these multimodal data through making multimodal inferences to analyze complex associations and relationships between various biological processes, health indicators, risk factors, and health outcomes, and developing exploratory and explanatory models.
3. Human-centered AI: With the intention to create AI models that “amplify and augment rather than displace human abilities. It seeks to preserve human control in a way that ensures AI meets human needs while also operating transparently, delivering equitable outcomes, and respecting privacy,”9 human-centered AI focuses on including the human in the loop. In this way, it provides substantial gains in transparency, fairness, accountability, reliability, and explainability of AI systems.10,11 The IEEE Global Initiative on Ethics of Autonomous and Intelligent Systems recently announced a few principles to advance discussion on the alignment between AI models and human rights and interests.12
4. Equitable AI: The Covid-19 pandemic once again exposed social, economic, and racial inequalities among under-represented and marginalized communities across the globe.13,14 Precision equity15 should be an integral part of precision health16 and health AI. According to the World Economic Forum’s call for more inclusive AI infrastructure,17 AI scientists and designers “should identify and partner with representatives of these impacted stakeholders on data collection methods, especially when identifying new or non-traditional resources for gathering data.” Algorithmic equity18 is also an important area that needs special attention to ensure that decisions and policies made based on AI algorithms are nondiscriminatory.
5. Ethical and value-based AI: Future AI solutions should consider ethical issues and incorporate human values, in their design and use.19 An important step here is listening to and understanding individuals’ concerns and respecting their personal autonomy and right to informed consent, and dissent.
Thematic issue on the future of AI
This thematic issue on the future of AI includes various contributions presenting results on theory, methods, systems, and applications of AI in medicine and healthcare. Boursalie et al.20 studied the challenges of evaluating deep learning–based imputation models by conducting a comparative analysis between root mean square error (RMSE), a predictive accuracy metric, and evaluation metrics used in statistical literature, including qualitative, predictive accuracy, statistical distance, and descriptive statistics metrics. Using two tabular data sets from the healthcare and financial sectors, they design an aggregated metric to evaluate deep learning–based imputation models called reconstruction loss (RL). Tanwar A et al.21 proposed an unsupervised method that leverages external clinical knowledge and contextualized word embeddings by ClinicalBERT for numerical reasoning in different phenotypic contexts. Jana et al.22 presented methods for predicting intensive care unit (ICU) length of stay as well as need for critical interventions for patients based on vital signs, laboratory measurements, and nursing notes prepared within the first 24 h of ICU stay. Their approach has been built and cross-validated over publicly available Medical Information Mart for Intensive Care (MIMIC-III v1.4) data set.
Xia et al.23 summarized publicly available data sets annotated by respiratory experts and reviewed the latest machine learning methods used for respiratory screening during the Covid-19 pandemic. Scaboro et al.24 compared some of the current systems for detecting adverse drug events using social media data and proposed strategies to increase the robustness of these systems. Using unstructured clinical notes, Karisani et al.25 created pipeline to infer the existence of alternative biological pathways from clinical phenotypes. Mohammadi et al.26 applied an existing weakly supervised learning algorithm to a real-world data set in histopathology, with over 90% validation accuracy. Then they extended this method to handle multiclass slide-level labels and presented an end-to-end saliency-mapping algorithm to segment regions of interest at the pixel level based only on slide-level labels.
Authors’ Contributions: A.S.N., M.M., S.B. conceptualized, drafted, reviewed, and edited the manuscript. J.S.B., D.L.B., R.L.D. reviewed and edited the manuscript.
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding: The author(s) received no financial support for the research, authorship, and/or publication of this article.
ORCID iD: Arash Shaban-Nejad https://orcid.org/0000-0003-2047-4759
==== Refs
References
1 Wynants L Van Calster B Collins GS Riley RD Heinze G Schuit E Bonten MMJ Dahly DL Damen JAA Debray TPA de Jong VMT De Vos M Dhiman P Haller MC Harhay MO Henckaerts L Heus P Kammer M Kreuzberger N Lohmann A Luijken K Ma J Martin GP McLernon DJ Andaur Navarro CL Reitsma JB Sergeant JC Shi C Skoetz N Smits LJM Snell KIE Sperrin M Spijker R Steyerberg EW Takada T Tzoulaki I van Kuijk SMJ van Bussel B van der Horst ICC van Royen FS Verbakel JY Wallisch C Wilkinson J Wolff R Hooft L Moons KGM van Smeden M. Prediction models for diagnosis and prognosis of covid-19: systematic review and critical appraisal. BMJ 2020;369 :m1328. Update in: BMJ 2021;372:n236. Erratum in: BMJ 2020;369:m22X04.
2 Roberts M Driggs D Thorpe M Gilbey J Yeung M Ursprung S Aviles-Rivero AI Etmann C McCague C Beer L Weir-McCall JR Teng Z Gkrania-Klotsas E , AIX-COVNET, Rudd JHF Sala E Schönlieb C-B. Common pitfalls and recommendations for using machine learning to detect and prognosticate for COVID-19 using chest radiographs and CT scans. Nat Mach Intell 2021;3 :199–217.
3 Brakefield WS Ammar N Olusanya OA Shaban-Nejad A. An urban population health observatory system to support COVID-19 pandemic preparedness, response, and management: design and development study. JMIR Public Health Surveill 2021;7 :e28269.34081605
4 World Health Organization (WHO). Ethics and governance of artificial intelligence for health. Geneva: WHO, 2021.
5 Krass M Henderson P Mello MM Studdert DM Ho DE. How US law will evaluate artificial intelligence for covid-19. BMJ 2021;372 : n234.
6 World Bank. Harnessing artificial intelligence for development on the post-COVID-19 era: a review of national AI strategies and policies. analytical insights. Washington, DC: World Bank, 2021. https://openknowledge.worldbank.org/handle/10986/35619
7 IBM Research. Collaborative AI. https://researcher.watson.ibm.com/researcher/view_group.php?id=7806 (accessed 30 October 2022).
8 Shaban-Nejad A Michalowski M Bianco S . Multimodal artificial intelligence: next wave of innovation in healthcare and medicine. In: Shaban-Nejad A Michalowski M Bianco S (eds) Multimodal AI in healthcare: a paradigm shift in health intelligence (Studies in Computational Intelligence Series, Vol. 1060). Cham: Springer, 2022, pp. 1–9.
9 Geyer W Weisz J Pinhanez CS Daly E. What is human-centered AI? IBM Research Blog, 31 March 2022. https://research.ibm.com/blog/what-is-human-centered-ai (accessed 30 October 2022).
10 Shaban-Nejad A Michalowski M Brownstein JS Buckeridge DL. Guest editorial explainable AI: towards fairness, accountability, transparency and trust in healthcare. IEEE J Biomed Health 2021;25 : 2374–5.
11 Shaban-Nejad A Michalowski M Buckeridge DL. Explainability and interpretability: keys to deep medicine. In: Shaban-Nejad A Michalowski M Buckeridge DL (eds) Explainable AI in healthcare and medicine (Studies in Computational Intelligence, Vol. 914). Cham: Springer, 2021, pp.1–10.
12 The IEEE Global Initiative on Ethics of Autonomous and Intelligent Systems. Ethically aligned design: a vision for prioritizing human well-being with autonomous and intelligent systems, Version 2. New York: IEEE, 2017. http://standards.ieee.org/develop/indconn/ec/autonomous_systems.html (accessed 30 October 2022).
13 Brakefield WS Olusanya OA White B Shaban-Nejad A. Social determinants and indicators of COVID-19 among marginalized communities: a scientific review and call to action for pandemic response and recovery. Disaster Med Public Health Prep. Epub ahead of print 2 May 2022. DOI: 10.1017/dmp.2022.104.
14 McCarthy G Shore S Ozdenerol E Stewart A Shaban-Nejad A Schwartz DL. History repeating-how pandemics collide with health disparities in the United States. J Racial Ethn Health Disparities. Epub ahead of print 20 May 2022. DOI: 10.1007/s40615-022-01331-5.
15 Shaban-Nejad A Michalowski M Peek N Brownstein JS Buckeridge DL. Seven pillars of precision digital health and medicine. Artif Intell Med 2020;103 :101793.32143798
16 Shaban-Nejad A Michalowski M (eds). Precision health and medicine: a digital revolution in healthcare (Studies in Computational Intelligence Series, Vol. 843 ). Cham: Springer Nature, 2020.
17 The World Economic Forum. A blueprint for equity and inclusion in artificial intelligence. White Paper, June 2022. https://www.weforum.org/whitepapers/a-blueprint-for-equity-and-inclusion-in-artificial-intelligence/ (accessed 30 October 2022).
18 Osoba OS Boudreaux B Saunders JJ Irwin L Mueller PA Cherney S. Algorithmic equity: a framework for social applications. Santa Monica, CA: RAND Corporation, 2019.
19 van de Poel I. Embedding values in artificial intelligence (AI) systems. Mind Mach 2020;30 :385–409.
20 Boursalie O Samavi R Doyle TE. Evaluation methodology for deep learning imputation models. Exp Biol Med. Epub ahead of print 21 September 2022. DOI: 10.1177/1535370222112160.
21 Tanwar A Zhang J Ive J Gupta V Guo Y. Phenotyping in clinical text with unsupervised numerical reasoning for patient stratification. Exp Biol Med. Epub ahead of print 11 October 2022. DOI: 10.1177/15353702221118092.
22 Jana S Dasgupta T Dey L. Predicting medical events and ICU requirements using a multimodal multiobjective transformer network. Exp Biol Med. Epub ahead of print 16 October 2022. DOI: 10.1177/15353702221126559.
23 Xia T Han J Mascolo C. Exploring machine learning for audio-based respiratory condition screening: a concise review of databases, methods, and open issues. Exp Biol Med. Epub ahead of print 16 August 2022. DOI: 10.1177/15353702221115428.
24 Scaboro S Portelli B Chersoni E Santus E Serra G. Increasing ADE extraction robustness on social media: a case study on negation and speculation. Exp Biol Med. Epub ahead of print 31 October 2022. DOI: 10.1177/15353702221128577.
25 Karisani N Platt DE Basu S Parida L. Topology and redescriptions detect multiple alternative biological pathways from clinical phenotypes. Exp Biol Med 2022. in press.
26 Mohammadi M Cooper J Arandjelovic O Fell C Morrison D Syed S Konanahalli P Bell S Bryson G Harrison DJ Harris-Birtill D. Weakly supervised learning and interpretability for endometrial whole slide image diagnosis. Exp Biol Med. Epub ahead of print 25 October 2022. DOI: 10.1177/15353702221126560.
| 36426683 | PMC9703021 | NO-CC CODE | 2022-11-29 23:21:06 | no | Exp Biol Med (Maywood). 2022 Nov 25;:15353702221140406 | utf-8 | Exp Biol Med (Maywood) | 2,022 | 10.1177/15353702221140406 | oa_other |
==== Front
Can Pharm J (Ott)
Can Pharm J (Ott)
CPH
spcph
Canadian Pharmacists Journal : CPJ
1715-1635
1913-701X
SAGE Publications Sage CA: Los Angeles, CA
10.1177/17151635221136552
10.1177_17151635221136552
Original Research
Experiences of community pharmacists administering COVID-19 vaccinations: A qualitative study
Gerges Sandra BScPhm, PharmD, MSc the Faculty of Health Sciences and Wellness, Pharmacy Technician Program, Humber College, Toronto
Gudzak Victoria BScH, MSc the Leslie Dan Faculty of Pharmacy, University of Toronto, Ontario
https://orcid.org/0000-0003-0821-3222
Bowles Susan BScPhm, PharmD, MSc the Department of Pharmacy, Nova Scotia Health and College of Pharmacy, Dalhousie University, Halifax, Nova Scotia
Logeman Charlotte MPH The Hospital for Sick Children, Toronto, Ontario
Fadaleh Sarah Abu BSc, MSc the Leslie Dan Faculty of Pharmacy, University of Toronto, Ontario
https://orcid.org/0000-0003-2713-0975
Bucci Lucie M. MA Bucci-Hepworth Health Services, Pincourt, Quebec
https://orcid.org/0000-0003-4432-8975
Taddio Anna BScPhm, MSc, PhD the Leslie Dan Faculty of Pharmacy, University of Toronto, Ontario
The Hospital for Sick Children, Toronto, Ontario
Contact [email protected].
25 11 2022
25 11 2022
1715163522113655223 9 2022
17 10 2022
© The Author(s) 2022
2022
Canadian Pharmacists Association
This article is made available via the PMC Open Access Subset for unrestricted re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the COVID-19 pandemic or until permissions are revoked in writing. Upon expiration of these permissions, PMC is granted a perpetual license to make this article available via PMC and Europe PMC, consistent with existing copyright protections.
Background:
Throughout the COVID-19 pandemic, community pharmacists have played an important role in the provision of patient care, including the delivery of COVID-19 vaccines. The additional workload and related demands arising from these extended services might affect worklife burnout. This qualitative study explored the experiences of Canadian community pharmacists in providing COVID-19 vaccines during the COVID-19 pandemic.
Methods:
Eighteen community pharmacists across 10 provinces were asked about vaccination processes and perceptions about their role in separate, virtual semistructured interviews. Interviews were transcribed verbatim and analyzed using a deductive approach using the Areas of Worklife Burnout framework, with pharmacists’ self-reported descriptions of their activities summarized using the Association of Faculties of Pharmacy of Canada’s (AFPC) professional competencies framework.
Results:
Participants identified aspects of their role that were rewarding and challenging. Some challenges included lack of control, increased workload, inadequate communication, unfair treatment and conflicting values. They described being able to meet challenges and demonstrating resiliency via adaptability, developing communities and valuing their contribution to ending the COVID-19 pandemic. Self-identified AFPC competencies contributing to their ability to manage their worklife included care-provider, professional, leader-manager, collaborator and scholar.
Conclusion:
Pharmacists accepted their additional responsibility of managing COVID-19 vaccines during the COVID-19 pandemic. However, they expressed some challenges with this role. To ensure sustainability of these services, prioritizing adequate resources, work processes and efficient communication with all relevant stakeholder groups, including public health, government and corporate leaders, is recommended for the future. Can Pharm J (Ott) 2023;156(Suppl):xx-xx.
Canadian Institutes of Health Research https://doi.org/10.13039/501100000024 Foundation Grant (FRN 159905) Public Health Agency of Canada https://doi.org/10.13039/100011094 Immunization Partnership Fund (1921-HQ-000220) edited-statecorrected-proof
typesetterts1
==== Body
pmcIntroduction
The World Health Organization declared COVID-19 as a pandemic on March 11, 2020, with the first confirmed case in Canada reported on January 25 2020.1,2 Canada has reported 3.8 million cases of COVID-19 and more than 41,000 deaths as of early June 2022.3 Throughout this time, community pharmacists have played an important role in helping to prevent morbidity and mortality of COVID-19 in addition to providing their regular patient care services. Many Canadian jurisdictions expanded prescribing privileges for pharmacists and allowed for point-of-care testing, including for COVID-19 and administering COVID-19 vaccines.4,5 To date, it is estimated that Canadian pharmacists have provided more than 17 million doses of COVID-19 vaccine.6
While the rationale for integrating community pharmacy practitioners into the delivery of COVID-19 vaccines was to vaccinate as many people as quickly as possible, it is important to examine the personal impact that participation in this effort has had on pharmacists. Although the effect of pandemic-related stressors on the mental well-being and burnout of nurses and physicians is well documented,7-10 less is known about pharmacists’ perceptions of their roles during the COVID-19 pandemic and more specifically how vaccinations have contributed to burnout experienced within the profession during the pandemic thus far.11-14 The aim of this qualitative study was to explore the perceptions of pharmacists about their vaccinator role during the COVID-19 pandemic. This study is part of a series included in this supplement to the Canadian Pharmacists Journal that together summarize the findings of a program of research funded by the Public Health Agency of Canada, examining community pharmacy vaccination services and integrating a vaccination delivery framework (Comfort Ask Relax Distract; CARD) to improve the experiences of vaccine clients and pharmacy staff.15-22
Methods
This study used a qualitative descriptive design to gather an in-depth understanding of pharmacists’ experiences as COVID-19 vaccinators. Pharmacists practising in community pharmacies from across all 13 Canadian provinces and territories and administering COVID-19 vaccinations were eligible for participation. We used a combination of purposive and snowball sampling, aiming to include respondents from varied geographical locations, population densities, roles and genders.
All interviews were moderated by 1 interviewer and conducted virtually using an online platform (Zoom). A semistructured interview guide was used, and each interview lasted about an hour. Pharmacists were asked about the circumstances leading to their becoming a COVID-19 vaccinator, how they prepared for this role, what the vaccination processes were in the pharmacy where they worked and their experiences and satisfaction in this role. At the end, they were asked about their perceptions of CARD as a framework for vaccination delivery. The results giving their perceptions of CARD are reported separately in this supplement.15 The study was approved by the Research Ethics Board of the University of Toronto (40916), and all participants provided informed signed consent.
Sample Size and Analysis
Data saturation was used to guide the required sample size. In prior related work by our group, saturation was achieved with 12 interviews in pharmacist vaccinators.23 We planned to include up to 30 interviews in the present study to account for variability that may have been present due to a more heterogenous participant sample.
Interviews were recorded and transcribed verbatim. The qualitative analysis initially followed an inductive approach.24 Line-by-line coding was performed independently by 4 researchers (S.G., S.B., V.G., A.T.), who then met to discuss their results and identify relationships between codes. After coding the first 2 interviews, the analysis was altered to a deductive approach as emerging themes were determined to be consistent with the preexisting Areas of Worklife Burnout Framework by Leiter and Maslach.25
In their comprehensive model of job burnout, Maslach and Leiter describe person-job incongruence within 6 domains of worklife: control, workload, community, fairness, rewards and values. These domains predict the level of experienced burnout (characterized by exhaustion, cynicism and inefficiency), which in turn predicts work outcomes such as job turnover.25 These 6 domains of job burnout are characterized as complex and interrelated, rather than simple and/or linear.25 Control is the starting point, as it will influence the extent to which people can attain congruence in workload, reward, fairness and community. Workload is described as directly contributing to exhaustion, whereas community, fairness and reward all affect the individual’s values. Values are integral to the model and affect all 3 characteristics of burnout (exhaustion, cynicism and inefficiency), which ultimately lead to negative work outcomes and employee turnover. When values are incongruent with demands/stressors, employees perceive their workplace’s mission to be incompatible with their well-being.25
In addition, initial coding discussions revealed that pharmacists’ self-reported activities, in terms of their roles and competencies, aligned with entry-to-practice pharmacy competencies as defined by the 2017 Association of Faculties of Pharmacies of Canada (AFPC).26 As a result, the AFPC taxonomy was used as the framework to describe how roles influenced the different 6 domains of the burnout. Key roles that were identified included professional, leader-manager, care provider, collaborator and scholar.
One researcher (S.G.) coded all of the transcripts using both frameworks. Regular meetings were held with the other 3 researchers (S.B., V.G., A.T.) to discuss the results until all transcripts were coded. Disagreements were resolved using consensus. Data saturation was determined to be achieved by the lack of substantive new information.
Results
The study was conducted between May 28 and August 16, 2021. Altogether, 18 pharmacists from 10 provinces participated. Demographic characteristics are shown in Table 1. Pharmacists had experience vaccinating with multiple COVID-19 vaccine brands, including Pfizer-BioNTech Comirnaty, Moderna Spikevax and Astra-Zeneca Vaxzevria.
Table 1 Characteristics of pharmacist participants (N = 18)
Characteristic Value
Gender, No. male (%) 13 (72.2)
Ethnicity, No. Caucasian (%) 10 (55.5)
Age,* median in years 31-40
Pharmacy type, No. independent (%) 11 (61.1)
Role, No. staff pharmacist (%) 8 (44.4)
* Reported age ranges (in years): 21-30, 31-40, 41-50, 51-60; n = 17.
Drivers of pharmacists’ worklife burnout were identified in all 6 domains of the Worklife Burnout Framework25 and are described in detail below (with sample quotes in Table 2).
Table 2 Selected pharmacist participant quotes pertaining to the Areas of Worklife Burnout Framework25 (N = 18)
Domain (as per Maslach and Leiter25) AFPC attributes of pharmacists Example quotes*
Control
Having the opportunity to make choices and decisions, including problem solving and contributing to the responsible actions related to one’s work Leader-Manager
Engage with others to optimize the safety, effectiveness and efficacy of health care and contribute to a vision of a high-quality health care system “So, we’re able to more clearly articulate the guidelines and plan ahead of time to make it more of a streamlined process . . . because throughout the whole experience, a lot of times it just felt like certain pieces of news were just dropped on us, and then we had about a night or couple of hours to figure it out and book all the appointments for the next day. So, definitely there could be more communication so we’re a little bit better prepared as well.” P12
“I would tell a patient at 1 o’clock in the afternoon that ‘Sorry, I can’t book you because you’re not eligible,’ and then at 1:30 the media announcement comes out and now are eligible, which would create just . . . total chaos.” P04
“Labour has been a big, big challenge, where corporate wants to have minimum labour, if they make $100, they only want to spend $19 on the labour.” P18
“The flow of our ‘vaccine clinic,’ I call it . . . is worked out primarily by myself and my staff, right? We understood what needed to get done. It’s unique to our environment . . . but I also have 2 other good pharmacist friends and we did a lot more than just creating . . . a vaccine flow within our pharmacy . . . right? And as soon as we knew when we were getting vaccines, we created a template that we could follow and a process that we could follow; almost like a standard operating procedure without writing it down.” P09
“We have . . . recently hired someone to fill the gaps, and . . . the new graduates too were coming out of school this year ready to go, so that was a big help to have them ready to jump in and help . . . ” P17
“We ended up with waitlists of hundreds of people, and we were making phone calls. . . . So very shortly after we started, I subscribed to a platform that would allow . . . much more automation. The platform allowed people to register and then I would offer appointments. . . . I would create appointments within the platform, send out automated messages, and people would book their own time.” P05
Workload
The amount of work to complete in a day; the frequency of surprising, unexpected events Care-provider
Provide patient-centred pharmacy care by using knowledge, skills and professional judgment to facilitate management of a patient’s medication and overall health needs across the care continuum “I was a bit hesitant at first, because we had enough on our plates with everything we’ve been doing due to COVID and lack of regular medical care, but it became apparent that we’re a very small community and we need to be front and centre and be able to offer these vaccines to our community. So we quickly got on board.” P07
“When news breaks out that, you know, AstraZeneca causes blood clots during the pandemic, who are patients going to call first? They’re not calling their family doctor because they’re inaccessible right now, ever since the start of the pandemic. So, pharmacists really did have to take on a larger role for their community. So, the first one they call would be their [laugh] community pharmacist to ask about it. And it’s happened, time and time again, where, you know, the same patients and even new patients keep coming to ask us about these things.” P01
Professional
Take responsibility and accountability for delivering pharmacy care to patients, communities and society through ethical practice and the high standards of behaviour that are expected of self-regulated professionals “Pharmacists and pharmacies have the ability, the logistically ability to participate as well as the skill and knowledge. For me, this is something that we can do well. Like we know, we know about vaccines and we know how to administer them and we have the facilities and the logistics to do it fairly quickly. So, for me, it was basically a no-brainer that I would participate in that if given the opportunity. . . . It felt like a way for me to contribute to the beginning of the end of the pandemic.” P05
Scholar
Applying medication therapy expertise, learning continuously, creating new knowledge and disseminating knowledge when teaching others “A lot of reading outside of work, researching, what the vaccine efficacy is, side effects, how to mix the doses, dilute them, anything like that, so it was a lot of extra work outside of my day-to-day job to prepare. A lot of research; a lot of looking up information . . . taking all that information, taking the research that was done and putting it in terms that could help them understand . . . so preparing people that way.” P17
Leader-Manager
Engage with others to optimize the safety, effectiveness and efficacy of health care and contribute to a vision of a high-quality health care system “The biggest downfall of this whole vaccination experience has been we’re dealing with 3 different programs. I’m dealing with a scheduling program, I’m dealing with the [province name] program and then I’m dealing with a pharmacist software program to bill. . . . The biggest downfall of this whole rollout was that there wasn’t a way to link everything.” P04
“We don’t want to have anybody burn out so we just can’t use one person to inoculate everyone and pharmacists are much more valued as knowledge and the registered technicians are much more valued in the technical aspects. So, we’re able to tease that apart, a little further and have each other perform to our excellence in our highest capacity. So, I could be doing a med review over the phone with a patient and documenting all that, while a technician’s giving an injection. So, the technician’s able to do that while I’m able to do the cognitive functions that I’m much better trained for, rather than having to do the injection.” P14
“I would say in a 3-day period we probably answered . . . 1000 phone calls . . . just from people looking to see where they could get their shots. Nothing else. It got to the point where honestly, we told our staff . . . if the caller ID is not one of our patients, let it go to voicemail, because the voicemail set up with the instructions on what to do if you’re looking just for a vaccine. Um, and I mean in my career, I never thought I’d see the day where I would . . . tell somebody on staff don’t answer the phone. But I mean . . . it got to a point where, in order for patients to get through, who really needed medications . . . we had to do that. You had to sort of monitor your phone calls so that . . . you were taking care of the patients who needed . . . refills or needed medications or had a medication question for you and filter out all the vaccine calls.” P04
“Corporate wouldn’t let me hire somebody. Even though I had 2 qualified PAs [pharmacy assistants] living close by . . . I knew that I was going to lose the assistant that I had, so I had brought them in 2 months earlier timeframe before this other assistant went away . . . and corporate didn’t hire them for 6 months, because they felt like there was not really a need right now, even though there was a need.” P18
Community
Community relates to an organization’s social environment; includes relationships with different stakeholder groups Collaborators
Work collaboratively with patients and intra- and interprofessional teams to provide safe, effective, efficient health care, thus fulfilling the needs of the community and society at large “There is a pharmacist Facebook group that started when COVID first started, just to keep everyone up to date on all the changes that were going on. So, a lot of people would post some helpful information there. . . . If there’s a particular resource that someone finds helpful, then they put it in there.” P15
“I think the pharmacy becomes your support system. So, even after shifts on some days, like we just . . . you know, stick around and like just chat with coworkers about things that happened during the day . . . if anyone has like a negative experience, it’s just easier to find someone to confide to who kind of has that similar experience as you because everyone’s going through the same thing. I think it solidified the pharmacy team, because sometimes . . . it just felt like, you know, all hands-on deck.” P12
“We had hundreds of thousands of doses of vaccines sitting in storage in a central facility and yet pharmacies are crying and Public Health is crying for doses up North. We’re not getting anything. It just seems so bizarre and no transparency. I mean, maybe they had a great reason why they were doing this, but they didn’t tell us why, so it’s very frustrating from a provider perspective to know that the, that the vaccines are out there and available, but they are not getting shipped to us, right? So that was the hardest part.” P08
Fairness
Individual and collective evaluation by employees about justice and fairness regarding decisions at work Professional
Take responsibility and accountability for delivering pharmacy care to patients, communities and society through ethical practice and the high standards of behaviour that are expected of self-regulated professionals “You didn’t have time, so we had to do our regular dispensing, consultations, our regular jobs on top of now having to answer all these questions that we don’t get paid to do over the counter—questions, phone calls, we don’t get paid for any of that. . . . At [XX dollars] an injection—to prep all the vaccine ahead of time, prep all the paperwork ahead of time, do the vaccinations, do the watching after and run the clinic as well—for that kind of money. . . . It’s a little bit frustrating how underpaid we are. I mean, that also goes along with flu shots too, that we are underpaid for giving flu shots.” P17
“The other added pressure right now to us is physicians aren’t working. I know patients who haven’t seen their family physician face to face in a year and a half and guess who’s picking up that slack? It’s coming down to us.” P04
“I don’t know of any other health care practitioner that [reassures patients] on their dime. Right? That’s the point I want to get across really. . . . There’s no other health care practitioner bar none that does all this on their own dime. Right? The doctor will say make an appointment, come and see me and bill the ministry to see you. Right? Our own dime and that’s the one thing that people appreciate. . . . So, yeah. We did a lot of reassurance.” P09
Rewards
The reward area of work life addresses the extent to which rewards—monetary, social and intrinsic—are consistent with expectations Care-provider
Provide patient-centred pharmacy care by using their knowledge, skills and professional judgment to facilitate management of a patient’s medication and overall health needs across the care continuum “So, if it’s just immunizing, it’s really satisfying. It’s a very professionally and personally satisfying goal to have right now because, like I mentioned . . . it’s something that people have been looking forward to since day one of the pandemic and oftentimes patients tell you how much this vaccination means to them once you’ve done . . . once you’ve completed the process they might be in tears, they might be overjoyed, they’re taking photos, they’re telling their friends, you know? They’re leaving the pharmacy incredibly happy. So that’s the one positive knowing how much of an impact you’re making right now in contributing to overcoming this pandemic, it’s incredible.” P01
“I’ve really become even more amazed at the influence we do have . . . when I explain this to people and people are saying ‘Yeah, okay, just give me Astra Zeneca. That’s great. I trust you more than I trust the media outlets. I trust you more than these committees.’ And I think that has been rewarding in a way to show that and I think we always knew it was there. It’s just one of those things that it makes you feel a lot better, knowing that, hey patients really do take our word for what’s going on.” P04
Values
The potential of work to contribute to the larger community; confidence that the organization’s mission is meaningful Care-provider
Provide patient-centred pharmacy care by using knowledge, skills and professional judgment to facilitate management of a patient’s medication and overall health needs across the care continuum “There are times we’ve actually felt that . . . we’ve let some things go that normally, we would be taking care of, because we are so focused on getting people vaccinated. And I think . . . talking to some of my colleagues. . . . The 2 things that really stand out right now . . . are (1) the same feeling that we’re not taking care of our patients how we really want to, and (2) we’re just starting to get burnt out. There are just way too many demands.” P04
“It has felt meaningful, but there are a lot of times, where it was like very physically demanding, and tiring as well, just because I feel like a lot of responsibilities were put on the pharmacies at once . . . a lot of times that communication wasn’t really great, but the vaccination experience overall . . . providing the vaccinations, I think that was a positive experience.” P12
“I’m not able to keep up because there are added tasks. . . . It was just the unsafe practice that I was getting drawn towards and then I realized it’s not something I could do forever. I ended up seeing doctors well before I stopped work in [month], but I started seeing the doctor in [earlier month], my family doctor, and asking for help and I couldn’t do it. . . . The last day before I stopped, I ended up just crying at the doctor’s office and she’s like, ‘yeah, it’s just unsafe for you to go back. You need to take the time off.’ So, she has given me [number] months to start with and then reassess after that.” P18
Professional
Take responsibility and accountability for delivering pharmacy care to patients, communities and society through ethical practice and the high standards of behaviour that are expected of self-regulated professionals “That’s just . . . not how I’m built. I mean, I’m not going to abandon people in their time of need. I’m relatively young and relatively healthy. I didn’t see my family at all because you know, I have elderly grandparents, I have elderly parents, so I mean, that was certainly a sacrifice because I knew the risk—I was probably at the highest risk of contracting it and giving it to someone else. My fiancée was working from home at the time, so I was taking my temperature every morning. I was, you know, monitoring myself for signs and symptoms. . . . You just can’t abandon these people. There’s no one else to take your place. So, it’s either you do it, or no one does.” P15
* Participants are denoted by their number (e.g., P01 = participant 1, P02 = participant 2, etc.).
Control
Pharmacists described feeling a lack of control over their work. This included usual activities, such as dispensing medications and patient counselling, and additional COVID-19 related activities, including COVID-19 testing and COVID-19 vaccine administration. They described their work situation as constantly changing. Communications regarding who qualified for COVID-19 vaccine administration were inconsistent and sometimes unpredictable. Pharmacists reported receiving little to no advance notice by health authorities regarding updates to patient vaccine eligibility. Information was often received through the same communication channels that were used to inform the general public (i.e., news reports, government websites). There were limited resources to be able to accommodate the work demands, including staff shortages, inability to hire extra staff due to corporate office refusals and unpredictability of vaccine shipments, which made it challenging for pharmacists to meet public demand.
Despite these challenges, pharmacists successfully leveraged their leader-manager role to adapt to the situation. They prioritized vaccination-related activities over other clinical services such as medication reviews. Some also reported hiring staff and developing and purchasing software to manage the complex vaccination processes.
Workload
Pharmacists described increases in workload during the pandemic. As care-providers, pharmacists filled in gaps in health services resulting from reduced access to other health care providers. Some pharmacists reported being concerned about not being able to address all patient health care needs due to the prioritization of vaccination services. Pharmacists reported an increase in telephone and in-person consultations.
Pharmacists also reported more time preparing, delivering and documenting pandemic-related clinical services, including vaccinations. Their role as professionals led them to accept the added responsibilities involved in the complex processes of vaccine storage, preparation, administration and documentation, which ultimately contributed to their workload.
Pharmacists participated in self-directed learning, including reviewing the primary literature to educate and prepare themselves about vaccination, demonstrating their role as scholars. This was part of their increased workload and was often done during “off” hours.
Pharmacists expressed frustration with inadequate technology support. Processes were reported as inefficient and often increasing rather than reducing workload, due to lack of integration of systems (i.e., vaccine acquisition, appointment booking, vaccine administration, reimbursement). As leader-managers, they altered work processes. Some created and purchased their own booking systems and hired staff to offset the additional work. Many staff pharmacists reported insufficient staffing, and their requests to have additional staff or hours were often denied by head office.
Community
Pharmacists reported that community factors contributed to both support and challenges. As collaborators, pharmacists described their relationships with patients, pharmacy staff and external pharmacies as positive, while those with physicians, public health, government and corporate offices were often strained. Pharmacists stated that patients appreciated this additional service and had increased trust in them. Pharmacy staff worked together more closely, and this strengthened their relationships. Pharmacists interacted with colleagues at other pharmacies more frequently than before, via pharmacy networks, social media and pharmacy associations. Most pharmacists described being disturbed by the unavailability of physicians for their patients, inadequate communication of public health and government with them and their lack of involvement in decision-making related to vaccination processes. They described finding out new information about the vaccines and their role at the same time as the public via mass media. Pharmacists hoped for more support from their corporate head offices than they received.
Fairness
Most pharmacists perceived that the compensation provided to them for their efforts and time was insufficient. Some pharmacists stated that the fee for administering the COVID-19 vaccines did not cover the costs associated with offering vaccination services. Many times, pharmacists were involved with assisting clients in making vaccination appointments, maintaining manual logs of waiting lists and personally contacting them about their appointments. They also reported providing additional services without compensation such as counselling about more ailments, providing advice about COVID-19 vaccinations and answering other COVID-19 pandemic-related queries. Despite this, they persevered, as they believed it was part of their professional role.
Rewards
Although pharmacists felt that they were not being compensated fairly for their services, they intrinsically valued caring for their patients, as a result of their role as care-providers, above all else. They also valued the positive feedback and gratitude expressed by their patients. They appreciated patients stating that they relied on and trusted them.
Values
Ultimately, pharmacists’ overall ability and willingness to persevere with their COVID-19 immunizer role was related to the perceived value they brought to their communities as care-providers. Pharmacists reported wanting to help their community and to help end the COVID-19 pandemic. Pharmacists also reported being concerned about patient safety and ensuring that patients’ health needs were met when there were so many gaps in health services. They worked extra shifts and longer hours; however, they still found themselves having to prioritize vaccinations over other patient health care needs, which increased their stress.
Pharmacists struggled when their own values of patient care and patient safety were not aligned with what they were able to provide for their patients. Despite this, pharmacists felt obligated to continue to work due to their professional role. Concern about practising in an “unsafe manner” with respect to making errors and causing harm to patients, however, did lead one pharmacist to take a leave of absence from their job.
Discussion
This study examined community pharmacists’ experiences and satisfaction with their role as COVID-19 vaccinators. Pharmacists reported that administering COVID-19 vaccinations was consistent with their professional role. While generally satisfied, they described a busy, chaotic and inefficient working environment with multiple competing demands, inadequate resources and suboptimal communication with key stakeholders (such as government and public health), that together contributed to increased stress. Factors that helped them to deal with their work situation included their ability to adapt to the changing conditions, leveraging supportive communities and valuing their contribution to their patients and to ending the COVID-19 pandemic.
Our findings add to an increasing body of literature examining COVID-19 pandemic-related challenges faced by health care providers.7-10 In the first large national survey on this topic, including 768 pharmacists, the INSPIRE study documented added workload due to an increase in number of patients seeking pharmacists’ counselling (vs. other settings), including guidance regarding COVID-19 vaccinations.27 Other studies with pharmacists have documented that increased workload, lack of control over the work environment, rapidly changing information, staffing shortages and conflicts with the head office or other management structures contributed to negative effects on mental health.11,12,28 Mitigating factors have been reported to include pharmacists’ community connections, positive feedback from patients and the ability to adapt to change.12,29
To date, pharmacies are estimated to have delivered an impressive 21% of all COVID-19 vaccine doses administered to the Canadian public.6 Based on the prominence of pharmacies in the delivery of COVID-19 vaccines during the pandemic, we believe that community pharmacy-based vaccination services are likely to take on a more central role in Canada in the future. Even prior to the COVID-19 pandemic, community pharmacies were increasingly relied on to deliver vaccinations due to their accessibility and demonstrated efficiency with vaccinating large numbers of individuals.30 Across the country, vaccination privileges for pharmacists have been continually expanding over time, including the ability to administer more vaccines and to vaccinate younger children. Vaccination administration privileges are also beginning to be extended to pharmacy technicians, in some provinces, as a result of needing to meet the uptake of COVID-19 vaccination.31 In April 2022, Nova Scotia pharmacists were given COVID-19 vaccination privileges for children 6 months of age and older, in anticipation of upcoming regulatory approval of COVID-19 vaccines for this population.32 Other provinces (e.g., Ontario) have followed suit.
It is important to note that while pharmacists in the present study were generally able to manage the increased work demands and associated stressors that arose from their COVID-19 vaccinator role, the time frame for the study was relatively short compared with the entire duration of the COVID-19 pandemic. While some work-related stressors may have subsided over time for some participants, others may have emerged. The ongoing impact of persistent demands and stressors on worklife burnout cannot be underestimated. For instance, new challenges have included administering vaccines in children aged 5 to 11 years, which takes more time and is more stressful for pharmacists than administering vaccines in adults.19,33 Recent expansion in the delivery of COVID-19 vaccines in infants and young children, along with other expanded scope services (e.g., common ambulatory conditions)34 that are taking place across the country are likely to place additional strains on pharmacists in their worklife in the ensuing months.
To this end, there is in fact emerging evidence that the ongoing COVID-19 pandemic is starting to take a toll on pharmacy professionals that is profoundly negative. In a recent national survey (N = 1399) led by the Canadian Pharmacists Association in January and February of 2022, 92% of pharmacy professionals (pharmacists and pharmacy technicians) were identified to be at risk of burnout. This was assessed using a validated measure—the Oldenburg Burnout Inventory.35 To support pharmacists to be able to continue to provide vaccination services, we recommend a number of changes, organized into 4 categories: (1) practice, (2) compensation, (3) communication and (4) support and well-being (Table 3). These recommendations are described in further detail below.
Table 3 Recommendations for vaccination service delivery in community pharmacies
Category Recommended change Explanation
Practice Increase staffing Increase staffing to account for increased vaccination (and other expanded scope) services, including pharmacists and pharmacy technicians
Improve vaccination delivery process Integrate evidence-based vaccination delivery framework to improve efficiency and safety of vaccination delivery (e.g., CARD system)
Efficient technology Scale up technology; integrate software tools like booking systems to help manage workflows by reducing client surges; health authorities can facilitate use of interoperable systems with central registries and repositories and subsidize adoption
Continuing education Provide continuous professional education and skills training to ensure the latest evidence is incorporated into care delivery to ensure quality care, including pharmacists and pharmacy technicians and resources and tools to support professional learning
Compensation Increase compensation for vaccine administration, vaccination-related services and non–vaccine-related cognitive services Provide compensation commensurate for services rendered and aligned with fair market value (for equivalent services by other providers)
Communication Optimize internal communications Maintain intra-organization communication to support a positive workplace environment. Consider staff/organization needs, including (but not limited to): knowledge dissemination, team building, problem solving and rewards
Optimize interpharmacy communications Maintain interpharmacy communications to facilitate knowledge acquisition and social support
Optimize external communications with important stakeholder groups Maintain external communications with public health and governments; include pharmacists in policy making
Ensure direct communications from health authorities to pharmacy/pharmacy groups that are timely, consistent and pragmatic
Support and well-being Taking breaks Alter current pharmacy work culture, which discourages breaks; pharmacists are currently exempt from having a maximum number of work hours, rest periods and eating periods in labour board policies
Administer self-assessment tools to track mental health/burnout Deploy self-assessment tools to monitor and measure work-related stressors in pharmacy staff (e.g., Maslach Burnout Inventory)
Implement employee assistance programs Create formal employee assistance programs to support pharmacy staff who are experiencing mental health issues (or other health issues)
First, with respect to practice, we recommend increased staffing to account for increased services and demands as well as using pharmacy technicians to provide pharmacists with more time for clinical services.36 The integration of evidence-informed resources and tools (e.g., the CARD system)20-22 is recommended to assist with establishing efficient vaccination delivery and promoting a healthy pharmacy environment for both pharmacy staff and patients. This may require alterations to work processes, including vaccination setup and flow. Community pharmacists can benefit from scaling up technology in their practice. Software tools like booking systems can help manage workflows by reducing surges in appointments. Health authorities can facilitate development and use of technology that is interoperable with central registries and repositories and provide subsidies for implementation.
Second, the rate of compensation for community pharmacists should be commensurate with vaccination services provided and similar to compensation for other health professionals providing the same services. This includes compensation for vaccine administration and other vaccine-related services (such as providing advice regarding vaccinations) as well as any other cognitive services provided.
Third, enhancements in communication and collaboration are recommended across stakeholder groups, including the inclusion of pharmacists in policy-making decisions. Direct communications from health authorities and government should be timely, consistent and pragmatic.
Finally, we recommend efforts to integrate tools and processes that provide staff support and well-being. There is a need to alter a long-standing pharmacy work culture of “no breaks” that is perpetuated by inadequate labour board policies.37 Surveys are recommended for staff to monitor work-related stress so that it can be identified and addressed before it becomes harmful and causes individuals to exit the workplace. Formal employee assistance programs are also recommended to address arising mental health issues as well as other health-related concerns.
There are several strengths to this study. The qualitative methods used allowed for an in-depth understanding of pharmacists’ perceptions. Virtual interviews improved the feasibility of participation and allowed for inclusion of pharmacists from across the country with different characteristics, ensuring that a wide range of experiences and perspectives were described. Involvement of multiple coders also increased the credibility of the findings.
This study has several limitations worthy of discussion. While participants from all 10 provinces were included, there was a lack of representation for the 3 territories, and some perspectives may not have been captured. In addition, the perspectives of pharmacists may have changed since participating in the study due to ongoing pandemic-related factors. The continual outbreaks and lockdowns, coupled with COVID-19 testing, administration of booster COVID-19 vaccines and vaccination of children, however, suggests that work-related stressors have been ongoing.
In summary, most community pharmacists demonstrated their ability to adapt to the addition of COVID-19 vaccination services. This added responsibility, however, created a work environment that negatively affected resiliency and threatens the sustainability of broadened community pharmacy-based vaccination services. We recommend examining current approaches to pharmacy service delivery that will allow pharmacists to fully embrace and excel in opportunities to lead vaccination delivery, as well as other emerging services, in order to address the evolving health needs of society, including the potential next pandemic, while also maintaining a healthy workforce.
Funding: Support for this project was received from the Public Health Agency of Canada Immunization Partnership Fund (1921-HQ-000220) and a Canadian Institutes of Health Research Foundation Grant (FRN 159905) awarded to A. Taddio. The funding agencies had no input into the study.
A. Taddio reports a University of Toronto Section 9 Trademark No. 924835 for CARD™.
ORCID iDs: Susan Bowles https://orcid.org/0000-0003-0821-3222
Lucie M. Bucci https://orcid.org/0000-0003-2713-0975
Anna Taddio https://orcid.org/0000-0003-4432-8975
==== Refs
References
1. World Health Organization. WHO director-general’s opening remarks at the media briefing on COVID-19. March 11, 2020. Available: https://www.who.int/director-general/speeches/detail/who-director-general-s-opening-remarks-at-the-media-briefing-on-covid-19—11-march-2020 (accessed Apr. 29, 2022).
2. Canadian Broadcasting Corporation. Timeline of COVID-19 cases across Canada. Available: https://www.cbc.ca/news/health/canada-coronavirus-timeline-1.5482310 (accessed Apr. 29, 2022).
3. Government of Canada. COVID-19 daily epidemiology update. Available: https://health-infobase.canada.ca/covid-19/epidemiological-summary-covid-19-cases.html (accessed Apr. 29, 2022).
4. Canadian Pharmacists Association. Pharmacists’ scope of practice during COVID-19 (2021). Available: https://www.pharmacists.ca/cpha-ca/function/utilities/pdf-server.cfm?thefile=/cpha-on-the-issues/CovidSOP.pdf (accessed Apr. 29, 2022).
5. Pharmacy Association of Nova Scotia. Healthcare services available from pharmacists. Available: https://novascotia.ca/dhw/pharmacare/healthcare-services.asp (accessed Apr. 29, 2022).
6. Neighbourhood Pharmacy Gazette. Community pharmacy & COVID vaccinations. 2022. Available: https://neighbourhoodpharmacies.ca/sites/default/files/2022-03/NPG-Spring2022-Issue-8.pdf (accessed Jun. 17, 2022).
7. Munn LT Liu TL Swick M , et al . Original research: well-being and resilience among health care workers during the COVID-19 pandemic: a cross-sectional study. Am J Nurs 2021;121 (8 ):24-34.
8. Vanhaecht K Seys D Bruyneel L , et al . COVID-19 is having a destructive impact on health-care workers’ mental well-being. Int J Qual Health Care 2021;33 (1 ):1-6.33432980
9. Galanis P Vraka I Fragkou D Bilali A Kaitelidou D . Nurses’ burnout and associated risk factors during the COVID-19 pandemic: a systematic review and meta-analysis. J Adv Nurs 2021;77 (8 ):3286-302.
10. Amanullah S Shankar RR . The impact of COVID-19 on physician burnout globally: a review. Healthcare 2020;8 (4 ).
11. Langran C Mantzourani E Hughes L Hall K Willis S . “I’m at breaking point”; exploring pharmacists’ resilience, coping and burnout during the COVID-19 pandemic. Explor Res Clin Soc Pharm. 2022;5 :100104.35072149
12. Johnston K O’Reilly CL Scholz B Georgousopoulou EN Mitchell I . Burnout and the challenges facing pharmacists during COVID-19: results of a national survey. Int J Clin Pharm 2021;43 (3 ):716.33851288
13. Carpenter DM Hastings T Westrick S , et al . Rural community pharmacists’ ability and interest in administering COVID-19 vaccines in the Southern United States. J Am Pharm Assoc 2022;62 (4 ):1379-83.
14. Turcu-Stiolica A Kamusheva M Bogdan M , et al . Pharmacist’s perspectives on administering a COVID-19 vaccine in community pharmacies in four Balkan countries. Front Public Health 2021;9 :1822.
15. Gudzak V Logeman C Freedman T Bucci L Taddio A . Community pharmacists’ perceptions of the CARD (Comfort Ask Relax Distract) system for use during vaccinations. Can Pharm J (Ott) 2023;156 (suppl):xx-xx.
16. Taddio A Gerges S Rocchi M Gudzak V Ilersich A . Perceptions of pharmacy technician students of the CARD (Comfort Ask Relax Distract) e-module introduced as part of vaccine injection training. Can Pharm J (Ott) 2023;156 (suppl):xx-xx.
17. Taddio A Rocchi M Shea C , et al . CARD (Comfort Ask Relax Distract) implementation guide for community pharmacy. Can Pharm J (Ott) 2023;156 (suppl):xx-xx.
18. Taddio A Morrison J Gudzak V , et al . CARD (Comfort Ask Relax Distract) for community pharmacy vaccinations in children: effect on stress-related responses and satisfaction with vaccination. Can Pharm J (Ott) 2023;156 (suppl):xx-xx.
19. Taddio A Morrison J Gudzak V , et al . Integration of CARD (Comfort Ask Relax Distract) for COVID-19 community pharmacy vaccination in 5-11-year old children: effect on implementation outcomes. Can Pharm J (Ott) 2023;156 (suppl):xx-xx.
20. Tetui M Grindrod K Waite N VanderDoes J Taddio A . Integrating the CARD (Comfort Ask Relax Distract) system in a mass vaccination clinic to improve the experience of individuals during COVID-19 vaccination: a pre-post implementation study. Hum Vaccin Immunother 2022;18 (5 ):2089500.35723609
21. Taddio A Gudzak V Jantzi M , et al . Impact of the CARD (Comfort Ask Relax Distract) system on school-based vaccinations: a cluster randomized trial. Vaccine 2022;40 (19 ):2802-9.
22. Freedman T Taddio A Alderman L , et al . The CARDTM system for improving the vaccination experience at school: results of a small-scale implementation project on student symptoms. Paediatr Child Health (Canada). 2019;24 :S42-53.
23. Gerges S Peter E Bowles SK , et al . Pharmacists as vaccinators: An analysis of their experiences and perceptions of their new role. Hum Vaccin Immunother 2018;14 (2 ):471-7.
24. Hsieh HF Shannon SE . Three approaches to qualitative content analysis. Qual Health Res 2005;15 (9 ):1277-88.
25. Leiter M Maslach C . A mediation model of job burnout. Cheltenham (UK): Edward Elgar; 2005.
26. Association of Faculties of Pharmacy of Canada. AFPC educational outcomes for first professional degree programs in pharmacy in Canada. Ontario (Canada): Association of Faculties of Pharmacy of Canada; 2017.
27. Lee DH Watson KE al Hamarneh YN . Impact of COVID-19 on frontline pharmacists’ roles and services in Canada: the INSPIRE survey. Can Pharm J (Ott) 2021;15 :368-73.
28. Jones AM Clark JS Mohammad RA . Burnout and secondary traumatic stress in health-system pharmacists during the COVID-19 pandemic. Am J Health Syst Pharm. 2021;78 (9 ):818-24.
29. Lam W Mey A King MA Woods P . The nature of conflict in community pharmacy—a pilot study of pharmacists’ experiences during the COVID-19 pandemic. Res Soc Admin Pharm 2022;18 (7 ):3164-73.
30. Isenor JE Alia TA Killen JL , et al . Impact of pharmacists as immunizers on influenza vaccination. Hum Vaccin Immunother 2016;12 (5 ):1225-8.
31. National Association of Pharmacy Regulatory Authorities. Pharmacy technicians scope of practice in Canadian jurisdictions. 2022. Available: https://www.napra.ca/pharmacy-technicians/pharmacy-technicians-scope-practice-canadian-jurisdictions (accessed Jun. 22, 2022).
32. Nova Scotia College of Pharmacists. Standards of practice: drug administration. 2022. Available: https://www.nspharmacists.ca/wp-content/uploads/2022/06/SOP_DrugAdministration_April_2022.pdf
33. Gerges S Peter E Bowles SK , et al . Pharmacists as vaccinators: An analysis of their experiences and perceptions of their new role pharmacists as vaccinators: an analysis of their experiences and perceptions of their new role. Hum Vaccin Immunother 2017;14 (2 ):471-7.
34. Ontario College of Pharmacists. Minor ailments. 2022. Available: https://www.ocpinfo.com/practice-education/expanded-scope-of-practice/minor-ailments/ (accessed Jun. 29, 2022).
35. Abacus Data. The Canadian Pharmacy Mental Health and Workforce Wellness Survey. 2022. Available: https://www.pharmacists.ca/cpha-ca/assets/File/cpha-on-the-issues/CPhA_Mental_Health_Survey_2022_EN.pdf (accessed Jun. Jun 14, 2022).
36. Gyse S Watson K Tsuyuki R . Appropriate staffing for pharmacists’ full scope of practice. Can Pharm J (Ott) 2022;155 (3 ):136-8.
37. Government of Ontario. Industries and jobs with exemptions or special rules. 2017. Available: https://www.ontario.ca/document/industries-and-jobs-exemptions-or-special-rules/ems-healthcare-and-health-professionals#section-21 (accessed Jun. 22, 2022).
| 0 | PMC9703022 | NO-CC CODE | 2022-11-29 23:21:06 | no | Can Pharm J (Ott). 2022 Nov 25;:17151635221136552 | utf-8 | Can Pharm J (Ott) | 2,022 | 10.1177/17151635221136552 | oa_other |
==== Front
Omega (Westport)
Omega (Westport)
spome
OME
Omega
0030-2228
1541-3764
SAGE Publications Sage CA: Los Angeles, CA
10.1177_00302228221141126
10.1177/00302228221141126
Original Manuscript
The Effectiveness of Integrated Group Therapy on Prolonged Grief Disorder of Bereaved People from COVID-19 Randomized Controlled Trial
Bardideh Fatemeh 1
Jarareh Jamshid 2
Mofrad Mohammad 3
https://orcid.org/0000-0002-8107-9894
Bardideh Kosar 1
1 Department of Counselling and Psychology, 68106 Islamic Azad University , Kish International, Kish, Iran
2 Department of Teacher Training, 121571 Shahid Rajaee University , Tehran, Iran
3 Department of Psychology, Khayyam Institute of Higher Education , Mashhad, Iran
Kosar Bardideh, Department of Counselling and Psychology, Islamic Azad University, Kish International Branch, Free Zone Organization Square, Senaii Blvd., Kish Island, Kish 7941775883, Iran. Email: [email protected]
24 11 2022
24 11 2022
00302228221141126© The Author(s) 2022
2022
SAGE Publications
This article is made available via the PMC Open Access Subset for unrestricted re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the COVID-19 pandemic or until permissions are revoked in writing. Upon expiration of these permissions, PMC is granted a perpetual license to make this article available via PMC and Europe PMC, consistent with existing copyright protections.
This study aimed to evaluate the integrated cognitive-behavioral group therapy and Gestalt empty chair technique on bereaved individuals with COVID-19-caused PGD (prolonged grief disease). Thirty-six patients with PGD resultant from COVID-19 were randomly assigned intervention and control groups. The intervention group underwent 16 90-minute integrated group therapy sessions twice a week. Both groups completed the BDI II depression, NAI anger, and GASP guilt scale before, after, and 2 months after the study’s conclusion. The intervention and control groups significantly differed in the depression, anger, and guilt indices after the therapeutic intervention (p < .001). This difference remained in the follow-up phase. Integrated group therapy in treating could help with some of the symptoms of PGD resulting from the corona-caused loss of loved ones. This reduction in symptoms was also stable over time.
integrated group therapy
prolonged grief disease
complicated grief
COVID-19
edited-statecorrected-proof
typesetterts10
==== Body
pmcIntroduction
The coronavirus has been the biggest pandemic since the Spanish influenza outbreak in 1918. Almost 6,500,000 people have died of this virus from 2019 to October 2022 (“Coronavirus 2019 Reported Cases and Deaths,” (Worldometer, 2022).
In America, it has been shown that every corona-resultant death leads to the grief of nine family members (Verdery et al., 2020). This statistic can escalate due to the intimacy of the family members in some eastern countries and the bereavement of intimate friends (Murphy, 2008). Accordingly, we can assert that this disease has made at least 58 million people suffer from corona-caused mourning. This number of grieving individuals can bring about a serious challenge to mental health worldwide. The management of these people is one of the most crucial responsibilities of psychologists.
In the past, grief and mourning were used interchangeably. However, today, with research in this field, grief is used as a concept to describe cognitive and emotional reactions, changes in performance, and behavior to the loss of a person (Schneider, 1980). In normal grief, at the beginning of loss, people usually experience extreme sadness, unfamiliar feelings, preoccupation with thoughts and memories of the deceased, difficulty concentrating, and a lack of interest in people and daily activities. Nevertheless, after a while (depending on the nature of the loss), the wounds start to heal, and the bereaved person finds a way to return to normal life. In contrast, prolonged grief is unresolved or traumatic and impairs the grieving person’s social functioning. People’s daily and social activities are disrupted by the thoughts and memories of the deceased, even after years (De Stefano et al., 2021). Recently, a new classification of grief called “Prolonged Grief Disorder” was added to the DSM-5 and ICD-11 to describe continuous and pervasive grief with longing and constant mental preoccupation for the deceased person (Szuhany et al., 2021). Based on the ICD-11 classification, PGD is usually accompanied by severe distress with feelings such as sadness, guilt, anger, denial, blame, difficulty accepting death, feeling like a part of oneself has been lost, inability to experience a positive mood, emotional numbness, and difficulty engaging with emotions. Also, PGD is usually manifested by the impairment of social activities, which goes beyond the social, cultural, or religious norms and damages the individual’s daily tasks (Eisma et al., 2020).
While PGD has been classified in ICD-11 and DSM-5TR, there are several differences between these classifications. The duration criteria for ICD-11 and DSM-5TR for the PGD are 6 months and 12 months, respectively. Also, while the DSM-5TR has 10 diagnosis criteria, the ICD-11 has two more criteria (denial and the inability to experience a positive mood). The feelings of despair, guilt, and anger are consistent in both the ICD-11 and DSM-5TR classifications. (Boelen, 2021; Eisma et al., 2020).
In their systematic reviews, Lundorf et al. and Dielantik et al. observed that the probability of prolonged grief is 9% among bereaved adults and 49% among individuals mourning in the face of unnormal deaths. However, higher statistics have been reported in non-western countries (Djelantik et al., 2020; Lundorff et al., 2017). Furthermore, the prolonged grief of the grieving adults due to COVID-19 has been reported at 37.8%, and there were no differences between the symptoms of individuals whose loved ones had died of the coronavirus longer or shorter than the past 6months (Tang & Xiang, 2021).
Furthermore, there is a direct relationship between guilt and an individual’s degree of grieving. In particular, concerning the grief arising from loved ones’ death due to COVID-19, the inability to take part in funerals, and the emerging feeling of guilt can be the primary factors giving rise to PGD in these people (Diolaiuti et al., 2021).
PGD could lead to avoidance behaviors, lack of control of emotions, and loss of social relationships (Stroebe et al., 2007). Prolonged grief in the long term could lead to negative outcomes that lower the life expectancy of the grieving person (Bowling, 1987; Song et al., 2019). The lack of treatment for PGD causes increased suicidal thoughts and activities, depression, and post-traumatic stress disorder (Latham & Prigerson, 2004). Suicidal thoughts and tendencies have been reported in 20–50% of PGD patients (Simon et al., 2007). Furthermore, prolonged grief disorder leads to physical and mental diseases like cancer, cardiovascular diseases, and substance abuse (Chen et al., 1999; Parisi et al., 2019; Stahl et al., 2016).
Based on Stroebe and Schut’s dual process of coping with bereavement, the person experiencing grief could go through periods of oscillation between confronting and avoiding the grief process due to the experience of loss or restoration-oriented stressors (Stroebe & Schut, 1999). Loss-oriented stressors refer to one’s feelings and experiences about losing a deceased person, such as anger, nostalgia, and longing for the return of the deceased person. Restoration-oriented stressors pertain to activities that are done to distract from the sadness and despair caused by bereavement and to deal with the stress and anxiety caused by the experience of loss (Fiore, 2021). Furthermore, the grief-to-personal growth model explains that the essence of grief remains the same, and it is the person and the life and new experiences that are added to it. Based on these models, successful treatment of grief includes gradually and continuously detaching from the decedent and building new relationships with others (Hogan & Schmidt, 2002). However, the novel treatment designs do not focus on reaching acceptance and terminating relationships with the deceased but rather on fitting feelings, awareness, and prior experiences with new realities associated with the loss (Eisma & Stroebe, 2021).
Different therapeutic methods, including CGT (Cognitive Grief Therapy) (Glickman et al., 2016), CBT (Cognitive Behavioral Therapy) (Breen et al., 2022), support group (Robinson & Pond, 2019), Gestalt (Seen et al., 2021), and pharmacotherapy are used for PGD treatment (Gang et al., 2021). It has been observed that the CBT treatment has a moderate and statistically significant effect on the alleviation of symptoms related to prolonged grief, such as anger, a feeling of guilt, and depression (Eisma & Stroebe, 2021; Szuhany et al., 2021). Individual and collective psychological therapies have been influential in grief treatment, and these effects stay until the follow-up phase (Wittouck et al., 2011).
Research has unveiled that group therapy can reduce the effects of undesirable experiences against the bereavement phenomenon rooted in the death of loved ones and, contrary to individual therapy, can impede the extensiveness of the social isolation, anger, and anxiety stemming from the death of these people (Para, 2009; Supiano et al., 2021).
There are different factors for accepting and adapting to the death of loved ones. One of the most significant factors constitutes the thoughts and beliefs of the person toward the death conditions of the decedent, feelings toward the decedent, relationships with and closeness to the decedent, and the attitudes of the person toward themself (Dolan et al., 2022). CBT is a method that can identify inefficient thoughts and feelings of people and examine and change them. Likewise, the embedded behavioral techniques of CBT could alter the lifestyle of grieving patients and therefore introduce restoration-oriented stressors in their daily life and help them complete their coping process.
Gestalt therapy, a humanistic-existential form of psychotherapy, is a grief treatment approach and emphasizes the personal responsibility and present experience of the person and the client-therapist relationships (Seen et al., 2021). Among the Gestalt therapy techniques, we can refer to the empty chair, which allows clients to discuss their blocked feelings and attitudes. This technique is used to help clients and reach unsolved feelings that make individuals experience difficulties (Rosner et al., 2011, 2015). Expressing suppressed feelings and examining individuals’ unmet needs can help them correctly perceive their feelings and behaviors (Greenberg et al., 2008; Roulston et al., 2018). Therefore, integrating the Gestalt empty chair technique and cognitive-behavioral interventions can be a helpful therapy for grieving individuals to express and alter their feelings. Also, Gestalt therapy has been shown to help with the personal growth of people, which, based on the grief-to-personal growth model, could help patients suffering from PGD cope with their grief (Leung et al., 2013). Combining CBT with Gestalt therapy will help patients change their destructive feelings and emotions towards the decedent and themselves by giving them a chance to say goodbye to the deceased (loss-oriented stressors). Also, this method alters their lifestyles using new behavioral habits, self-care, and personal growth (restoration-oriented stressors).
Numerous integrated therapies have examined prolonged grief; however, no study has so far, and some even addressed the integration of the group CBT and Gestalt empty chair has integrated some parts of Gestalt therapy, like the imaginal conversation, in their treatment method (Iglewicz et al., 2020). However, no integrated group therapy method has been evaluated on PGD or grief caused by COVID. Thus, the present research aimed to investigate the effectiveness of the integrated group CBT and Gestalt empty chair technique on individuals diagnosed with PGD and bereaved by the corona-caused death of their loved ones.
Methods
The inclusion criteria were the acquisition of >102 scores on the Grief Experience Questionnaire (GEQ-34) of Barrett and Scott (Barrett & Scott, 1989), the age range of 18–50, and the pass of >6 months from the corona-caused death of a loved person.
The exclusion criteria were substance abuse, simultaneously receiving psychiatric or pharmaceutical therapies, and possessing psychosis histories or symptoms.
Participants
The sample size for our study was calculated using based on the results of the BDI-II score of the study by Lacasta-Reverte (Lacasta-Reverte & Cruzado, 2021) and type I error/α = 0.05 and Type II error/β = 0.2: N=2×(1.96+0.8453.74)2×5.432=32.7. The sample size was calculated as 33 patients.
Participants included 18–50-year-old bereaved adults losing their loved ones due to the coronavirus. These people whose PGDs were diagnosed by psychologists or psychiatrists in their first session and were referred to us via six clinics in Mashhad city. None of these patients had received any psychiatric or psychological treatments between the death of their loved ones and the start of our study. We interviewed the selected individuals, recorded their medical histories, and administered the GEQ-34 test. The GEQ-34 (Grief Experience Questionnaire) is a self-reported measure of bereavement, consisting of subscales such as guilt, trying to justify and cope, physical reactions, feelings of abandonment, personal or other judgments, and embarrassment. The GEQ-34 test score above 102 was chosen as a measure of prolonged grief disorder based on the study by Treml et al. (Treml et al., 2021).
From the patients referred to us, the individuals scoring >102 on the GEQ test and experiencing pathological grief symptoms for more than 6 months (based on the ICD-11 classification) were included in our study. Out of 58 patients, 8 were excluded due to using psychological drugs, 11 due to not acquiring >102 scores, and 3 due to the non-fit of the intervention hours with their programs. The remaining 36 participants filled out the consent form (nobody refused to sign the form), and, finally, 36 individuals were entered into the study. Table 1 provides the demographic characteristics of the participants. If during the study, the participants developed any psychosis symptoms, suicidal tendencies, or need to use any psychological drugs, they would be excluded from the study. They would be referred to their original clinician for an emergency visit.Table 1. Demographic Properties of Sample Group.
Measure Number of Participants Percentage of Participants for Each Group
Control Experimental Control Experimental
Participants 18 18 100 100
Age
18–30 4 3 22.23 16.67
30–40 8 9 44.44 50.00
40–50 6 6 33.33 33.33
Sex
Male 8 8 44.45 44.45
Female 10 10 55.55 55.55
Academic status
Elementary school 1 2 5.55 11.11
High school 2 1 11.11 5.55
Associate degree 5 5 27.78 27.77
Bachelor degree 6 7 33.33 38.89
Master degree 3 2 16.66 11.11
Doctorate degree 1 1 5.55 5.55
Identity of deceased
Spouse/partner 2 3 11.11 16.67
Parent 9 7 50.00 38.88
Sibling 2 3 11.11 16.66
Aunt/uncle 3 2 16.67 11.11
Grandparent 1 2 5.55 11.11
Close friend 1 1 5.55 5.55
Other 0 0 0 0
Time since death (months) [mean ± SD]
Control group Experimental group
13.61 ± 15.34 12.73 ± 14.87
The participants were assigned to the experimental and control groups (18 per group) based on simple randomization through the www.randomizer.org site. Then, the experimental group was divided into two therapeutic groups (9 per group) to provide sufficient time for treatment. Both 9-subject groups (a total of 18 subjects) underwent a single treatment, and the control group received no intervention. Before the intervention, all 36 participants filled out Beck’s Depression Inventory-II, Novaco Anger Inventory (NAI), and Cohen’s Guilt and Shame Proneness (GASP) scale. Afterward, both 9-subject groups received 16 90-minute sessions of integrated therapy (Kim Paleg Grief Protocol, 2015 (Berghuis & Paleg, 2015)) twice a week. After the intervention, the experimental and control groups took the posttest and were followed up 2 months later (Figure 1). The patients in the control group also underwent integrated treatment after the study’s conclusion.Figure 1. Flowchart illustarting the study.
Instruments
Beck’s Depression Inventory
Beck’s Depression Inventory, built in 1961, consists of 21 questions comparing somatic, behavioral, and cognitive symptoms of depression and is scored based on a 5-point Likert scale (0–4) (Beck et al., 1961). It measures severe to mild depression. Its minimum and maximum scores, Cronbach alpha coefficient, and test-retest reliability equal 0, 63, 0.78, and 0.73, respectively. The validity and reliability of this questionnaire have been estimated at 0.70 and 0.77 in the Iranian context (García-Batista et al., 2018).
Novaco Anger Inventory
This inventory, built by Novaco in 1986, is a self-report scale consisting of 30 items that measure anger and aggression. It is scored based on a 5-point Likert scale with zero and 100 as its minimum and maximum scores. Individuals acquiring scores above the mean possess higher levels of aggression. The validity and reliability of this scale have been reported at 0.86 and 0.96, and its Cronbach alpha coefficient and test-retest reliability equal 0.86 and 0.73 (Jang, 2019).
Cohen’s Guilt and Shame Proneness scale
The Guilt and Shame Proneness (GASP) scale was designed by Cohen and Wolf in 2011. It is a self-report scale with 16 items identifying shame and guilt in two subscales of negative self-evaluation and withdrawal behaviors. The scale is scored based on a 5-point Likert scale, and its Cronbach alpha coefficient for reliability has been reported at 0.61 and 0.71 (Young et al., 2021).
Therapy
The grief-treating group therapy was implemented based on the group therapy protocol of Kim Paleg (2015), which was an integration of the cognitive-behavioral group therapy and the Gestalt empty chair technique. It was provided in 16 90-minute sessions held twice a week for 8 weeks. The group sessions consisted of nine members who participated in activities collectively. The treatment for both 9-person intervention groups was overseen by the same three-clinician team with at least 4 years of grief counseling experience. F.B, with a doctorate in counseling, was the facilitator; K.B, with a doctorate in counseling, acted as the assistant, and J.J, with a doctorate in psychology, was the supervisor of the group therapy. Table 2 displays the descriptions of the sessions.Table 2. The Curriculum for Integrated Group Therapy Sessions.
Sessions Content
1 Share the story of the loss, including who was lost, when, and how the loss occurred
2 Describe the impact of the loss on work, family, and relationships
3 Verbalize an increased understanding of the components of grief as parts of a process that must be experienced in order to heal
4 Identify personal grief coping strategies, including the use of substances, and note those that may have interfered with the grieving process
5 Accept the need for antidepressant medication and follow through on a referral to a physician for an evaluation
6 Demonstrate the ability to ask for help in group and with significant others
7 Write a letter to the deceased person saying goodbye and expressing all the feelings experienced in the aftermath of the loss
8 Verbalize the impact of the changed identity resulting from the loss
9 Articulate a realistic picture of the lost person—both positive and negative—and of the relationship with that person, and identify ways of remembering the special qualities
10 Report an increase in self-nurturing activities
11 Develop a plan or ceremony to facilitate memorializing the lost person
12 Verbalize self-care plans to cope with anniversary reactions
13 Read books on the grief process and discuss their impact
14 Verbalize an acceptance of the unique style of grieving of others close to the deceased
15 Verbalize the desire to and beginning of the process of letting go of bitter blame for the loss of the significant other
16 Verbalize a resolution of feelings of guilt or regret over actions toward the lost loved one. (Using empty chair technique to facilitate members’ saying goodbye to the deceased and saying things that were left unsaid or asking for forgiveness for actions regretted)
Ethical Considerations
Before the study, the researchers held a session for the subjects and explained the research procedure and its ethical issues, such as voluntary intervention leave. Then, the participants completed the informed consent form and were ensured that their identities would be confidential before, during, and after the release of the results. This study was approved by the ethics committee of the Iranian National Institute for Medical Research Development (NIMAD) with the registration number 4002721.
Results
Table 3 displays the means and standard deviations (SD) of the examined variables in the experimental and control groups in the pre-intervention, post-intervention, and follow-up phases. According to the results of this table, the intervention group’s mean scores in the anger, depression, and guilt indices are more degressive than the control group, and this degression continues until the follow-up phase. The ANCOVA test was used for testing the hypotheses. However, before running the test, the researchers investigated its basic assumptions using the Kolmogorov-Smirnov test for normality of data distribution, Levene’s test of homogeneity, box plots for the absence of irrelative variables, Box M test for covariance, matrice’s homogeneity, and linear regression for a linear relationship between the covariate and dependent variable.Table 3. Mean and Standard Deviation for Depression, Anger and Shame.
Experimental Group Control Group
Pre-test Post-test Follow-up Pre-test Post-test Follow-up
BDI-II M 42.39 35.44 32.94 34.94 32.50 33.00
SD 9.54 10.47 10.22 12.64 10.03 10.42
NAI M 73.44 51.88 48.88 68.88 66.11 66.77
SD 10.77 17.46 17.06 11.08 13.93 15.02
GASP M 24.67 16.00 15.00 18.44 15.44 16.55
SD 11.89 12.44 10.59 8.36 9.73 9.51
Note. BDI-II = Beck Depression Inventory-Second Edition; NAI = novaco anger scale; GASP = guilt and shame proneness scale test.
After the realization of the assumptions, the ANCOVA test was run for the comparison of the experimental and control groups in the posttest and follow-up phases (Tables 4 and 5).Table 4. Analysis of Covariance Results for Depression, Anger and Shame Scores From the Post-Intervention Stage (Between Groups).
SS df MS F Sig Es
BDI-II 3666.80 1 3666.80 208.75 .00 0.69
NAI 5922.39 1 5922.39 76.19 .00 0.88
GASP 3178.27 1 3178.27 261.76 .000 0.86
Note. BDI-II = Beck Depression Inventory-Second Edition; NAI = novaco anger scale; GASP = guilt and shame proneness scale test; df = degrees of freedom; ES = Eta squared; MS = mean square; SS = sum of squares.
Table 5. Analysis of Covariance Results for Depression, Anger and Shame Scores in Follow-up Phase.
SS df MS F Sig Es
BDI-II 2382.684 1 2382.684 63.295 .000 .657
NAI 4775.315 1 4775.315 39.283 .000 .543
GASP 2830.275 1 2830.275 150.602 .000 .820
Note. BDI-II = Beck Depression Inventory-Second Edition; NAI = novaco anger scale; GASP = guilt and shame proneness scale test; df = degrees of freedom; ES = Eta squared; MS = mean square; SS = sum of squares.
The results of Table 4 show that the intervention and control groups are significantly different in the depression (p < 0.00, f = 208.75), anger (p < .001, F = 76.19), and Shame (p < .001, F = 261.76) indices. Likewise, according to the results of Table 5, the difference between the two groups is also significant in the follow-up phase in the depression (p < .001, F = 63.29), anger (p < .001, F = 39.28), and guilt (p < .001, F = 150.60).
Discussion
After treatment, the intervention group showed lower scores for anger, depression and guilt and these results remained until the follow-up. These results show that integrated group therapy could be a promising approach for lessening the PGD symptoms, such as depression, anger, and feeling of guilt among bereaved individuals who have lost their loved ones because of the Corona Virus. Concerning the alleviation of depression symptoms, our findings are in line with the results of studies conducted by other on the subject (Boelen et al., 2007; Bryant et al., 2014; Bryant et al., 2017; Roberts et al., 2019; Rosner et al., 2015). Although the main reasons for these changes are unknown, we can claim that those individuals get acquainted with one another in integrated group therapy, learn effective social skills, and test them in the group. Group therapy aims to make individuals come to a common experience of grief, reduce the isolation stemming from their avoidance behavior and limited social support, and discover opportunities for exchanging support (Schuster et al., 2017). Likewise, in integrated group therapy, individuals decrease their depression by doing daily enjoyable activities, performing respiratory exercises, receiving cognitive restructuring techniques, and challenging and changing negative coping beliefs (107). Challenging group members, repeating exercises in the group, and witnessing positive impacts on others encourage them to continue doing these activities. In a study, Thimm et al. showed that group therapy decreased depression symptoms by 45% (Thimm & Antonsen, 2014). These effects could also justify 75% of the changes in the follow-up phase. This study revealed that integrated group therapy reduced anger and guilt in bereaved individuals with PGD arising from the corona-caused death of loved ones. This outcome may be due to employing different techniques, such as writing letters to decedents, cognitive restructuring, and the empty chair of Gestalt therapy (Gupta, 2018).
These techniques make individuals recognize and express their hidden and suppressed anger toward the decedent or themselves. Furthermore, using the empty chair, writing a letter to the decedent, and expressing unsaid words and regrets to the decedent decrease anger and increase grief acceptance in bereaved people (Tsvieli & Diamond, 2018).
Employing cognitive restructuring techniques and identifying and challenging cognitive distortions and inefficient basic beliefs can reduce the feeling of guilt resulting from inefficient thoughts (Meichsner et al., 2020). The impact of our intervention on the withdrawal of self and other blames may be rooted in the decline of negative cognitions and avoidance behaviors and can help individuals solve their family problems (de Groot et al., 2007). As observed in the study Arslan et al. conducted on normal bereaved families, possessing opportunities to participate in groups and receiving counseling assists individuals to notice that they have made no mistakes, and their feelings of shame and guilt considerably lessen (Şimşek Arslan & Buldukoğlu, 2019). Groups create atmospheres wherein the problems related to past relationships and insecure attachments can be expressed and evaluated and provide participants with opportunities to set goals and be responsive to their progress (Larsen et al., 2021). Participating in groups and challenging the thoughts of other group members make individuals aware of their inefficient thoughts and pave the way for them to challenge and change their own thoughts (Leiderman, 2019).
In our study, we assessed three negative symptoms (depression, anger, guilt) that are associated with prolonged grief disorder. While it was concluded that integrated therapy could help reduce these variables, prolonged grief disorder is a separate construct, and the effects of this integrated therapy on this construct were not directly measured (Rosner et al., 2011). This issue should be considered when reviewing the results of our study. Furthermore, there have been several different grief treatment protocols that have utilized both elements of CBT and Gestalt therapy. These treatment protocols have shown to be effective for treatment of prolonged grief disorder and so these probably be also effective for PGD arising from COVID loss.
Compared to past studies, the merits of this design include (1) Selecting the subjects with the help of skillful psychologists and psychiatrists, (2) Assigning the subjects into groups randomly, which could enormously prevent testing errors, (3) Dividing the participants into experimental and control groups to separate the effects of the intervention from spontaneous changes in the subjects and present more reliable results; and (4). Considering a follow-up phase besides the pretest and posttest to ensure the impacts of the intervention even 2 months after the experiment.
Some of the limitations of this research were a 2-month interval for administering the follow-up test and the fewness of subjects. It is suggested that future studies use larger samples and more extended follow-up periods.
Maybe the most significant limitation of our study was the use of depression, anger, and guilt inventories because of the lack of availability of better native-translated and validated questionnaires. These measurements only evaluate some aspects of the grief process and do not assess prolonged grief disorder as a construct as a whole. Also, these inventories only measure negative variables for the effects of integrated therapy. The use of only negative variables might introduce pathology-focused bias, which is somewhat in opposition to the spirit of the dual process of coping with bereavement models and the Gestalt therapy method. We recommend that future studies use more comprehensive and positive measurement methods like Personal Growth Initiative Scale or Prolonged Grief Disorder (PG-13) questionnaire.
Furthermore, the RCT design of our study might introduce its limitations. While using a randomized clinical trial design could eliminate the selection bias and decrease the effects of confounding factors, the generalizability of the results might be more limited than varied study designs. Moreover, finally, because the control group patients were treated after the study’s conclusion, no long-term assessment of the effects of integrated therapy on the PGD of the subjects included in our study could be performed.
Conclusion
This study provided evidence that integrated group therapy could help with some of the symptoms of PGD resulting from the corona-caused loss of loved ones. The increasing rise of the disease statistics, besides its resultant death toll, has caused many people to experience prolonged grief derived from the coronavirus. Owing to the sudden and unexpected death of loved ones and the impossibility of mourning and grieving because of long-term quarantines, these individuals remain in the grief cycle and experience different feelings, such as grief, sorrow, helplessness, anger, anxiety, confusion, nervousness, and guilt. Hence, by applying several techniques, such as employing cognitive restructuring, deep abdominal breathing and relaxation, the empty chair, writing a letter to the decedent, and perceiving and identifying different emotions and physical states of the person, integrated group therapy can assist with accepting the death of a loved one. Moreover, groups are sources of inspiration and exemplification and increase inter-group intimacy and relationships that can overstep group boundaries. Recognizing appropriate coping strategies to manage challenging situations by applying the cognitive-behavioral and mentioned techniques can help decrease some of the depression, anger, and guilt in bereaved individuals with corona-caused PGD symptoms.
Still, further studies using integrated therapy with more encompassing inventories are needed to understand how these treatments could help patients with PGD arising from COVID.
Author Biographies
Jamshid Jarareh, Ph.D. is an assistant professor of psychological counseling at the Humanities School of Shahid Rajaee University. He is the director of the psychological clinic of Shahid Rajaee university and his research includes the studies on effects of group therapy on anxiety and anger, different methods patients use for coping with trauma, and the treatment of patients with PTSD.
Kosar Bardideh is a doctorate student of psychological counseling at the psychology and counseling school of Azad University. Her research focuses on cognitive behavioral therapy, Mindfulness-Based Cognitive Therapy, and group therapy for the treatment of anxiety-related disorders like nocturnal enuresis, PTSD, and grief-related disorders.
Fatemeh Bardideh is a doctorate student of psychological counseling at the psychology and counseling school of Azad University. Her research focuses on group CBT, existential therapy, and its effects on personal anxiety disorders and adult grief.
Majid Monfared is a master’s student of psychology and biostatistics at Khayam University. His research interests include panic disorders and the use of mixed-design models for evaluating psychological treatments.
ORCID iD
Kosar Bardideh https://orcid.org/0000-0002-8107-9894
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding: The author(s) received no financial support for the research, authorship, and/or publication of this article.
==== Refs
References
Barrett T. W. Scott T. B. (1989). Development of the grief experience questionnaire. Suicide & Life-threatening Behavior, 19 (2 ), 201–215. 10.1111/j.1943-278x.1989.tb01033.x 2749862
Beck A. T. Ward C. H. Mendelson M. Mock J. Erbaugh J. (1961). An inventory for measuring depression. Archives of General Psychiatry, 4 (6 ), 561–571. 10.1001/archpsyc.1961.01710120031004 13688369
Berghuis D. J. Paleg K. (2015). The group therapy treatment planner, with DSM-5 updates. John Wiley & Sons.
Boelen P. A. (2021). Symptoms of prolonged grief disorder as per DSM-5-TR, posttraumatic stress, and depression: Latent classes and correlations with anxious and depressive avoidance. Psychiatry Research, 302 , 114033. 10.1016/j.psychres.2021.114033 34134030
Boelen P. A. de Keijser J. van den Hout M. A. van den Bout J. (2007). Treatment of complicated grief: A comparison between cognitive-behavioral therapy and supportive counseling. Journal of Consulting and Clinical Psychology, 75 (2 ), 277–284. 10.1037/0022-006x.75.2.277 17469885
Bowling A. (1987). Mortality after bereavement: A review of the literature on survival periods and factors affecting survival. Social Science & Medicine, 24 (2 ), 117–124. 10.1016/0277-9536(87)90244-9 3551087
Breen L. Greene D. Rees C. Black A. Cawthorne M. Egan S. (2022). A systematic review and meta-analysis of the efficacy of grief interventions for anxiety and depression in young people. PsyArXiv. Center for Open Science. 10.31234/osf.io/gy9wv
Bryant R. A. Kenny L. Joscelyne A. Rawson N. Maccallum F. Cahill C. Hopwood S. Aderka I. Nickerson A. (2014). Treating prolonged grief disorder. JAMA Psychiatry, 71 (12 ), 1332. 10.1001/jamapsychiatry.2014.1600 25338187
Bryant R. A. Kenny L. Joscelyne A. Rawson N. Maccallum F. Cahill C. Hopwood S. Nickerson A. (2017). Treating prolonged grief disorder. The Journal of Clinical Psychiatry, 78 (9 ), 1363–1368. 10.4088/jcp.16m10729 28445631
Chen J. H. Bierhals A. J. Prigerson H. G. Kasl S. V. Mazure C. M. Jacobs S. (1999). Gender differences in the effects of bereavement-related psychological distress in health outcomes. Psychological Medicine, 29 (2 ), 367–380. 10.1017/s0033291798008137 10218927
de Groot M. de Keijser J. Neeleman J. Kerkhof A. Nolen W. Burger H. (2007). Cognitive behaviour therapy to prevent complicated grief among relatives and spouses bereaved by suicide: Cluster randomised controlled trial. BMJ (Clinical Research ed.), 334 (7601 ), 994–994. 10.1136/bmj.39161.457431.55
De Stefano R. Muscatello M. R. A. Bruno A. Cedro C. Mento C. Zoccali R. A. Pandolfo G. (2021). Complicated grief: A systematic review of the last 20 years. International Journal of Social Psychiatry, 67 (5 ), 492–499. 10.1177/0020764020960202 32972293
Diolaiuti F. Marazziti D. Beatino M. F. Mucci F. Pozza A. (2021). Impact and consequences of COVID-19 pandemic on complicated grief and persistent complex bereavement disorder. Psychiatry Research, 300 , 113916–113916. 10.1016/j.psychres.2021.113916 33836468
Djelantik A. A. A. M. J. Smid G. E. Mroz A. Kleber R. J. Boelen P. A. (2020). The prevalence of prolonged grief disorder in bereaved individuals following unnatural losses: Systematic review and meta regression analysis. Journal of Affective Disorders, 265 , 146–156. 10.1016/j.jad.2020.01.034fv 32090736
Dolan N. Grealish A. Tuohy T. Bright A. M. (2022). Are mindfulness-based interventions as effective as cognitive behavioral therapy in reducing symptoms of complicated perinatal grief? A systematic review. Journal of Midwifery & Women’s Health, 67 (2 ), 209–225. 10.1111/jmwh.13335
Eisma M. C. Rosner R. Comtesse H. (2020). ICD-11 prolonged grief disorder criteria: Turning challenges into opportunities with multiverse analyses. Frontiers in Psychiatry, 11 , 752. 10.3389/fpsyt.2020.00752 32848929
Eisma M. C. Stroebe M. S. (2021). Emotion regulatory strategies in complicated grief: A systematic review. Behavior Therapy, 52 (1 ), 234–249. 10.1016/j.beth.2020.04.004 33483120
Fiore J. (2021). A systematic review of the dual process model of coping with bereavement (1999-2016). Omega, 84 (2 ), 414–458. 10.1177/0030222819893139 31829782
Gang J. Kocsis J. Avery J. Maciejewski P. K. Prigerson H. G. (2021). Naltrexone treatment for prolonged grief disorder: Study protocol for a randomized, triple-blinded, placebo-controlled trial. Trials, 22 (1 ), 110–110. 10.1186/s13063-021-05044-8 33522931
García-Batista Z. E. Guerra-Peña K. Cano-Vindel A. Herrera-Martínez S. X. Medrano L. A. (2018). Validity and reliability of the Beck depression inventory (BDI-II) in general and hospital population of Dominican republic. Plos One, 13 (6 ), Article e0199750. 10.1371/journal.pone.0199750 29958268
Glickman K. Shear M. K. Wall M. M. (2016). Mediators of outcome in complicated grief treatment. Journal of Clinical Psychology, 73 (7 ), 817–828. 10.1002/jclp.22384 27755654
Greenberg L. J. Warwar S. H. Malcolm W. M. (2008). Differential effects of emotion-focused therapy and psychoeducation in facilitating forgiveness and letting go of emotional injuries. Journal of Counseling Psychology, 55 (2 ), 185–196. 10.1037/0022-0167.55.2.185
Gupta T. (2018). Psychological management of bereavement among adolescents: A case series. Journal of Indian Association for Child & Adolescent Mental Health, 14 (2 ), 117–127. 10.1177/0973134220180208
Hogan N. S. Schmidt L. A. (2002). Testing the grief to personal growth model using structural equation modeling. Death Studies, 26 (8 ), 615–634. 10.1080/07481180290088338 12243195
Iglewicz A. Shear M. K. Reynolds C. F., 3rd Simon N. Lebowitz B. Zisook S. (2020). Complicated grief therapy for clinicians: An evidence-based protocol for mental health practice. Depression and Anxiety, 37 (1 ), 90–98. 10.1002/da.22965 31622522
Jang E. (2019). The relation between temperament and anger response among prisoners: Comparison of reinforcement sensitivity theory and the psychobiological model of temperament and character. Heliyon, 5 (7 ), Article e02103. 10.1016/j.heliyon.2019.e02103 31372554
Lacasta-Reverte M. Cruzado J. A. (2021). Effectiveness of a cognitive-behavioral group therapy for complicated grief in family members of patients with cancer. Preprint. 10.21203/rs.3.rs-1137722/v1
Larsen L. H. Guldberg A. Kring V. (2021). A group therapy program for parentally bereaved young adults with grief complications: Rationale, method and case examples. OMEGA - Journal of Death and Dying. Advance online publication. 10.1177/0030222821997705
Latham A. E. Prigerson H. G. (2004). Suicidality and bereavement: Complicated grief as psychiatric disorder presenting greatest risk for suicidality. Suicide & Life-threatening Behavior, 34 (4 ), 350–362. 10.1521/suli.34.4.350.53737 15585457
Leiderman L. M. (2019). Psychodynamic group therapy with hispanic migrants: Interpersonal, relational constructs in treating complex trauma, dissociation, and enactments. International Journal of Group Psychotherapy, 70 (2 ), 162–182. 10.1080/00207284.2019.1686704
Leung G. S. Leung T. Y. Ng M. L. (2013). An outcome study of gestalt-oriented growth workshops. International Journal of Group Psychotherapy, 63 (1 ), 117–125. 10.1521/ijgp.2013.63.1.117 23252818
Lundorff M. Holmgren H. Zachariae R. Farver-Vestergaard I. O’Connor M. (2017). Prevalence of prolonged grief disorder in adult bereavement: A systematic review and meta-analysis. Journal of Affective Disorders, 212 , 138–149. 10.1016/j.jad.2017.01.030 28167398
Meichsner F. O’Connor M. Skritskaya N. Shear M. K. (2020). Grief before and after bereavement in the elderly: An approach to care. The American Journal of Geriatric Psychiatry, 28 (5 ), 560–569. 10.1016/j.jagp.2019.12.010 32037292
Murphy M. (2008). Variations in kinship networks across geographic and social space. Population and Development Review, 34 (1 ), 19–49. 10.1111/j.1728-4457.2008.00204.x
Para E. A. (2009). Group counseling for complicated grief: A literature review. Graduate Journal of Counseling Psychology, 1 (2 ), 10.
Parisi A. Sharma A. Howard M. O. Blank Wilson A. (2019). The relationship between substance misuse and complicated grief: A systematic review. Journal of Substance Abuse Treatment, 103 , 43–57. 10.1016/j.jsat.2019.05.012 31229191
Roberts K. E. Walsh L. E. Saracino R. M. Fogarty J. Coats T. Goldberg J. Prigerson H. Lichtenthal W. G. (2019). A systematic review of treatment options for grieving older adults. Current Treatment Options in Psychiatry, 6 (4 ), 422–449. 10.1007/s40501-019-00191-x 32754423
Robinson C. Pond D. R. (2019). Do online support groups for grief benefit the bereaved? Systematic review of the quantitative and qualitative literature. Computers in Human Behavior, 100 (3 ), 48–59. 10.1016/j.chb.2019.06.011
Rosner R. Bartl H. Pfoh G. Kotoučová M. Hagl M. (2015). Efficacy of an integrative CBT for prolonged grief disorder: A long-term follow-up. Journal of Affective Disorders, 183 , 106–112. 10.1016/j.jad.2015.04.051 26001670
Rosner R. Pfoh G. Kotoučová M. (2011). Treatment of complicated grief. European Journal of Psychotraumatology, 2 . (1 ). 10.3402/ejpt.v3402i3400.7995
Roulston A. Clarke M. J. Donnelly M. Candy B. McGaughey J. Keegan O. Duffy M. (2018). Psychological therapies for major depressive disorder and prolonged grief in bereaved adults. Cochrane Database of Systematic Reviews. 2018 (12 ), CD013237. 10.1002/14651858.cd013237
Schneider J. M. (1980). Clinically significant differences between grief, pathological grief, and depression. Patient Counselling and Health Education, 2 (4 ), 161–169. 10.1016/s0738-3991(80)80097-8 23847793
Schuster R. Sigl S. Berger T. Laireiter A.-R. (2017). Patients experiences of blended group therapy for depression: Fit and implications for the group setting. JMIR Publications Inc.
Seen A. T. H. Ahmad N. S. Khalid N. F. (2021). Effect of empty chair technique towards psychological well-being individual experiencing grief: A case study. Jurnal Penyelidikan Sains Sosial (JOSSR), 4 (10 ), 21–34.
Simon N. M. Shear K. M. Thompson E. H. Zalta A. K. Perlman C. Reynolds C. F. Frank E. Melhem N. M. Silowash R. (2007). The prevalence and correlates of psychiatric comorbidity in individuals with complicated grief. Comprehensive Psychiatry, 48 (5 ), 395–399. 10.1016/j.comppsych.2007.05.002 17707245
Şimşek Arslan B. Buldukoğlu K. (2019). Yasın aile üzerine etkilerini azaltmak i?çin uygulanan yas destek programları. Psikiyatride Guncel Yaklasimlar - Current Approaches in Psychiatry, 11 (3 ), 402–417. 10.18863/pgy.444297
Song J. Mailick M. R. Greenberg J. S. Floyd F. J. (2019). Mortality in parents after the death of a child. Social Science & Medicine, 239 , 112522. 10.1016/j.socscimed.2019.112522 31487567
Stahl S. T. Arnold A. M. Chen J. Y. Anderson S. Schulz R. (2016). Mortality after bereavement: The role of cardiovascular disease and depression. Psychosomatic Medicine, 78 (6 ), 697–703. 10.1097/psy.0000000000000317 26894326
Stroebe M. Schut H. (1999). The dual process model of coping with bereavement: Rationale and description. Death Studies, 23 (3 ), 197–224. 10.1080/074811899201046 10848151
Stroebe M. Schut H. Stroebe W. (2007). Health outcomes of bereavement. Lancet, 370 (9603 ), 1960–1973. 10.1016/s0140-6736(07)61816-9 18068517
Supiano K. P. Larsen P. Riley C. Hutton A. Iacob E. Walker A. M. Morrison S. J. (2021). Complicated grief group therapy for community-residing persons diagnosed with serious mental illness. Social Work in Mental Health, 19 (5 ), 381–402. 10.1080/15332985.2021.1940419
Szuhany K. L. Malgaroli M. Miron C. D. Simon N. M. (2021). Prolonged grief disorder: Course, diagnosis, assessment, and treatment. Focus, 19 (2 ), 161–172. 10.1176/appi.focus.20200052 34690579
Tang S. Xiang Z. (2021). Who suffered most after deaths due to COVID-19? Prevalence and correlates of prolonged grief disorder in COVID-19 related bereaved adults. Globalization and Health, 17 (1 ), 19–19. 10.1186/s12992-021-00669-5 33573673
Thimm J. C. Antonsen L. (2014). Effectiveness of cognitive behavioral group therapy for depression in routine practice. BMC Psychiatry, 14 , 292. 10.1186/s12888-014-0292-x 25330912
Treml J. Nagl M. Linde K. Kündiger C. Peterhänsel C. Kersting A. (2021). Efficacy of an internet-based cognitive-behavioural grief therapy for people bereaved by suicide: A randomized controlled trial. European Journal of Psychotraumatology, 12 (1 ), 1926650. 10.1080/20008198.2021.1926650 34992754
Tsvieli N. Diamond G. M. (2018). Therapist interventions and emotional processing in attachment-based family therapy for unresolved anger. Psychotherapy, 55 (3 ), 289–297. 10.1037/pst0000158 30179035
Verdery A. M. Smith-Greenaway E. Margolis R. Daw J. (2020). Tracking the reach of COVID-19 kin loss with a bereavement multiplier applied to the United States. Proceedings of the National Academy of Sciences of the United States of America, 117 (30 ), 17695–17701. 10.1073/pnas.2007476117 32651279
Wittouck C. Van Autreve S. De Jaegere E. Portzky G. van Heeringen K. (2011). The prevention and treatment of complicated grief: A meta-analysis. Clinical Psychology Review, 31 (1 ), 69–78. 10.1016/j.cpr.2010.09.005 21130937
Worldometer. (2022). Coronavirus, cases, reported, & deaths. https://www.worldometers.info/coronavirus/
Young I. F. Razavi P. Cohen T. R. Yang Q. Alabèrnia-Segura M. Sullivan D. (2021). A multidimensional approach to the relationship between individualism-collectivism and guilt and shame. Emotion, 21 (1 ), 108–122. 10.1037/emo0000689 31580090
| 0 | PMC9703023 | NO-CC CODE | 2022-11-29 23:21:06 | no | Omega (Westport). 2022 Nov 24;:00302228221141126 | utf-8 | Omega (Westport) | 2,022 | 10.1177/00302228221141126 | oa_other |
==== Front
Public Health Rep
Public Health Rep
PHR
spphr
Public Health Reports
0033-3549
1468-2877
SAGE Publications Sage CA: Los Angeles, CA
36062380
10.1177/00333549221120676
10.1177_00333549221120676
Public Health Evaluation
Moving the Needle: Association Between a Vaccination Reward Lottery and COVID-19 Vaccination Uptake in Louisiana
https://orcid.org/0000-0002-3554-656X
Wang Yin MA 1
Hernandez Julie PhD 2
Stoecker Charles PhD 1
1 Department of Health Policy and Management, Tulane University School of Public Health and Tropical Medicine, New Orleans, LA, USA
2 Department of International Health and Sustainable Development, Tulane University School of Public Health and Tropical Medicine, New Orleans, LA, USA
Charles Stoecker, PhD, Tulane University School of Public Health and Tropical Medicine, Department of Health Policy and Management, 1440 Canal St, New Orleans, LA 70112, USA. Email: [email protected]
5 9 2022
Jan-Feb 2023
138 1 6875
© The Author(s) 2022
2022
US Surgeon General’s Office
Objective:
On June 17, 2021, Louisiana launched a lottery campaign to reward residents who received a COVID-19 vaccination. We investigated the association between the lottery and vaccination uptake by characteristics of parishes.
Methods:
We constructed an interrupted time series based on daily parish-level data on COVID-19 vaccinations to analyze the association with the lottery. We used recursive partitioning to separate vaccination uptake due to the Delta variant from vaccination uptake due to the lottery and limited our study period to May 25 through July 20, 2021. We performed subanalyses that grouped parishes by political affiliation, hesitancy toward COVID-19 vaccines, race and ethnicity, and socioeconomic status to detect heterogeneous responses to the lottery by these characteristics. We ran models separately for parishes in the top and bottom tertiles of each sociodemographic indicator and used a z test to check for differences.
Results:
The lottery was associated with an additional 1.03 (95% CI, 0.61-1.45; P < .001) first doses per parish per day. Comparing lottery impacts between top and bottom tertiles, we found significantly larger associations in parishes with lower vaccine hesitancy rates, higher percentage of Hispanic population, higher median annual household income, and more people with a college degree.
Conclusions:
Results suggest that the lottery was associated with increased COVID-19 vaccination uptake in Louisiana. However, larger associations were observed in parishes with an already higher likelihood of accepting vaccines, which raises equity issues about the opportunity created by the lottery and its effectiveness as a long-term behavioral incentive.
COVID-19
vaccine uptake
lottery incentive
heterogeneous associations
Louisiana
typesetterts1
cover-dateJanuary/February 2023
==== Body
pmcDespite the widespread availability of safe and effective vaccines in the United States against COVID-19, vaccine uptake continues to lag, especially in the Deep South. 1 Just 1 month before the White House goal to have 70% of residents vaccinated by July 4, 2021, the vaccination rates of Mississippi, Louisiana, and Alabama were only 45% and sat at the bottom ranks among all states. 2 Vaccination rates also varied across counties according to sociodemographic characteristics. Nationally, low income, low educational attainment, identifying as Black, and conservative political leaning were associated with high levels of vaccine hesitancy,3-5 and vaccination uptake was typically low in rural counties, 6 counties with a high proportion of Republican Party (GOP) voters, 7 and counties with high poverty rates. 6
Sixteen state governors created COVID-19 lottery-based incentives with a jackpot of at least $1 million to increase vaccination rates, including the first vaccination lottery with a million-dollar prize in Ohio in May 2021. 8 Following that strategy, Louisiana launched the first and only vaccine lottery campaign in the Deep South on June 17, 2021. 9 This 6-week campaign, dubbed “Shot at a Million,” offered one $1-million jackpot and 4 additional $100 000 cash prizes to adults who had received at least 1 dose of COVID-19 vaccine and registered for the lottery by July 31, 2021. 10
In contrast to smaller, guaranteed rewards, large but uncertain prizes such as cash lotteries are often used to nudge people toward health behaviors they would not spontaneously adopt or are reluctant to adopt. Lottery incentives capitalize on people’s psychology of overestimating small probabilities and have been successfully used to promote onetime behaviors such as vaccination and screening. 11 Experimental evidence also suggests that lottery incentives tend to be more effective than guaranteed but smaller bonuses in promoting health interventions.12,13
Evidence of the effects of the first COVID-19 vaccine lottery incentive campaign in Ohio is mixed. Robertson et al 14 analyzed 12 state vaccine lotteries, including the Ohio lottery, using cumulative vaccination data. Although they detected an increase in vaccination uptake for Ohio and 9 other states, the impact in Arkansas and California was not significant. In contrast, Walkey et al 15 argued that increases in vaccinations in Ohio after the lottery could have resulted from the nearly contemporaneous authorization of the Pfizer vaccine for use in teenagers. In addition to these mixed findings, studies have not examined which attributes of local populations are most likely to be associated with increased lottery effectiveness.
Our work contributes to this existing literature in 2 important ways. First, we examined the association of the Louisiana lottery, the only COVID-19 vaccination reward lottery in the Deep South, with COVID-19 vaccination uptake. Second, we examined heterogeneous associations of the lottery with vaccination uptake by parish (county) attributes.
Methods
We used an interrupted time-series design to measure changes in daily parish-level counts of first-dose vaccinations in Louisiana before and after the launch, on June 17, 2021, of the “Shot at a Million” lottery. Our study did not include human subjects; analysis was limited to publicly available data aggregated to the parish level and, as such, was exempt from institutional review board review.
Data
We collected data for all parishes (N = 64) in Louisiana from several sources. First, we obtained the number of first doses, from any manufacturer, of COVID-19 vaccine administered on each day to working-age adults (people aged 18 to 64 years) from the COVID-19 Data Tracker published by the Centers for Disease Control and Prevention (CDC). 16 To control for the severity of the pandemic, we also collected data on the daily COVID-19 death count from the Center for Systems Science and Engineering COVID-19 GitHub Repository of Johns Hopkins University. 17 We used COVID-19 death counts instead of COVID-19 cases because we aimed to capture factors that influence vaccination choices and that may be more accurately proxied by a severe outcome such as death than by a less severe outcome (illness).
We analyzed heterogeneous lottery associations across 4 dimensions of parish-level sociodemographic characteristics: (1) political affiliation, (2) level of hesitancy toward COVID-19 vaccine, (3) racial and ethnic distribution, and (4) socioeconomic status. We measured political affiliation with the percentage of votes for the GOP in the 2020 presidential election. 18 Vaccine hesitancy was measured by the percentage of the population with strong hesitancy toward the COVID-19 vaccine, data for which were adapted from the federal Household Pulse Survey from May 26–June 7, 2021, by the Office of the Assistant Secretary for Planning and Evaluation. 19 We obtained parish-level data on racial and ethnic distribution from the US Census Bureau 2019 release. 20 To measure socioeconomic characteristics, we used parish-level median annual household income, unemployment rate, and percentage of the population with ≥bachelor’s degree; data were obtained from the Economic Research Service at the US Department of Agriculture. 21
Methodology
We used the following interrupted time-series regression model to evaluate the lottery policy:
Dose1ct=α+βTt+γDt+δTt×Dt+πCovidDeathct+τc+εct
where Dt is a dummy indicating the days on and after the lottery announcement date (June 17, 2021), Tt is the running variable measured in days relative to the lottery announcement date, δ measures the change in slope in the postlottery announcement period, and γ measures the discontinuity at the lottery intervention point. We used 7-day moving averages of both counts of first doses administered and COVID-19 deaths as dependent and control variables, respectively, to reduce the noise in the data resulting from fluctuations between weekdays and weekends and different working days of vaccine clinics. We included parish fixed-effect terms ( τc ) to absorb time-invariant unobserved differences across parishes. SEs were clustered at the parish level. To examine whether the lottery impact varied among parishes according to characteristics, we ran the model separately for parishes in the top and bottom tertiles of each sociodemographic indicator. We also used z tests 22 to analyze whether δ, the coefficient indicating the lottery effect, differed significantly between the models limited to the top and bottom tertiles of selected sociodemographic variables. The regression models were conducted in Stata version 16.0 (StataCorp LLC).
To isolate the association with Louisiana’s lottery from other factors that influenced COVID-19 vaccination uptake, we limited the study period to a few weeks before and after the lottery announcement. Our analysis window started on May 25, 2021; most COVID-19 restrictions were lifted on that date as COVID-19 hospitalizations in Louisiana dropped to one of their lowest points since the beginning of the pandemic. 23 This start date was 2 weeks after the day the US Food and Drug Administration authorized the Pfizer-BioNTech COVID-19 vaccine for emergency use in adolescents aged 12 to 15 years, on May 10, 2021. 24 We selected this start date to avoid the confounding that may have been present in the analyses of the Ohio lottery.
To remove the effect of the spread of the Delta variant on vaccination uptake, we closed our analysis window before the surge in Delta cases. We used a model-based recursive partitioning method to identify the kink point in the trend in vaccination uptake. This method fits a designated parametric model to subsets of data generated according to a partitioning variable and tests for parameter instability during the recursive process to identify the partition with the highest parameter instability. 25 We applied this recursive partitioning method to our data from June 17 through July 31, 2021, to identify the date after which the coefficient of the running variable T changed the most (ie, the date after which news of the Delta coronavirus variant started to affect vaccine uptake). To perform this analysis, we used the “partykit” package version 1.2-15 in R 4.1.1 (R Core Team). 26 We found a change in trend on July 20, 2021. Thus, we set our final analysis window from May 25 through July 20, 2021.
Finally, we performed a simple calculation to estimate the total number of additional doses associated with the lottery for the 64 parishes during the 34 days after the lottery announcement (from June 17 through July 20). Using the parameters of the regression model, the total additional doses would be (γ × 34 + δ × 34 × 34/2) × 64, where γ is the parish-level discontinuity value and δ is the daily marginal increase in parish-level vaccinations.
Results
Statewide Louisiana vaccinations increased more rapidly after the kink point identified by our recursive partitioning algorithm (July 20) (Figure panel A). The raw mean number of first-dose vaccinations rose from 43.7 doses per parish per day before the lottery announcement to 47.6 doses per parish per day afterward (Table 1). Statewide vaccine uptake also trended upward after the lottery was announced, whereas previously it was trending downward (Figure panel B). The interrupted time-series regression demonstrated a positive association between the lottery and vaccination uptake among working-age adults in Louisiana: the change in slope after the lottery announcement was both positive and significant (change in slope = 1.03; 95% CI, 0.61-1.45; P < .001) (Table 2).
Figure. Trend in number of first doses of COVID-19 vaccine administered in Louisiana, May 25–August 15, 2021. Scatters indicate 7-day moving average of first doses administered, including 1 dose of the Pfizer or Moderna COVID-19 vaccine or 1 dose of the single-dose Johnson & Johnson/Janssen vaccine. A, Vertical lines indicate vaccine lottery announcement (June 17) and the Delta variant–related kink point (July 20). B, Fitted lines depict vaccination trends before and after vaccine lottery announcement; shaded areas indicate 95% CIs. Data source: Centers for Disease Control and Prevention. 16
Table 1. Descriptive statistics for parish-level COVID-19 vaccinations, COVID-19 deaths, and sociodemographic characteristics, in an analysis of the policy impact of a COVID-19 vaccination reward lottery campaign, Louisiana, May 25 through July 20, 2021 a
Variable Statistics
Mean (SD) Median (IQR)
COVID-19–related variables
Daily no. of first-dose vaccines administered b
Prelottery (May 25–June 16) (N = 1472 parish-days) 43.7 (100.0) 10 (1-38)
Postlottery (June 17–July 20) (N = 2176 parish-days) 47.6 (124.3) 12 (1-40)
Daily no. of COVID-19 deaths c
Prelottery (May 25–June 16) (N = 1472 parish-days) 0.1 (0.4) 0 (0-0)
Postlottery (June 17–July 20) (N = 2176 parish-days) 0.1 (0.4) 0 (0-0)
Parish-level sociodemographic variables (N = 64 parishes)
Voted for GOP in 2020 presidential election, % d 64.6 (14.5) —
People with strong hesitancy toward vaccines, % e 11.7 (1.2) —
Non-Hispanic Black, % f 31.7 (14.3) —
Non-Hispanic White, % f 61.3 (13.3) —
Hispanic of all races, % f 3.8 (2.4) —
Median annual household income, $ g 48 080.00 (10 442.30) —
Unemployment, % g 7.9 (1.6) —
People with ≥bachelor’s degree, % g 17.3 (6.8) —
Abbreviations: —, not applicable; GOP, the Republican Party; IQR, interquartile range.
a On June 17, 2021, Louisiana launched a 6-week campaign that offered one $1-million jackpot and 4 additional $100 000 cash prizes to adults who had received ≥1 dose of COVID-19 vaccine and registered for the lottery by July 31, 2021.
b Vaccination data are from the Centers for Disease Control and Prevention’s COVID-19 Data Tracker. 16
c COVID-19 death data are from Johns Hopkins University Center for Systems Science and Engineering COVID-19 GitHub Repository. 17
d Percentage of votes for GOP in the 2020 presidential election is from a public release on GitHub. 18
e Vaccine hesitancy data are from the Office of the Assistant Secretary for Planning and Evaluation. 19
f Parish-level race and ethnicity proportion data are from the US Census Bureau. 20
g Other parish-level socioeconomic data are from the Economic Research Service at the US Department of Agriculture. 21
Table 2. Regression results for all parishes and those in the top tertile and bottom tertile of political affiliation and attitudes toward COVID-19 vaccines, in an analysis of the policy impact of a COVID-19 vaccination reward lottery campaign, Louisiana, May 25 through July 20, 2021 a
Item All Percentage of residents with strong hesitancy toward vaccines b Percentage of residents who voted for GOP in 2020 presidential election c
Bottom tertile Top tertile Bottom tertile Top tertile
Postlottery slope change d 1.03 (0.61 to 1.45) [<.001] 1.88 (0.79 to 2.97) [.002] 0.34 (−0.11 to 0.80) [.13] 1.21 (0.28 to 2.15) [.01] 0.70 (0.05 to 1.35) [.04]
Lottery discontinuity d 0.48 (−2.81 to 3.77) [.77] 1.10 (−6.62 to 8.81) [.77] −0.95 (−6.57 to 4.67) [.73] 1.15 (−5.51 to 7.81) [.72] −0.71 (−5.45 to 4.02) [.76]
Prelottery slope e −0.74 (−1.07 to −0.42) [<.001] −1.38 (−2.19 to −0.56) [.002] −0.21 (−0.61 to 0.18) [.27] −0.88 (−1.58 to −0.19) [.02] −0.42 (−0.90 to 0.06) [.08]
No. of observations 3648 1254 1140 1254 1197
Adjusted R2 0.90 0.88 0.72 0.91 0.89
z test f — 2.71 [.006] 0.93 [.35]
Abbreviations: —, not applicable; GOP, the Republican Party.
a On June 17, 2021, Louisiana launched a 6-week campaign that offered one $1-million jackpot and 4 additional $100 000 cash prizes to adults who had received ≥1 dose of COVID-19 vaccine and registered for the lottery by July 31, 2021. Vaccination data are from the Centers for Disease Control and Prevention’s COVID-19 Data Tracker. 16 All values are coefficient (95% CI) [P value] unless otherwise indicated. Coefficients were gained from interrupted time-series regressions controlled for parish-level fixed effects and the number of COVID-19 deaths. P < .05 was considered significant.
b Vaccine hesitancy data are from the Office of the Assistant Secretary for Planning and Evaluation. 19
c Percentage of votes for GOP in the 2020 presidential election is from a public release on GitHub. 18
d Postlottery slope change and lottery discontinuity indicate the impact of the lottery on COVID-19 vaccination uptake. Postlottery period is from June 17 to July 20, 2021.
e Prelottery period is from May 25 to June 16, 2021.
f A z test was used to compare the coefficients of postlottery slope change between samples limited to bottom and top tertiles. 22 Values are z score [P value].
Using the regression results, we calculated 39 146 (95% CI, 16 450-61 841) additional doses associated with the lottery for the whole state in the 34 days after the lottery announcement. Based on the 2020 Louisiana population, we calculated an increase of 1.37 (95% CI, 0.58-2.17) percentage points in first-dose vaccination rate from a baseline rate of 37.5% (the vaccination rate on June 17, 2021). Given $1.4 million in total lottery prizes, we estimated that the cost per marginal vaccination was $35.76 (95% CI, $22.64-$85.10).
The response to the lottery differed significantly between parishes in the top and bottom tertiles of vaccine hesitancy (Table 2). Although the association between the lottery and vaccination uptake was nominal in high-hesitancy parishes, the association was more pronounced in low-hesitancy parishes: in low-hesitancy parishes, the change in slope after the lottery announcement was positive and significant (change in slope = 1.88; 95% CI, 0.79-2.97; P = .002). Response to the lottery also differed to different degrees between parishes according to political affiliation but not significantly according to the z test (Table 2). This result suggests that the parishes in the top tertile of GOP voters also responded to the lottery, although the response was moderate (change in slope = 0.70; 95% CI, 0.05-1.35; P = .04) compared with the response in counterpart parishes in the bottom tertile (change in slope = 1.21; 95% CI, 0.28-2.15; P = .01).
The association of the lottery with vaccination uptake among parishes according to racial and ethnic structure was mixed (Table 3). We found no significant differences according to percentages of non-Hispanic Black and non-Hispanic White populations based on the z test. However, we found a significantly larger association (change in slope = 2.13; 95% CI, 1.01-3.24; P < .001) in parishes in the top tertile than in the bottom tertile of Hispanic population percentage.
Table 3. Regression results for parishes according to race and ethnicity structures, in an analysis of the policy impact of a COVID-19 vaccination reward lottery campaign, Louisiana, May 25 through July 20, 2021 a
Item Percentage of residents who are non-Hispanic Black b Percentage of residents who are Hispanic b Percentage of residents who are non-Hispanic White b
Bottom tertile Top tertile Bottom tertile Top tertile Bottom tertile Top tertile
Postlottery slope change c 0.84 (0.22 to 1.46) [.01] 1.07 (0.18 to 1.96) [.02] 0.45 (0.09 to 0.81) [.02] 2.13 (1.01 to 3.24) [<.001] 1.32 (0.38 to 2.26) [.008] 0.88 (0.21 to 1.54) [.01]
Lottery discontinuity c 1.08 (−3.82 to 5.99) [.65] 2.18 (−4.65 to 9.00) [.51] −0.07 (−3.97 to 3.83) [.97] 1.82 (−7.11 to 10.76) [.67] 1.28 (−5.21 to 7.77) [.69] 1.14 (−4.29 to 6.57) [.67]
Prelottery slope d −0.58 (−1.04 to −0.11) [.02] −0.80 (−1.50 to −0.10) [.03] −0.35 (−0.68 to −0.03) [.04] −1.49 (−2.34 to −0.64) [.002] −0.95 (−1.66 to −0.23) [.01] −0.58 (−1.08 to −0.09) [.02]
No. of observations 1254 1197 1254 1197 1254 1197
Adjusted R2 0.89 0.89 0.86 0.88 0.91 0.90
z test e −0.44 [.66] −2.98 [.003] 0.80 [.42]
a On June 17, 2021, Louisiana launched a 6-week campaign that offered one $1-million jackpot and 4 additional $100 000 cash prizes to adults who had received ≥1 dose of COVID-19 vaccine and registered for the lottery by July 31, 2021. Vaccination data are from the Centers for Disease Control and Prevention’s COVID-19 Data Tracker. 16 All values are coefficient (95% CI) [P value] unless otherwise indicated. Coefficients were gained from interrupted time-series regressions controlled for parish-level fixed effects and the number of COVID-19 deaths. P < .05 was considered significant.
b Parish-level race and ethnicity proportion data are from the US Census Bureau. 20
c Postlottery slope change and lottery discontinuity indicate the impact of the lottery on COVID-19 vaccine uptake. Postlottery period is from June 17 to July 20, 2021.
d Prelottery period is from May 25 to June 16, 2021.
e A z test was used to compare the coefficients of postlottery slope change between samples limited to the bottom and top tertiles. 22 Values are z score [P value].
The association between the lottery and COVID-19 vaccination uptake did not differ among parishes according to unemployment rates (Table 4). However, we found a significantly greater response to the lottery in parishes in the top tertile of median annual household income (change in slope = 1.81; 95% CI, 0.88-2.74; P < .001) than in the bottom tertile. Parishes in the top tertile of population with a college degree also showed a significantly greater boost in COVID-19 vaccination uptake after the lottery announcement (change in slope = 2.39; 95% CI, 1.40-3.39; P < .001) than their counterparts in the bottom tertile.
Table 4. Regression results for parishes with different socioeconomic characteristics, in an analysis of the policy impact of a COVID-19 vaccination reward lottery campaign, Louisiana, May 25 through July 20, 2021 a
Item Median annual household income b Percentage of residents who are unemployed b Percentage of residents with ≥bachelor’s degree b
Bottom tertile Top tertile Bottom tertile Top tertile Bottom tertile Top tertile
Postlottery slope change c 0.38 (0.09 to 0.68) [.01] 1.81 (0.88 to 2.74) [<.001] 1.18 (0.49 to 1.86) [.002] 1.19 (0.50 to 1.89) [.002] 0.21 (0.04 to 0.38) [.02] 2.39 (1.40 to 3.39) [<.001]
Lottery discontinuity c −1.06 (−4.73 to 2.62) [.56] −0.68 (−7.82 to 6.46) [.84] 1.29 (−5.35 to 7.93) [.69] 0.30 (−4.11 to 4.71) [.89] −1.84 (−4.58 to 0.90) [.18] 5.16 (−2.90 to 13.21) [.20]
Prelottery slope d −0.27 (−0.54 to 0.00) [.05] −1.28 (−1.95 to −0.61) [<.001] −0.86 (−1.39 to −0.32) [.003] −0.79 (−1.26 to −0.32) [.002] −0.11 (−0.27 to 0.05) [.17] −1.84 (−2.56 to −1.12) [<.001]
No. of observations 1254 1197 1368 1197 1254 1197
Adjusted R2 0.88 0.91 0.87 0.90 0.86 0.87
z test e −3.04 [.002] −0.02 [.98] −4.49 [<.001]
a On June 17, 2021, Louisiana launched a 6-week campaign that offered one $1-million jackpot and 4 additional $100 000 cash prizes to adults who had received ≥1 dose of COVID-19 vaccine and registered for the lottery by July 31, 2021. Vaccination data are from the Centers for Disease Control and Prevention’s COVID-19 Data Tracker. 16 All values are coefficient (95% CI) [P value] unless otherwise indicated. Coefficients were gained from interrupted time-series regressions controlled for parish-level fixed effects and the number of COVID-19 deaths. P < .05 was considered significant.
b Parish-level socioeconomic data are from the Economic Research Service at the US Department of Agriculture. 21
c Postlottery slope change and lottery discontinuity indicate the impact of the lottery on COVID-19 vaccination uptake. Postlottery period is from June 17 to July 20, 2021.
d Prelottery period is from May 25 to June 16, 2021.
e A z test was used to compare the coefficients of postlottery slope change between samples limited to the bottom and top tertiles. 22 Values are z score [P value].
Discussion
Our analysis suggests a positive association between the COVID-19 vaccination reward lottery and vaccination uptake in Louisiana. The lottery was associated with 1.03 additional first doses per parish per day. This estimate is likely conservative because we excluded the lottery eligibility period that overlapped the Delta variant surge (July 21–July 31, 2021); there may have been additional lottery effects during that time.
Our analysis suggests a cost of $35.76 per marginal COVID-19 vaccine dose for the Louisiana lottery. Studies of other state lotteries found costs per induced dose of $68 or $75 (Ohio), $20.90 (New York), and $769.60 (West Virginia), with the average marginal cost per lottery of $55 across several states.14,27,28 Generally, the lottery in Louisiana had similar economic efficiencies to those of other statewide lotteries.
Our research showed an increase of 1.37 (95% CI, 0.58-2.17) percentage points in the first-dose COVID-19 vaccination rate, from a baseline rate of 37.5%. Similar increases, ranging from 1.06 to 4.20 percentage points, were found in studies on text-reminder and small-value voucher incentives.29,30 However, for mandate interventions, results varied from no impact to some positive impact.31-33
Our study provided empirical evidence of the heterogeneous associations between the lottery and COVID-19 vaccination uptake across parishes with various sociodemographic characteristics. Our findings echoed the findings that predominantly Democratic parishes are more likely to show high levels of vaccination uptake7,34 and that higher education level is associated with a higher probability of vaccine acceptance.35,36 We expanded on these findings by showing that parishes in the top quartile of education levels also responded to the lottery to a greater extent than parishes in the bottom quartile of education levels. Moreover, we found that parishes with strong antivaccine attitudes, regardless of the determinants of their antivaccine attitudes, were less likely than vaccine-accepting parishes to be nudged by monetary incentives in the form of a lottery.
Our analysis also produced some counterintuitive findings. First, parishes in the top tertile of median annual household income tended to be more responsive to the lottery than the counterpart parishes in the bottom tertile, despite the presumed greater need for income in the latter. This finding might be partly explained by the concentration of wealthy populations in urban areas, where exposure to media could increase both vaccine awareness 37 and information on the lottery, and where the abundance of health resources may have facilitated access to vaccines. 38 However, in the absence of individual-level data, the demographic characteristics of lottery participants are unknown, so some proportion of lottery participants may have come from low-income populations in urban areas. Second, we detected a stronger response to the lottery in parishes in the top tertile (vs bottom tertile) of Hispanic population. Although this finding is veiled by the absence of individual-level data on lottery participants, it may again signal a stronger effect of the lottery on urbanized parishes than on rural parishes in Louisiana, because Hispanic people are highly concentrated in metropolitan areas of the state.39,40
While the heterogeneous effects of the lottery detected in the analysis may reflect differences between urban parishes, which may be wealthier, more educated, and more ethnically diverse than their rural counterparts in Louisiana, they nonetheless point toward the importance of tailoring large lottery-type incentive campaigns to the intended audiences.41,42 Our findings indicate that the Louisiana lottery may only have incentivized populations who were predisposed to taking the vaccine but could not nudge residents with deeply entrenched antivaccine attitudes. Furthermore, the lottery may have exacerbated existing differences in vaccination uptake because urban residents—with better exposure and access to health services than rural residents—had a “better shot” at participating in the lottery. Finally, it is impossible to discount the potential counterproductive effects of the incentive campaign; previous research noted that intended beneficiaries might become suspicious of the reasons behind the incentive and may be even less likely to adopt the promoted behavior. 43
Limitations
Our study had several limitations. First, because of the lack of daily individual-level data, we could not compare the sociodemographic characteristics of the lottery participants or those of people who received the vaccine after the lottery started, thus running the risk of ecological fallacy in our parish-level analysis. Further research is needed to understand the individual-level response to the lottery. Second, the study had no control group, and the pre–post design of the interrupted time-series method could have led to bias because the 2 segments did not cover the same period. For example, the intensity of vaccine misinformation circulating online may have differed between the periods before and after the lottery announcement, or local vaccination events may have taken place to coincide with the lottery in certain geographies. Either of these factors may have led to bias in our ecological study design. Finally, although we used the recursive partitioning method to identify the kink point (when the Delta variant started to confound our research), it is difficult to completely disentangle the impact of the Delta variant on vaccination uptake, which happened not long after the lottery campaign.
Conclusions
Our research found a positive association between the COVID-19 vaccination reward lottery and COVID-19 vaccination uptake in Louisiana. We contribute to the current research on vaccine incentive lotteries by identifying heterogeneous associations of the lottery by parish characteristics. Typically, we found larger associations in parishes with higher COVID-19 vaccination uptake before the lottery was launched. This set of circumstances could raise equity issues if the public resources consumed by the lottery did not create equal opportunities across populations. Also, the effectiveness of lottery as a public behavioral incentive in the long run is unclear because it may bring about a spiral of expectations on rewards for socially beneficial behaviors such as vaccination.
Louisiana is the only Deep South state that launched a COVID-19 vaccination reward lottery. The positive association suggested by our research may add support to similar practices in other states with politically conservative populations and relatively low vaccination rates. Although effective on the margin, a lottery alone may not be a particularly strong nudge for people with predisposed vaccine hesitancy. Other nonmonetary interventions targeted to spread vaccine knowledge and ease the vaccination process may need to be considered.
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding: The authors received no financial support for the research, authorship, and/or publication of this article.
ORCID iD: Yin Wang, MA https://orcid.org/0000-0002-3554-656X
==== Refs
References
1. Mollalo A Tatar M . Spatial modeling of COVID-19 vaccine hesitancy in the United States. Int J Environ Res Public Health. 2021;18 (18 ):9488. doi:10.3390/ijerph18189488 34574416
2. Gamio L Walker AS . See which states are falling behind Biden’s vaccination goal. The New York Times. Updated June 3, 2021. Accessed August 2, 2021. https://www.nytimes.com/interactive/2021/06/03/us/virus-vaccine-states.html
3. Wright A . Lowest rates, highest hurdles: southern states tackle vaccine gap. Pew Charitable Trusts. June 17, 2021. Accessed August 2, 2021. https://pew.org/3xsSemk
4. Ruiz JB Bell RA . Predictors of intention to vaccinate against COVID-19: results of a nationwide survey. Vaccine. 2021;39 (7 ):1080-1086. doi:10.1016/j.vaccine.2021.01.010 33461833
5. Viswanath K Bekalu M Dhawan D Pinnamaneni R Lang J McLoud R . Individual and social determinants of COVID-19 vaccine uptake. BMC Public Health. 2021;21 (1 ):818. doi:10.1186/s12889-021-10862-1 33910558
6. Tolbert J Orgera K Garfield R Kates J Artiga S . Vaccination is local: COVID-19 vaccination rates vary by county and key characteristics. Kaiser Family Foundation. May 12, 2021. Accessed June 10, 2021. https://www.kff.org/coronavirus-covid-19/issue-brief/vaccination-is-local-covid-19-vaccination-rates-vary-by-county-and-key-characteristics
7. Kates J Tolbert J Orgera K . The red/blue divide in COVID-19 vaccination rates. Kaiser Family Foundation. September 14, 2021. Accessed October 6, 2021. https://www.kff.org/policy-watch/the-red-blue-divide-in-covid-19-vaccination-rates
8. Governor DeWine announces vaccine incentives, end date for health orders. News release. Office of the Governor, State of Ohio. May 12, 2021. Accessed June 10, 2021. https://governor.ohio.gov/wps/portal/gov/governor/media/news-and-media/vaccine-incentives-end-date-for-health-orders-05122021
9. Louisianans vaccinated against COVID-19 by July 31 get a “Shot at a Million” in $2.3 million lottery for cash, scholarship prizes. News release. Office of the Governor, State of Louisiana. June 17, 2021. Accessed June 19, 2021. https://gov.louisiana.gov/index.cfm/newsroom/detail/3216
10. Louisiana Department of Health. Official rules. Accessed August 2, 2021. https://shotatamillion.com/official-rules
11. Vlaev I King D Darzi A Dolan P . Changing health behaviors using financial incentives: a review from behavioral economics. BMC Public Health. 2019;19 (1 ):1059. doi:10.1186/s12889-019-7407-8 31391010
12. Haisley E Volpp KG Pellathy T Loewenstein G . The impact of alternative incentive schemes on completion of health risk assessments. Am J Health Promot. 2012;26 (3 ):184-188. doi:10.4278/ajhp.100729-arb-257 22208418
13. Kim DG . Vaccination lottery. Econ Lett. 2021;208 :110059. doi:10.1016/j.econlet.2021.110059
14. Robertson C Schaefer KA Scheitrum D . Are vaccine lotteries worth the money? Econ Lett. 2021;209 :110097. doi:10.1016/j.econlet.2021.110097 35153345
15. Walkey AJ Law A Bosch NA . Lottery-based incentive in Ohio and COVID-19 vaccination rates. JAMA. 2021;326 (8 ):766-767. doi:10.1001/jama.2021.11048 34213530
16. Centers for Disease Control and Prevention. COVID-19 vaccination data in the United States. Updated April 29, 2022. Accessed October 21, 2021. https://www.cdc.gov/coronavirus/2019-ncov/vaccines/distributing/reporting-counties.html
17. Johns Hopkins University Center for Systems Science and Engineering. COVID-19 data repository by the Center for Systems Science and Engineering (CSSE) at Johns Hopkins University. Accessed August 20, 2021. https://github.com/CSSEGISandData/COVID-19
18. McGovern T . United States general election presidential results by county from 2008 to 2020. Updated October 7, 2021. Accessed July 30, 2021. https://github.com/tonmcg/US_County_Level_Election_Results_08-20
19. Office of the Assistant Secretary for Planning Evaluation. Vaccine hesitancy for COVID-19: state, county, and local estimates. June 16, 2021. Accessed July 30, 2021. https://aspe.hhs.gov/reports/vaccine-hesitancy-covid-19-state-county-local-estimates
20. US Census Bureau. County population by characteristics: 2010-2019. Updated October 8, 2021. Accessed July 30, 2021. https://www.census.gov/data/tables/time-series/demo/popest/2010s-counties-detail.html
21. US Department of Agriculture, Economic Research Service. County-level data sets. Updated June 3, 2022. Accessed July 30, 2021. https://www.ers.usda.gov/data-products/county-level-data-sets
22. Paternoster R Brame R Mazerolle P Piquero A . Using the correct statistical test for the equality of regression coefficients. Criminology. 1998;36 (4 ):859-866. doi:10.1111/j.1745-9125.1998.tb01268.x
23. With COVID-19 hospitalizations dropping to their lowest point, Gov. Edwards signs updated public health emergency order ending most restrictions. News release. Office of the Governor, State of Louisiana. May 25, 2021. Accessed August 2, 2021. https://gov.louisiana.gov/index.cfm/newsroom/detail/3153
24. US Food and Drug Administration. Coronavirus (COVID-19) update: FDA authorizes Pfizer-BioNTech COVID-19 vaccine for emergency use in adolescents in another important action in fight against pandemic. News release. US Food and Drug Administration. May 10, 2021. Accessed August 2, 2021. https://www.fda.gov/news-events/press-announcements/coronavirus-covid-19-update-fda-authorizes-pfizer-biontech-covid-19-vaccine-emergency-use
25. Zeileis A Hothorn T Hornik K . Model-based recursive partitioning. J Comput Graphic Stat. 2008;17 (2 ):492-514. doi:10.1198/106186008x319331
26. Hothorn T Zeileis A . partykit: a modular toolkit for recursive partytioning in R. J Machine Learning Res. 2015;16 (118 ):3905-3909.
27. Brehm M Brehm P Saavedra M . The Ohio vaccine lottery and starting vaccination rates. Am J Health Econ. 2022;8 (3 ). Posted online October 13, 2021. doi:10.1086/718512
28. Barber A West J . Conditional cash lotteries increase COVID-19 vaccination rates. J Health Econ. 2022;81 :102578. doi:10.1016/j.jhealeco.2021.102578 34986437
29. Dai H Saccardo S Han MA , et al . Behavioural nudges increase COVID-19 vaccinations. Nature. 2021;597 (7876 ):404-409. doi:10.1038/s41586-021-03843-2 34340242
30. Campos-Mercade P Meier AN Schneider FH Meier S Pope D Wengström E . Monetary incentives increase COVID-19 vaccinations. Science. 2021;374 (6569 ):879-882. doi:10.1126/science.abm0475 34618594
31. Kreps SE Kriner DL . How do COVID-19 vaccine mandates affect attitudes toward the vaccine and participation in mandate-affected activities? Evidence from the United States. Vaccine. 2022. Posted online March 2, 2022. doi:10.1016/j.vaccine.2022.02.083
32. Sprengholz P Betsch C Böhm R . Reactance revisited: consequences of mandatory and scarce vaccination in the case of COVID-19. Appl Psychol Health Well-Being. 2021;13 (4 ):986-995. doi:10.1111/aphw.12285 34032388
33. Batteux E Mills F Jones LF Symons C Weston D . The effectiveness of interventions for increasing COVID-19 vaccine uptake: a systematic review. Vaccines (Basel). 2022;10 (3 ):386. doi:10.3390/vaccines10030386 35335020
34. Ivory D Leatherby L Gebeloff R . Least vaccinated U.S. counties have something in common: Trump voters. The New York Times. Updated April 17, 2021. Accessed June 16, 2021. https://www.nytimes.com/interactive/2021/04/17/us/vaccine-hesitancy-politics.html
35. Martinelli M Veltri GA . Do cognitive styles affect vaccine hesitancy? A dual-process cognitive framework for vaccine hesitancy and the role of risk perceptions. Soc Sci Med. 2021;289 :114403. doi:10.1016/j.socscimed.2021.114403 34547544
36. Galarce EM Minsky S Viswanath K . Socioeconomic status, demographics, beliefs and A(H1N1) vaccine uptake in the United States. Vaccine. 2011;29 (32 ):5284-5289. doi:10.1016/j.vaccine.2011.05.014 21621577
37. Chen W Stoecker C . Mass media coverage and influenza vaccine uptake. Vaccine. 2020;38 (2 ):271-277. doi:10.1016/j.vaccine.2019.10.019 31699506
38. Murthy BP Sterrett N Weller D , et al . Disparities in COVID-19 vaccination coverage between urban and rural counties—United States, December 14, 2020–April 10, 2021. MMWR Morb Mortal Wkly Rep. 2021;70 (20 ):759-764. doi:10.15585/mmwr.mm7020e3 34014911
39. US Census Bureau. Racial and ethnic diversity in the United States: 2010 Census and 2020 Census. Accessed October 10, 2021. https://www.census.gov/library/visualizations/interactive/racial-and-ethnic-diversity-in-the-united-states-2010-and-2020-census.html
40. Who lives in New Orleans and metro parishes now? The Data Center. August 12, 2021. Accessed October 10, 2021. https://www.datacenterresearch.org/data-resources/who-lives-in-new-orleans-now
41. Boxer ML . Improving Underserved Population Health: Attitudes on Medicaid Financial Incentives for Medication Adherence in Asthmatic Children. Dissertation. Medical University of South Carolina; 2018.
42. Roseleur J Harvey G Stocks N Karnon J . Behavioral economic insights to improve medication adherence in adults with chronic conditions: a scoping review. Patient. 2019;12 (6 ):571-592. doi:10.1007/s40271-019-00377-8 31332723
43. Judah G Darzi A Vlaev I , et al . Financial disincentives? A three-armed randomised controlled trial of the effect of financial Incentives in Diabetic Eye Assessment by Screening (IDEAS) trial. Br J Ophthalmol. 2018;102 (8 ):1014-1020. doi:10.1136/bjophthalmol-2017-311778 29793929
| 36062380 | PMC9703024 | NO-CC CODE | 2022-12-14 23:26:51 | no | Public Health Rep. 2022 Sep 5; 138(1):68-75 | utf-8 | Public Health Rep | 2,022 | 10.1177/00333549221120676 | oa_other |
==== Front
JMC
spjmc
Journalism & Mass Communication Educator
1077-6958
2161-4326
SAGE Publications Sage CA: Los Angeles, CA
10.1177/10776958221135969
10.1177_10776958221135969
Research Article
What Do Employers Expect for Jobs Requiring Media Analytics? A Semantic Network Analysis of Job Descriptions of In-Person and Remote Positions During the COVID-19 Pandemic
Jiang Ke 1
Xu Qian 1
Afromsky Ashleigh 1
1 Elon University, NC, USA
Ke Jiang, Assistant Professor, School of Communications, Elon University, 50 Campus Drive, Elon, NC 27244, USA. Email: [email protected]
24 11 2022
24 11 2022
10776958221135969© AEJMC 2022
2022
Association for Education in Journalism & Mass Communication
This article is made available via the PMC Open Access Subset for unrestricted re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the COVID-19 pandemic or until permissions are revoked in writing. Upon expiration of these permissions, PMC is granted a perpetual license to make this article available via PMC and Europe PMC, consistent with existing copyright protections.
Using text mining and semantic network analysis, this study analyzed the job descriptions of 34,787 positions about media analytics from Indeed.com to compare how the in-person and remote jobs differ to inform educators about integrating analytics in the media and communications curriculum. We found that the in-person positions emphasized more on the skills of verbal, interpersonal, and organizational communication, whereas the remote positions asked more for written communication. While the in-person positions had higher expectations of using general data management and analysis tools, the remote positions emphasized more on the use of social media analytics and digital marketing tools.
media analytics
remote
in-person
semantic network analysis
edited-statecorrected-proof
typesetterts1
==== Body
pmcDue to media digitization, a large amount of data about media users and how they consume media content have become more accessible (Kapatamoyo, 2019). Media analytics has thus integrated into the functioning of media and communications professions to drive daily decision-making (Hollifield, 2020). It has become an increasingly sought ability for new college graduates seeking media and communication jobs (Adams & Lee, 2021; Freberg & Kim, 2018). Educators have called attention to integrate media analytics into undergraduate curriculum in media and communications (Neill & Schauster, 2015). However, the lack of clarification on the roles and responsibilities of media analytics makes it challenging for media and communication educators to figure out the industry expectations and prepare undergraduates with relevant skills (Stansberry & MacKenzie, 2020). To address this challenge, this study analyzes the description of job postings with the keyword of media analytics.
Due to the outbreak of the COVID-19 pandemic, many jobs have transitioned to remote work since Spring 2020, which provides us a unique context to compare how the expectations of the in-person and the remote positions about media analytics differ. This remote working trend is said to extend beyond the pandemic at least partially if not to the full extent (Coate, 2021; Parker et al., 2020). Therefore, this study chooses to situate the analysis of job descriptions in the context of COVID-19 to better understand the industry expectations and gain insights into how to prepare our students for both the in-person and remote positions of media analytics. Specifically, this study aims to uncover the differences between the in-person and remote media analytics positions regarding job location, job title, employer type, and how they describe expected skills and required tools through text mining and semantic network analysis.
Analytics and Media Analytics
The term analytics has been used in different ways and in different contexts. While some definitions highlight its quantitative rigor (Cooper, 2012; Daniel, 2015; Tandoc, 2019), others underline its goal-driven nature (Hawkins, 2008; Hollifield, 2020). Following Dinsmore (2016), this study considers analytics as the process of developing useful insights from data for problem-solving. Media analytics focuses on the data about media users and media content (Manovich, 2018). In media and communication industries, analytics includes a wide range of practices, such as general media monitoring, social media listening, evaluation of advertising and marketing effectiveness, assessment of user experience, and distribution forecasting. (Granados, 2019; PwC, n.d.). Despite the differences in purpose, these analytical practices all deal with data about either users or how they consume media and media content. Therefore, this study defines media analytics as the process and practice of gathering and analyzing data about both media users and media content for problem-solving. This problem-solving orientation requires media analytics to go beyond statistical analysis to include the components of asking the right questions, identifying data to answer the questions, gaining actionable insights from data, and effectively presenting insights (Grady, 2020; Hollifield, 2020). With the proliferation of web and social media, more communication positions are particularly looking for employees with media analytics skills to understand the data of user interaction with online and interactive media (Meng et al., 2019).
Data Competence and Analytics in Media and Communication Education
According to a recent study released by the Plank Center for Leadership in Public Relations, communication professionals in North America have identified large gaps between perceived importance and personal qualification level of data, technology, and management competencies, with data as the weakest area of developed competence (Meng et al., 2021). Compared with older professionals, younger professionals were found to be particularly lagging in skills and knowledge about data (Meng et al., 2021). Data competence is critical to media analytics. It addresses the ability to conduct analysis of data about media users and content as well as the knowledge and practices about research methods, data collection, and interpretation of results. Media analytics also require technology competence, which facilitates the learning and use of relevant data tools.
To meet the increasing needs of new hires with data and technology competences, educators in media and communications have called for an integration of analytics at both the program and the course levels (Adams & Lee, 2021). For example, Commission on Public Relations Education (CPRE, 2018) has identified analytics as one of the most important topics for undergraduate public relations curriculum. Accrediting Council on Education in Journalism and Mass Communications (ACEJMC, 2012) has also emphasized the assessment of students’ ability to conduct research and evaluate information with appropriate methods and apply basic numerical and statistical concepts.
However, universities have been slow in integrating analytics into curriculum. O’Boyle and Sturgill (2020) found that most programs of ACEJMC-accredited universities did not offer any analytics-focused course, with only about half offering some courses with analytics components. The programs with a dedicated focus on media analytics are mainly at the graduate level (O’Boyle & Sturgill, 2020). The lack of clarification on the role and responsibilities of media analytics makes it challenging for media and communication educators to figure out how to prepare and integrate analytics into the undergraduate curriculum (Stansberry & MacKenzie, 2020). To help educators understand what is expected from an industry perspective, this study examines the descriptions of jobs requiring media analytics on Indeed.com. It is important to note that the positions requiring media analytics are not just limited to media industries. Other industries provide the relevant positions too. Therefore, in addition to studying the skills and tools described in the job ads, this study also examines the location, the title, and the employer type of these jobs.
Rise of Remote Work During the COVID-19 Pandemic
The outbreak of the COVID-19 pandemic ushered many offices and workplaces to pivot to remote work in spring 2020. By May 2020, more than one-third of employees in United States reported to work from home during the past 4 weeks due to COVID-19 (Coate, 2021). Among them, employees in management, business, and professional occupations had a higher likelihood to work remotely than others (Coate, 2021). Pew Research Center reported that employees with bachelor’s degree or higher were more likely to report that their work could be done at home than those without a college degree (Parker et al., 2020). Shifting to remote work was also more commonly found for people with more education during the pandemic (Bartik et al., 2020).
Internship, the important career launching pad for college students, has also been significantly influenced by the COVID-19 pandemic. During the peak of the pandemic in 2020, around half of all internship opportunities had been canceled (Martin, 2021), making the demand for internship opportunities far exceed the available positions (Hora et al., 2021). Despite the plunge in overall internship opportunities, remote internships have increased since the start of this pandemic (Feldman, 2021). A recent report from Indeed Hiring Lab showed that the share of all internship postings on its U.S. website dropped 39% compared with the year before the pandemic, while the share of remote internships went up almost seven times in 2020 (Konkel, 2021).
Several of the most common fields with emerging remote jobs and internships, such as marketing, social media, and arts and entertainment (Konkel, 2021; Lusinski & Ward, 2020), are highly relevant to students majoring in media and communications. It is important to understand the similarities and differences between the in-person and the remote positions requiring media analytics. Therefore, this study proposes the following research question.
RQ1: What are the differences and similarities between the in-person and the remote positions requiring media analytics regarding (a) location, (b) job title, and (c) employer type?
By analyzing the expected skills and mastery of tools listed in job ads, this study also informs the development and revision of media and communications curriculum to better prepare students for the new shifts in the job market.
RQ2: What are the differences and similarities between the in-person and remote positions requiring media analytics regarding the skills mentioned in job descriptions?
RQ3: What are the differences and similarities between the in-person and remote positions requiring media analytics regarding the tools mentioned in job descriptions?
Semantic Networks of Job Descriptions
Aside from the mentions of skills and tools, this study further compares the differences between in-person and remote positions by analyzing the semantic networks of job descriptions. Semantic network analysis (SMA) is a form of content analysis identifying the network of associations between words expressed in a text (Doerfel, 1998). This method assumes that the text represents a network of words. The position of concepts within a text network thus “provides insight into the meaning or prominent themes of the text as a whole” (Hunter, 2014, p. 350). Its theoretical foundation rests on the cognitive paradigm (D’Angelo, 2002) and the tradition of frame semantics in linguistics (e.g., Fillmore, 2008). Collins and Quillian (1972) argue that words are hierarchically clustered in memory. Thus, the spatial models illustrating the relations among words are representative of meaning (Barnett & Woelfel, 1988). Through examining the co-occurrence of words in texts, we can identify salient words in specified concept clusters and explain the related framing strategies. SMA has been applied to media and communications scholarship in various areas, such as issue framing (e.g., Jiang et al., 2016), public opinions (e.g., Kwon et al., 2016), and crisis management (e.g., Liu et al., 2018). This study uses SMA to examine and visualize the associations between mentioned skills and tools in job descriptions as well as salient skills or tools in specified concept clusters.
The comparison between the sematic networks of the in-person and the remote positions is conducted through the analyses of Quadratic Assignment Procedure (QAP) correlation, modularity, and centrality. QAP correlation (Borgatti et al., 2002) is a nonparametric technique that does not rely on the assumption of independence. A higher correlation suggests a higher level of similarity in structure between two networks. Modularity analysis addresses how words in a network are clustered into groups (Blondel et al., 2008). It calculates how the mentioned skills and tools are grouped into smaller communities in each semantic network. Together, QAP correlations and modularity analyses help answer the following question about the structural differences in semantic networks.
RQ4: How do the semantic networks of the in-person and remote positions differ from each other regarding (a) network structure and (b) number of clusters?
Each word within a semantic network could have different levels of influence on the network, which is measured through normalized eigenvector centrality (Bonacich, 1972; Freeman, 1978). It is important to examine the overall salience of a particular skill or tool mentioned in each cluster of the semantic networks of in-person and remote positions.
RQ5: What were the most central skills and/or tools in each cluster of the two semantic networks?
Method
Using “media analytics” as the search string, this study identified 34,787 jobs posted on Indeed.com (one of the largest job sites) from May 19, 2020, to January 11, 2021. Among them, more than 25% (n = 7464) were remote positions. The web scraping service parsehub.com was used to collect information about job title, location, employer, and job description. Based on the online database of Dun & Bradstreet (https://www.dnb.com/business-directory/company-search.html), we identified the employer type for the top 100 employers that appeared most frequently.
The text corpus was divided into the in-person and remote positions. The tidytext package for R was used to clean the textual data by removing the stop words that are typically extremely common words in English (e.g., “the,” “of,” “to,” etc.). The most frequently used phrases identified in job descriptions, such as “marketing strategies” and “social media strategies,” were coded as one concept. Frequency analysis in texting mining was used to answer RQ1 to RQ3.
Semantic networks of job descriptions were created based on the bigrams of mentioned skills and tools. QAP correlation, modularity, and normalized eigenvector centrality were calculated to explore the correlation between the two semantic networks and to identify the clusters within each network and the salient skills and/or tools in each cluster (RQ4 and RQ5). The ForceAtlas2 layout in Gephi (Jacomy et al., 2014) was used to create visual maps of semantic networks (Figures 1 and 2) to supplement the discussions of modularity and centrality results.
Figure 1. Semantic Network of Skills and Tools in Job Descriptions of Media Analytics In-Person Positions.
Note. The labels with same color are in the same cluster. The size of each skill or tool’s label depends on its eigenvector centralities, such that the larger the object, the more central it is in the job description. Lines on the visualizations indicate the presence of a relationship between each pair of nodes. The thicker lines represent a stronger relationship between two nodes.
Figure 2. Semantic Network of Skills and Tools in Job Descriptions of Media Analytics Remote Positions.
Note. The labels with same color are in the same cluster. The size of each skill or tool’s label depends on its eigenvector centralities, such that the larger the object, the more central it is in the job description. Lines on the visualizations indicate the presence of a relationship between each pair of nodes. The thicker lines represent a stronger relationship between two nodes.
Results
Top Job Locations, Job Titles, and Employer Types (RQ1)
Both the in-person and remote jobs were mainly from the metropolitan areas with the highest economic outputs, especially along the East and West Coast. Table 1 lists the top 20 cities with the greatest numbers of in-person and remote positions, respectively. Among them, New York City had the most in-person and remote positions. While the ratios of remote to in-person positions were relatively lower in Charlotte, Seattle, San Francisco, Chicago, and New York City; Washington, D.C., Portland, Los Angeles, Houston, and Nashville had relatively higher ratios of remote to in-person positions.
Table 1. Top Locations With the Greatest Number of Media Analytics Jobs.
In-Person Remote
Rank City F R City F R
1 New York, NY 1,973 0.19 New York, NY 470 0.19
2 Chicago, IL 795 0.18 Los Angeles, CA 257 0.32
3 San Francisco, CA 664 0.18 Washington, DC 201 0.39
4 Los Angeles, CA 555 0.32 Austin, TX 194 0.27
5 Austin, TX 530 0.27 Atlanta, GA 189 0.27
6 Atlanta, GA 512 0.27 Chicago, IL 178 0.18
7 Boston, MA 400 0.23 San Francisco, CA 145 0.18
8 Seattle, WA 391 0.17 Boston, MA 122 0.23
9 Charlotte, NC 338 0.14 Denver, CO 107 0.29
10 Washington, DC 318 0.39 San Diego, CA 100 0.25
11 San Diego, CA 305 0.25 Dallas, TX 98 0.27
12 Denver, CO 267 0.29 Philadelphia, PA 93 0.34
13 Dallas, TX 265 0.27 Seattle, WA 79 0.17
14 Tampa, FL 206 0.26 Miami, FL 76 0.29
15 Miami, FL 186 0.29 Houston, TX 72 0.30
16 Philadelphia, PA 184 0.34 Tampa, FL 72 0.26
17 Columbus, OH 176 0.09 Minneapolis, MN 65 0.28
18 Houston, TX 169 0.30 Nashville, TN 60 0.30
19 Minneapolis, MN 164 0.28 Portland, OR 60 0.38
20 Orlando, FL 161 0.25 Charlotte, NC 56 0.14
Note. F represents the frequency of cities appeared on the job postings. R is the ratio of remote to in-person jobs in these cities.
As shown in Table 2, the top 20 most frequently mentioned job titles of both in-person and remote positions requiring media analytics are related to marketing or social media. Marketing manager, digital marketing specialist, and digital marketing manager emerged as the three most frequently mentioned job titles for both in-person and remote positions. Compared with the in-person positions, the remote positions involved more opportunities for internships.
Table 2. Top 20 Most Frequent Job Titles for the Media Analytics Jobs.
In-Person Remote
Job title F Job title F
1 Marketing Manager 776 Digital Marketing Specialist 264
2 Digital Marketing Specialist 736 Marketing Manager 242
3 Digital Marketing Manager 650 Digital Marketing Manager 230
4 Marketing Coordinator 588 Social Media Manager 160
5 Marketing Specialist 403 Marketing Coordinator 156
6 Social Media Manager 396 Social Media Intern 104
7 Social Media Specialist 318 Marketing Specialist 95
8 Marketing Assistant 245 Social Media Specialist 76
9 Marketing Analyst 235 Social Media Coordinator 75
10 Social Media Coordinator 221 Marketing Intern 71
11 Marketing Director 198 Marketing Assistant 63
12 Director of Marketing 190 Marketing Director 57
13 Marketing Associate 167 Digital Marketing Account Manager 55
14 Digital Marketing Coordinator 163 Digital Marketing Coordinator 49
15 Product Marketing Manager 156 Marketing Associate 46
16 Marketing Intern 151 Director of Marketing 44
17 Digital Marketing Analyst 118 Content Marketing Manager 40
18 Digital Marketing Strategist 101 Product Marketing Manager 40
19 Content Marketing Manager 95 Digital Marketing Intern 39
20 Email Marketing Specialist 91 Social Media Marketing Intern 39
Note. F represents the frequency of job titles appeared on the job postings.
Among the top 20 employer types offering the greatest number of positions (Table 3), e-commerce employers offered the most in-person positions requiring media analytics, followed by employers in wireless & telecommunications, advertising & marketing services, staffing & recruiting, bank & credit unions, information technology services, pharmaceutical manufacturing, and social media. Employers from advertising & marketing services offered the greatest number of remote positions requiring media analytics, followed by employers in staffing and recruiting, insurance carriers, agencies and brokerages, information technology services, education and training services, e-commerce, and publishing.
Table 3. Top 20 Most Frequent Employer Types for the Media Analytics Jobs.
In-Person Remote
Employer types F Employer types F
E-commerce 516 Advertising & Marketing Services 411
Wireless Telecommunications 421 Staffing & Recruiting 161
Advertising & Marketing Services 396 Insurance Carriers, Agencies & Brokerages 85
Staffing & Recruiting 396 Information Technology Services 74
Banks & Credit Unions 327 Education & Training Services 43
Information Technology Services 265 E-Commerce 34
Pharmaceutical Manufacturing 242 Publishing 34
Social Media 214 Consulting Services 22
Insurance Carriers, Agencies & Brokerages 189 Wireless Telecommunications 22
Computer Software 162 Social Assistance 20
TV Production, Broadcast & Cable Networks 153 Computer Software 15
Nursing Homes & Long-Term Care Facilities 77 Sci. & Tech. Instruments Manufacturing 15
Scientific & Technical Instruments Manufacturing 73 Manufacturing Sector 14
Publishing 59 Scientific Research & Development Services 14
Colleges & Universities 57 Radio Broadcasting & Programming 13
Managed Application 57 Managed Application 12
Consulting Services 50 Computer Manufacturing 11
Medical Equipment & Supplies Manufacturing 49 Discount Department Stores 11
Physicians 45 Lending 11
Real estate 40 Personal care products manufacturing 11
Note. F represents the frequency of employer types for in-person and remote positions.
Most Frequently Mentioned Skills (RQ2)
To explore how in-person and remote positions requiring media analytics differ regarding expected skills, we identified a total of 89 skills and calculated their mention frequencies respectively. Table 4 lists the top 50 most frequently mentioned skills as well as their percentage of frequency among the 89 identified skills. As shown in Table 4, both in-person and remote positions highly emphasized the skills of communication, writing, research, collaboration, and development of marketing strategies.
Table 4. Top 50 Most Frequent Skills in Job Descriptions of Media Analytics Positions.
In-Person F P Remote F P
1 Communication 44,015 19.51 Communication 7566 17.02
2 Writing 18,540 8.22 Writing 3770 8.48
3 Research 14,564 6.46 Research 2,767 6.22
4 Collaboration 11,141 4.94 Collaboration 2,411 5.42
5 Marketing strategies 7,549 3.35 Marketing strategies 1,570 3.53
6 Verbal 6,346 2.81 Content creation 1,312 2.95
7 Project management 6,008 2.66 Verbal 1,168 2.63
8 Organizational 6,001 2.66 Marketing campaigns 1,147 2.58
9 Marketing campaigns 5,428 2.41 Organizational 1,075 2.42
10 Content creation 5,071 2.25 Email marketing 1,056 2.38
11 Marketing automation 4,192 1.86 Project management 1,021 2.30
12 Interpersonal 4,083 1.81 Marketing automation 912 2.05
13 Email marketing 3,927 1.74 Social media platforms 825 1.86
14 Presentation 3,752 1.66 Content marketing 731 1.64
15 Marketing plans 3,695 1.64 Social media marketing 679 1.53
16 Social media platforms 3,060 1.36 Interpersonal 619 1.39
17 Marketing experience 2,666 1.18 Html 593 1.33
18 Html 2,519 1.12 Presentation 593 1.33
19 Oral 2,471 1.10 Marketing plans 564 1.27
20 Content marketing 2,386 1.06 Team player 516 1.16
21 Team player 2,369 1.05 Marketing experience 480 1.08
22 Analytics 2,360 1.05 Brand awareness 475 1.07
23 Brand awareness 2,248 1.00 Marketing channels 467 1.05
24 Customer experience 2,156 0.96 Online marketing 448 1.01
25 Marketing channels 2,035 0.90 Time management 435 0.98
26 Time management 1,998 0.89 Content management 385 0.87
27 Content management 1,921 0.85 Email campaigns 384 0.86
28 Social media marketing 1,904 0.84 Marketing management 372 0.84
29 Programming 1,820 0.81 Content strategy 371 0.83
30 Marketing communication 1,784 0.79 CSS 360 0.81
31 Business development 1,759 0.78 Social media accounts 357 0.80
32 Marketing management 1,748 0.77 Social media strategies 352 0.79
33 Social media channels 1,739 0.77 Oral 351 0.79
34 Content strategy 1,633 0.72 Social media management 351 0.79
35 Online marketing 1,614 0.72 Social media channels 333 0.75
36 CSS 1,516 0.67 Business development 313 0.70
37 Email campaigns 1,492 0.66 Customer experience 307 0.69
38 User experience 1,395 0.62 Advertising campaigns 297 0.67
39 Digital channels 1,389 0.62 User experience 297 0.67
40 Social media strategies 1,365 0.61 Digital advertising 293 0.66
41 Digital advertising 1,308 0.58 Media buying 290 0.65
42 Digital marketing campaigns 1,291 0.57 Social media content 284 0.64
43 Advertising campaigns 1,288 0.57 Digital marketing campaigns 267 0.60
44 Management experience 1,180 0.52 Analytics 249 0.56
45 Digital content 1,171 0.52 Digital channels 240 0.54
46 Product development 1,171 0.52 Programming 240 0.54
47 Social media management 1,163 0.52 Marketing communication 233 0.52
48 Business intelligence 1,126 0.50 Social media campaigns 225 0.51
49 Social media content 1,112 0.49 Data science 200 0.45
50 Data science 1,105 0.49 Strategy development 197 0.44
Note. F represents the frequency of mentioned skills for in-person and remote positions. P represents the percentage of frequency among the identified skills (n = 90).
Compared with remote positions, the job descriptions of in-person positions had a higher percentage of frequency for the skills about communication and presentation, analytics, marketing plans, project and brand management, programming, customer experience, research, product development, management, business intelligence, digital content, data visualization, and business development. In contrast, the job descriptions of remote positions had a higher percentage of frequency for the skills related to content (e.g., content creation, content strategy, and content marketing), social media (e.g., the platform itself, marketing, management, and strategies, and analytics of social media), email marketing and campaigns, marketing strategies, and campaigns in general, collaboration and teamwork, writing, time management, and web development (e.g., HTML and CSS).
Most Frequently Mentioned Tools (RQ3)
After identifying 42 tools mentioned in job descriptions, we calculated their mention frequencies by in-person and remote positions. As shown in Table 5, both the in-person and the remote positions emphasized Google Analytics and Facebook the most. Excel and Instagram were the third most frequently mentioned tools for in-person and remote positions, respectively. Compared with remote positions, in-person positions placed more emphasis on the tools for content creation and communication, such as Microsoft Office, PowerPoint, Word, Adobe Creative, and InDesign, as well as for data management and analysis, such as SQL, Tableau, Salesforce, Python, SAS, Adobe Analytics, Nielsen, ComScore, SPSS, and Kenshoo.
Table 5. Frequency of Mentioned Tools in Job Descriptions of Media Analytics Positions.
In-Person F P Remote F P
1 Google Analytics 7,850 10.81 Google Analytics 2,254 12.91
2 Facebook 7,251 9.99 Facebook 2,076 11.89
3 Excel 6,906 9.51 Instagram 1,321 7.56
4 Salesforce 4,096 5.64 Google Ads 1,147 6.57
5 Instagram 4,092 5.64 Excel 1,109 6.35
6 Twitter 3,925 5.41 Twitter 992 5.68
7 Google ads 3,526 4.86 LinkedIn 956 5.47
8 Word 3,302 4.55 Salesforce 845 4.84
9 PowerPoint 3,256 4.48 YouTube 720 4.12
10 Microsoft office 3,120 4.30 Facebook Page 659 3.77
11 LinkedIn 3,092 4.26 Word 575 3.29
12 YouTube 2,688 3.70 Photoshop 498 2.85
13 SQL 2,536 3.49 Microsoft Office 487 2.79
14 Photoshop 2,075 2.86 PowerPoint 467 2.67
15 Tableau 1,943 2.68 Pinterest 376 2.15
16 Adobe Creative 1,671 2.30 SQL 357 2.04
17 Pinterest 1,328 1.83 Facebook Ads 322 1.84
18 Illustrator 1,123 1.55 Adobe Creative 319 1.83
19 Python 1,058 1.46 Illustrator 258 1.48
20 InDesign 1,022 1.41 Tableau 239 1.37
21 Facebook Ads 903 1.24 InDesign 204 1.17
22 Adobe Analytics 659 0.91 Google Tag 198 1.13
23 Hootsuite 610 0.84 Snapchat 168 0.96
24 Facebook Page 599 0.82 Hootsuite 160 0.92
25 Snapchat 585 0.81 Python 144 0.82
26 SAS 557 0.77 Adobe Analytics 95 0.54
27 Google Tag 531 0.73 Premiere 94 0.54
28 Nielsen 490 0.67 Nielsen 60 0.34
29 Premiere 375 0.52 Facebook Insights 53 0.30
30 ComScore 222 0.31 Sprinklr 51 0.29
31 Sprinklr 207 0.29 Instagram Ads 50 0.29
32 Facebook insights 154 0.21 SAS 37 0.21
33 SPSS 143 0.20 Display Network 33 0.19
34 Microsoft Suite 135 0.19 YouTube Ads 29 0.17
35 Instagram Ads 132 0.18 Twitter Ads 24 0.14
36 Kenshoo 123 0.17 Microsoft Suite 20 0.11
37 Display Network 99 0.14 Brandwatch 17 0.10
38 Twitter Ads 81 0.11 Twitter Analytics 17 0.10
39 Twitter Analytics 66 0.09 Kenshoo 11 0.06
40 Brandwatch 47 0.06 SPSS 11 0.06
41 Sysomos 21 0.03 ComScore 8 0.05
42 YouTube Ads 14 0.02 Sysomos 3 0.02
Note. F represents the frequency of mentioned tools for in-person and remote positions. P represents the percentage of frequency among the identified tools (n = 42).
In contrast, the job descriptions of remote positions placed more emphasis on the specific online and social media platforms and the relevant tools, such as Facebook, Facebook Page, Facebook Ads, and Facebook Insights, Instagram and Instagram Ads, Google Ads and Google Tag, LinkedIn, YouTube and YouTube Ads, Pinterest, Twitter, Snapchat, and Hootsuite.
Differences in Semantic Networks (RQ4 and RQ5)
QAP Correlation and Network Modularity (RQ4a and 4b)
The QAP correlation result showed that the semantic network structures of in-person and remote positions were similar to each other (r = .96, p < .001). The modularity results further confirmed this finding. The modularity scores of the in-person and the remote semantic networks were 0.666 and 0.631 respectively, which indicated dense connections between the skills and the tools within the same clusters and sparse connections between the skills and the tools across different clusters in each semantic network. There were seven clusters in the in-person network and six in the remote one (Table 6).
Table 6. QAP Correlation and Modularity Analysis Results.
In-Person Remote QAP: 0.96
Modularity .666 Modularity .631 (p = .000)
Clusters P (%) Hub Clusters P (%) Hub
Red 42.64 Research Red 36.44 Google Analytics
Blue 21.71 Google Analytics Blue 34.75 Research
Green 10.85 Communication Green 11.86 Facebook
Purple 10.85 Facebook Purple 8.47 SQL
Orange 6.20 Excel Orange 5.08 Excel
LBlue 4.65 HTML LBlue 3.39 Photoshop
LGreen 1 Photoshop
Note. P represents the percentage of each cluster in the semantic network. Hub represents the most salient skills or tool in each cluster in terms of eigenvector centrality. LBlue = Light Blue; LGreen = Light Green; QAP = quadratic assignment procedure.
Most Salient Skill and/or Tool in Each Cluster (RQ5)
Table 7 lists the normalized eigenvector centralities (ranging from 0 to 1, with 1 representing the greatest centrality) and modularity class of each skill and tool in the two semantic networks. For the in-person positions (Figure 1), the largest clusters covered 42.64% of the network and centered around the skill of research closely related to the skills of content creation and collaboration. The second largest cluster (21.71%) centered around the tool Google Analytics closely tied to a variety of digital advertising and marketing tools, such as Google Ads, Facebook Ads, Twitter Ads, and Salesforce. Communication was the hub of the third largest cluster (10.85%), which was closely associated with skills of writing, verbal, oral, interpersonal, and organizational communication as well as presentation. Facebook was the hub of another third largest cluster (10.85%), which was closely tied to other specific social media platforms, such as Instagram, Twitter, LinkedIn, YouTube, and Pinterest. This social media cluster was also closely connected to the skills in managing social media channels, accounts, and pages, as well as increasing social media presence across diverse platforms. Excel was the hub of the fifth largest cluster (6.2%) and was strongly associated with the tools of Word, PowerPoint, and Microsoft Office. The sixth largest cluster centered around the skill of HTML (4.65%), which was closely tied to the skills about CSS, basic HTML, and HTML coding. Photoshop was the hub of the smallest cluster (1%) and was strongly tied to InDesign, Illustrator, and Premiere.
Table 7. Eigenvector Centralities and Modularity Class of Nodes in the Two Semantic Networks.
In-Person Eigen Cluster Remote Eigen Cluster
1 Google Analytics 1.000 Blue Google Analytics 1.000 Red
2 Communication 0.955 Green Facebook 0.817 Green
3 Research 0.859 Red Research 0.814 Blue
4 Email Marketing 0.854 Red Google Ads 0.666 Red
5 Writing 0.793 Green Email marketing 0.653 Red
6 Project management 0.783 Green Writing 0.646 Blue
7 Google Ads 0.757 Blue LinkedIn 0.587 Green
8 Content creation 0.731 Red Communication 0.538 Blue
9 Facebook 0.670 Purple Content creation 0.533 Blue
10 Marketing automation 0.649 Red Twitter 0.525 Green
11 Social media platforms 0.595 Purple Social media platforms 0.509 Green
12 Salesforce 0.583 Blue Instagram 0.484 Green
13 LinkedIn 0.579 Purple Salesforce 0.450 Red
14 YouTube 0.571 Purple YouTube 0.434 Green
15 Collaboration 0.560 Red Project management 0.433 Blue
16 Excel 0.552 Orange Social media management 0.416 Red
17 Digital advertising 0.536 Red Facebook Ads 0.406 Red
18 Social media management 0.519 Red Excel 0.401 Orange
19 Content marketing 0.516 Red Marketing strategies 0.400 Blue
20 PowerPoint 0.499 Orange Hootsuite 0.395 Red
21 Twitter 0.496 Purple Social media marketing 0.382 Red
22 Email campaigns 0.483 Red Marketing automation 0.377 Red
23 Social media channels 0.479 Purple Microsoft Office 0.369 Orange
24 HTML 0.475 LBlue Marketing campaigns 0.360 Blue
25 Microsoft Office 0.470 Orange Word 0.344 Orange
26 Content management 0.465 Red Content marketing 0.330 Red
27 Marketing campaigns 0.455 Red Pinterest 0.319 Green
28 Hootsuite 0.444 Blue Content management 0.311 Red
29 Instagram 0.439 Purple Snapchat 0.297 Green
30 Campaign management 0.426 Red Presentation 0.272 Blue
31 Marketing strategies 0.418 Red PowerPoint 0.266 Orange
32 Photoshop 0.417 LGreen Photoshop 0.264 LBlue
33 Content strategy 0.416 Red Tableau 0.264 Red
34 Social media campaigns 0.397 Red Collaboration 0.261 Blue
35 Programming 0.395 Blue Content strategy 0.258 Red
36 Social media marketing 0.375 Red Social media channels 0.257 Green
37 Word 0.373 Orange Web development 0.249 Blue
38 Web development 0.373 LBlue SQL 0.248 Purple
39 Organizational 0.371 Green YouTube Ads 0.247 Red
40 Presentation 0.358 Green HTML 0.245 Purple
41 Pinterest 0.354 Purple Advertising campaigns 0.234 Red
42 SQL 0.348 Blue Team player 0.228 Blue
43 Adobe Analytics 0.346 Blue Adobe Analytics 0.225 Red
44 Advertising campaigns 0.344 Red InDesign 0.222 LBlue
45 Tableau 0.339 Blue Online marketing 0.222 Red
46 Facebook Ads 0.330 Blue Organizational 0.217 Blue
47 Analytics 0.321 Green Instagram Ads 0.211 Red
48 Marketing experience 0.311 Red Email campaigns 0.200 Blue
49 Media buying 0.296 Red Campaign management 0.192 Red
50 Strategy development 0.294 Red Marketing communication 0.182 Blue
51 Data management 0.292 Red Media buying 0.182 Green
52 InDesign 0.285 LGreen Marketing management 0.181 Blue
53 Social media strategies 0.284 Red Illustrator 0.180 LBlue
54 Paid social media 0.280 Red Social media content 0.178 Blue
55 Social media accounts 0.279 Purple Data science 0.176 Blue
56 Marketing communication 0.277 Red Interpersonal 0.173 Blue
57 Team player 0.272 Red Social media accounts 0.168 Green
58 Business development 0.271 Red Paid social media 0.168 Red
59 Adobe Creative 0.262 Orange Data analytic 0.168 Blue
60 Brand management 0.260 Red Twitter Ads 0.168 Red
61 Digital content 0.259 Green Email marketing campaigns 0.164 Red
62 CSS 0.257 LBlue Time management 0.155 Blue
63 Customer experience 0.256 Red Analytics 0.154 Blue
64 Marketing plans 0.256 Red Basic html 0.154 Purple
65 Verbal 0.255 Green Business development 0.153 Blue
66 Web content 0.253 Red Digital advertising 0.152 Red
67 Business intelligence 0.249 Red Adobe Creative 0.148 Orange
68 Online marketing 0.241 Red Programming 0.145 Purple
69 Basic html 0.238 LBlue Social media presence 0.143 Green
70 User experience 0.238 Red CSS 0.137 Purple
71 Marketing management 0.235 Red Google Tag 0.134 Red
72 Brand awareness 0.234 Red Sprinklr 0.132 Red
73 Email marketing campaigns 0.231 Red Verbal 0.131 Blue
74 Digital channels 0.229 Red Data visualization 0.130 Red
75 Social media content 0.228 Red Oral 0.130 Blue
76 Data visualization 0.228 Blue User experience 0.128 Blue
77 Marketing channels 0.221 LBlue Social media posts 0.120 Blue
78 Social media analytics 0.220 Red Data driven 0.117 Blue
79 Data driven 0.214 Red Digital marketing campaigns 0.111 Red
80 Oral 0.212 Green Python 0.111 Purple
81 Digital marketing experience 0.211 Red Marketing channels 0.110 Blue
82 Google Tag 0.206 Blue Web content 0.110 Blue
83 Illustrator 0.205 LGreen Display network 0.110 Red
84 Digital marketing campaigns 0.202 Red Digital content 0.110 Blue
85 Time management 0.202 Green Social media analytics 0.109 Red
86 Interpersonal 0.200 Green Social media campaigns 0.104 Green
87 Data science 0.200 Red Strategy development 0.104 Blue
88 SAS 0.197 Blue Kenshoo 0.100 Red
89 Python 0.197 Blue Campaign development 0.098 Red
90 Snapchat 0.197 Purple Digital channels 0.097 Blue
91 Campaign development 0.194 Red Integrated marketing 0.092 Blue
92 Social media posts 0.190 Green Nielsen 0.091 Red
93 Management experience 0.184 Blue Customer experience 0.091 Blue
94 Social listening 0.179 Red Facebook Insights 0.090 Red
95 Sprinklr 0.176 Blue Brand management 0.085 Blue
96 Facebook Insights 0.172 Blue Social listening 0.085 Red
97 ComScore 0.170 Blue Business intelligence 0.080 Blue
98 Data mining 0.162 Blue Facebook Page 0.077 Green
99 Display Network 0.161 Blue Microsoft Suite 0.075 Orange
100 Data analytic 0.148 Red marketing plans 0.071 Blue
Note. Red is the color of the largest cluster. Blue is the color of the second-largest cluster. Green is the color of the third largest cluster. Purple is the color of the fourth largest cluster. Orange is the color of the fifth largest cluster. LBlue represents Light Blue that is the color of the sixth largest cluster. LGreen represents Light Green that is the color of smallest cluster.
For the remote positions, the largest clusters took 36.44% of the network and centered around Google Analytics. Different from the in-person job descriptions, this cluster of tools about digital advertising was also strongly associated with skills about marketing automation, email marketing, social media marketing, social media management, and social media analytics. The second largest cluster (34.75%) centered around the skill of research which was closely tied to the skills of marketing strategies, web development, as well as writing. Different from the in-person network, there was not any salient cluster with the skill of communication as a hub. The skill of communication co-occurred most frequently with writing, a salient node in the research cluster, and was strongly associated with the skill of content creation. Similar to the in-person network, Facebook was also the hub of the third largest cluster (11.86%) and closely tied to a set of social media tools and skills. The fourth largest cluster (8.47%) centered around the tool of SQL, clustering with the tools of HTML, CSS, SAS, SPSS, Python, and programming skill. Compared with the in-person network, the remote network also had a relatively smaller Microsoft Office cluster (5.08%) and a larger Adobe Creative Suite cluster (3.39%).
Discussion
Through text mining and semantic network analysis, this study compared the in-person and remote jobs requiring media analytics to inform educators about integrating analytics in the media and communications curriculum. The findings indicate that both the in-person and remote positions emphasized the competencies of conducting research, developing strategies, communication, and collaborations as well as the mastery of the tools about data analysis, digital marketing, and content creation. These expectations are consistent with the CPRE’s (2018) and ACEJMC’s (2012) emphasis on integrating research, analysis, and data into media and communications curriculum. However, these expectations also suggest that knowledge and skills about data and analytics are rarely sought alone. Instead, employers expect the candidates to apply them in conjunction with other media and communication domain-specific knowledge to drive strategy development and decision-making, which is rooted in the original definition of analytics (Dinsmore, 2016).
Similar to the earlier research by Stansberry and MacKenzie (2020), communication emerged as a top skill desired by both types of positions. However, the nature of the working mode influenced which aspect of communication they emphasized on (Coate, 2021). Specifically, the in-person positions emphasized more on verbal, interpersonal, and organizational communication, whereas the remote positions asked more for written communication. In addition, while the in-person positions had higher expectations on the general data management and analysis tools, the remote positions underlined the use of analysis and advertising tools related to specific social media platforms and search engines.
Informed by these findings, the programs in media and communications should consider incorporating the following topics into the curriculum to prepare students for positions requiring media analytics. First, considering the high mention frequency of Excel in the descriptions of in-person positions, the use of spreadsheets should be integrated into the existing curriculum to lay the foundation for managing large datasets and creating effective visualizations for reporting. For example, Excel labs can be integrated into writing classes to introduce how to match various types of audience datasets with appropriate visualizations for creating business and research reports. As called by ACEJMC (2012), students need to be equipped with the knowledge of basic statistics and metrics for describing the audience’s interaction with traditional and emerging media. More importantly, the most salient skills that distinguish communication students from computer and data science students are storytelling, presentation, and collaboration. Group work should be provided to allow students to collaboratively analyze the user data. Plenty of oral presentation opportunities are needed for students to practice using data visualizations to communicate insights and better manage brands.
The COVID-19 pandemic not only increased the share of remote internships as reported by Indeed Hiring Lab (Konkel, 2021) but also led to more remote internship opportunities in digital and social media management and marketing, which are highly relevant to college students majoring in media and communications. To get them prepared with essential analytical skills for this kind of internship, we need to teach analytics beyond the spreadsheets and integrate more hands-on activities about social listening using both popular free platforms, such as Facebook Ads & Insights, Instagram Ads & Insights, Twitter Ads & Analytics, and YouTube Ads & Insights, as well as paid analytics platforms, such as Brandwatch. It is even more helpful to dedicate some courses to social listening and social media content creation. Furthermore, students need to be equipped with the ability to use Google Analytics to track and understand media users’ digital footprints on a variety of social media platforms, employing data to gauge audience engagement (e.g., acquisition and conversion), and developing/optimizing social media marketing and campaign strategies. For example, students could work on a project to first conduct market research through social listening, develop a social media marketing plan and/or campaign through content creation, and then use Google Tags and Google Analytics to track and assess the effectiveness of the plan.
Aside from informing curriculum development, the findings of this study also shed light on some actions to be taken by student professional development centers (SPDCs) to better prepare students for careers requiring media analytics. For example, SPDCs could use the keywords and key phrases of skills and tools identified in this study (listed in Tables 4 and 5) to help students better understand the employers’ expectations. By using the target keywords and key phrases recognized by employers, students will be able to build effective resumes and develop coherent personal narratives to demonstrate their knowledge and skills about media analytics in interviews. Media analytics is a fast-growing field. Expanding students’ connection with employers will allow them to obtain update-to-date know-hows from the field.
There are several limitations to this study. First, although text mining and semantic network analysis allow us to examine a large number of job ads, it is limited to quantitative analysis of keywords and key phrases. Future research could integrate qualitative analysis to gain a more nuanced understanding of the industry expectations. Second, the data were collected during the first and second waves of the COVID-19 pandemic in the United States when most companies were functioning in the remote mode. Future research could expand the data collection period to compare the in-person and remote positions at different time points, such as before, during, and post the pandemic, or at different waves. This comparative research could further enrich the findings discovered in this study.
Author Biographies
Ke Jiang (Ph.D.) is an assistant professor of media analytics in the School of Communications at Elon University. She is a computational social scientist in the areas of network analysis, natural language processing, and social media analytics. Her research interests focus on the media effects on the macro-society level, as well as the evolution of digital culture and humanity.
Qian Xu (Ph.D.) is a professor of strategic communications in the School of Communications at Elon University. She teaches courses on media analytics and strategies for interactive media and social media. Her research interests focus on the social and psychological effects of media technology, as well as the pedagogy of high impact practices.
Ashleigh Afromsky is a recent graduate from Elon University with bachelor degrees in Media Analytics and Communication Design.
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding: The author(s) received no financial support for the research, authorship, and/or publication of this article.
==== Refs
References
Accrediting Council on Education in Journalism and Mass Communications. (2012). A guide to assessment of learning outcomes for ACEJMC accreditation. http://www.acejmc.org/wp-content/uploads/2016/08/ACEJMC-Guide-to-Assessment-of-Learning-Outcomes.pdf
Adams M. B. Lee N. M. (2021). Analytics in PR education: Desired skills for digital communicators. Journal of Public Relations Education, 7 (2 ), 203–212. https://aejmc.us/jpre/2021/09/10/analytics-in-pr-education-desired-skills-for-digital-communicators/
Barnett G. A. Woelfel J. (Eds.). (1988). Readings in the Galileo system: Theory, methods and applications. Kendall/Hunt.
Bartik A. Cullen Z. Glaeser E. Luca M. Stanton C. (2020). What jobs are being done at home during the COVID-19 crisis? Evidence from firm-level surveys (NBER Working Paper No. 27422). National Bureau of Economic Research. https://www.nber.org/system/files/working_papers/w27422/w27422.pdf
Blondel V. D. Guillaume J. L. Lambiotte R. Lefebvre E. (2008). Fast unfolding of communities in large networks. Journal of Statistical Mechanics, 10 , P10008. 10.1088/1742-5468/2008/10/P10008
Bonacich P. (1972). Factoring and weighting approaches to status scores and clique identification. Journal of Mathematical Sociology, 2 (1 ), 113–120. 10.1080/0022250X.1972.9989806
Borgatti S. Everett M. Freeman L. C. (2002). UCINET: Network analysis software [Computer software]. Analytic Technologies.
Coate P. (2021, January 25). Remote work before, during, and after the pandemic. Quarterly Economics Briefing, Q4. https://www.ncci.com/SecureDocuments/QEB/QEB_Q4_2020_RemoteWork.html
Collins A. M. Quillian M. R. (1972). Experiments on semantic memory and language comprehension. In Gregg L. W. (Ed.), Cognition in learning and memory (pp. 117–138). Wiley.
Commission on Public Relations Education. (2018). Fast forward: Foundations + future state. Educators + practitioners: The Commission on Public Relations Education 2017 report on undergraduate education. http://www.commissionpred.org/wp-content/uploads/2018/04/report6-full.pdf
Cooper A. (2012). What is “analytics”? Definition and essential characteristics. CETIS Analytics Series, 1 (5 ), 1–10. https://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.258.5595&rep=rep1&type=pdf
D’Angelo P. (2002). News framing as a multiparadigmatic research program: A response to Entman. Journal of Communication, 52 (4 ), 870–888. 10.1111/j.1460-2466.2002.tb02578.x
Daniel B. K. (2015). Big data and analytics in higher education: Opportunities and challenges. British Journal of Educational Technology, 46 , 904–920. 10.1111/bjet.12230
Dinsmore T. W. (2016). Disruptive analytics: Charting your strategy for next-generation business analytics. Apress.
Doerfel M. L. (1998). What constitutes semantic network analysis? A comparison of research and methodologies. Connections, 21 (2 ), 16–26.
Feldman E. (2021). Virtual internships during the COVID-19 pandemic and beyond. New Horizons in Adult Education and Human Resource Development, 33 (2 ), 45–51. 10.1002/nha3.20314
Fillmore C. J. (2008). Frame semantics. In Geeraerts D. (Ed.), Cognitive linguistics: Basic readings (pp. 373–400). De Gruyter.
Freberg K. Kim C. M. (2018). Social media education: Industry leader recommendations for curriculum and faculty competencies. Journalism & Mass Communication Educator, 73 (4 ), 379–391. 10.1177/1077695817725414
Freeman L. C. (1978). Centrality in social networks conceptual clarification. Social Networks, 1 (3 ), 215–239.
Grady D. A. (2020). Media analytics and audiences. In Grady D. A. Hollifield C. A. Sturgill A. (Eds.), The golden age of data: Media analytics in study and practice (pp. 3–17). Routledge.
Granados N. (2019, December 9). Vision 2020: How analytics will transform media and entertainment in the next decade. Forbes. https://www.forbes.com/sites/nelsongranados/2019/12/09/vision-2020-how-analytics-will-transform-media-and-entertainment-in-the-next-decade/?sh=387cd41262a6
Hawkins B. L. (2008). Accountability, demands for information, and the role of the campus IT organization. In Katz R. N. (Ed.), The tower and the cloud (pp. 98–104). EDUCAUSE.
Hollifield C. A. (2020). The pioneers of media analytics in higher education. In Grady D. A. Hollifield C. A. Sturgill A. (Eds.), The golden age of data: Media analytics in study and practice (pp. 31–40). Routledge.
Hora M. T. Lee C. Chen Z. Hernandez A. (2021). Exploring online internships amidst the COVID-19 pandemic in 2020: Results from a mixed-methods study. Center for Research on College-Workforce Transitions, University of Wisconsin–Madison.
Hunter S. (2014). A novel method of network text analysis. Open Journal of Modern Linguistics, 4 (2 ), 350–366.
Jacomy M. Venturini T. Heymann S. Bastian M. (2014). ForceAtlas2, a continuous graph layout algorithm for handy network visualization designed for the Gephi software. PLOS ONE, 9 (6 ), Article e98679. 10.1371/journal.pone.0098679
Jiang K. Barnett G. A. Taylor L. D. (2016). News framing in an international context: A semantic network analysis. International Journal of Communication, 10 , 3710–3736.
Kapatamoyo M. (2019). Data analytics in mass communication: New methods for an old craft. Iberian Journal of Information Systems and Technologies, 20 , 504–515.
Konkel A. (2021, April 27). Remote internships jump during pandemic. Indeed Hiring Lab. https://www.hiringlab.org/2021/04/27/remote-internships-jump-during-pandemic/
Kwon K. H. Bang C. C. Egnoto M. Raghav Rao H. (2016). Social media rumors as improvised public opinion: Semantic network analyses of twitter discourses during Korean saber rattling 2013. Asian Journal of Communication, 26 (3 ), 201–222. 10.1080/01292986.2015.1130157
Liu W. Lai C. H. Xu W. W. (2018). Tweeting about emergency: A semantic network analysis of government organizations’ social media messaging during hurricane Harvey. Public Relations Review, 44 (5 ), 807–819. 10.1016/j.pubrev.2018.10.009
Lusinski N. Ward M. (2020, April 21). The 13 most common work-from-home jobs in America, and how much money you can make doing them. Business Insider. https://www.businessinsider.com/jobs-work-from-home-2019-3
Manovich L. (2018). 100 billion data rows per second: Media analytics in the early 21st century. International Journal of Communication, 12 , 473–488. https://ijoc.org/index.php/ijoc/article/view/6160/2249
Martin A. (2021, August 26). The pandemic made internships hypercompetitive. CNBC. https://www.cnbc.com/2021/08/26/the-pandemic-made-internships-hypercompetitive.html
Meng J. Jin Y. Lee Y.-I. Kim S. (2019). Can Google Analytics certification cultivate PR students’ competency in digital analytics? A longitudinal pedagogical research. Journalism & Mass Communication Educator, 74 (4 ), 388–406. 10.1177/1077695818816916
Meng J. Reber B. H. Berger B. K. Gower K. K. Zerfass A. (2021). North American Communication Monitor 2020-2021. The impact of COVID-19 pandemic, ethical challenges, gender issues, cyber security, and competence gaps in strategic communication. The Plank Center for Leadership in Public Relations. http://plankcenter.ua.edu/wp-content/uploads/2021/06/NACM-Report-2020-21-1.pdf
Neill M. S. Schauster E. (2015). Gaps in advertising and public relations education: Perspectives of agency leaders. Journal of Advertising Education, 19 (2 ), 5–17. 10.1177/109804821501900203
O’Boyle J. Sturgill A. (2020). Prevalence of media analytics content in accredited colleges and universities. In Grady D. A. Hollifield C. A. Sturgill A. (Eds.), The golden age of data: Media analytics in study and practice (pp. 137–145). Routledge.
Parker K. Horowitz J. M. Minkin R. (2020, December 9). How the coronavirus outbreak has—and hasn’t—changed the way Americans work. Pew Research Center. https://www.pewresearch.org/social-trends/2020/12/09/how-the-coronavirus-outbreak-has-and-hasnt-changed-the-way-americans-work/
PwC. (n.d.). Data and analytics in the entertainment and media sector. https://www.pwc.co.uk/issues/data-analytics/industries/data-and-analytics-in-the-entertainment-and-media-sector.html
Stansberry K. MacKenzie M. (2020). Equipping the media analytics toolbox: A study of the skills required for entry and mid-level media analytics jobs. In Grady D. A. Hollifield C. A. Sturgill A. (Eds.), The golden age of data: Media analytics in study and practice (pp. 56–69). Routledge.
Tandoc E. C. Jr. (2019). Analyzing analytics: Disrupting journalism one click at a time. Routledge.
| 0 | PMC9703025 | NO-CC CODE | 2022-11-29 23:21:06 | no | 2022 Nov 24;:10776958221135969 | utf-8 | null | null | null | oa_other |
==== Front
Cross Cult Res
Cross Cult Res
spccr
CCR
Cross-Cultural Research
1069-3971
1552-3578
SAGE Publications Sage CA: Los Angeles, CA
10.1177_10693971221141478
10.1177/10693971221141478
Original Research Article
How National Culture Influences the Speed of COVID-19 Spread: Three Cross-Cultural Studies
https://orcid.org/0000-0003-0487-7814
Huang Xiaoyu 1†
Gupta Vipin 1†
Feng Cailing 2
https://orcid.org/0000-0003-4385-2011
Yang Fu 3
Zhang Lihua 4
Zheng Jiaming 4
https://orcid.org/0000-0001-9243-4479
Van Wart Montgomery 1
1 122429 California State University, San Bernardino College of Business and Public Administration , San Bernardino, CA, USA
2 School of Public Management, 70578 Nanjing Agricultural University , Nanjing, China
3 School of Business Administration, 12603 Southwestern University of Finance and Economics , Chengdu, China
4 School of Labor and Human Resources, 12471 Renmin University of China , Beijing, China
Cailing Feng, School of Public Management, Nanjing Agricultural University. 1 Jinling Rd, Xuanwu District, Nanjing, Jiangsu Province, China. 210095. Email: [email protected]
† Xiaoyu Huang and Vipin Gupta contributed equally to this work.
25 11 2022
25 11 2022
10693971221141478© 2022 SAGE Publications
2022
SAGE Publications
This article is made available via the PMC Open Access Subset for unrestricted re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the COVID-19 pandemic or until permissions are revoked in writing. Upon expiration of these permissions, PMC is granted a perpetual license to make this article available via PMC and Europe PMC, consistent with existing copyright protections.
The COVID-19 pandemic has affected 222 countries and territories around the globe. Notably, the speed of COVID-19 spread varies significantly across countries. This cross-cultural research proposes and empirically examines how national culture influences the speed of COVID-19 spread in three studies. Study 1 examines the effects of Hofstede’s national cultural dimensions on the speed of COVID-19 spread in 60 countries. Drawing on the GLOBE study (House et al., 2004), Study 2 investigates how GLOBE cultural dimensions relate to the speed of the pandemic’s spread in 55 countries. Study 3 examines the effect of cultural tightness in 31 countries. We find that five national cultural dimensions – power distance, uncertainty avoidance, humane orientation, in-group collectivism, and cultural tightness – are significantly related to the speed of COVID-19 spread in the initial stages, but not in the later stages, of the pandemic. Study 1 shows that the coronavirus spreads faster in countries with small power distance and strong uncertainty avoidance. Study 2 supports these findings and further reveals that countries with low humane orientation and high in-group collectivism report a faster spread of the disease. Lastly, Study 3 shows that COVID-19 spreads slower in countries with high cultural tightness.
cross-cultural study
national culture
speed of COVID-19 spread
power distance
uncertainty avoidance
humane orientation
in-group collectivism
cultural tightness
National Social Science Foundation of China No.22BGL140 edited-statecorrected-proof
typesetterts10
==== Body
pmcIntroduction
The COVID-19 pandemic has spread across 222 countries and territories worldwide since the first reported case at the end of 2019. Notably, the speed of COVID-19 spread varies significantly across the globe, which raises the important question of why COVID-19 spreads faster in certain countries and slower in others. It could be broadly explained by how individuals react toward the pandemic at the micro level as well as how macro-level forces, such as government interventions and lockdowns, jointly influence the actions and behaviors of individuals within a country (Dizikes, 2020; Mayer et al., 2020). Much of the discussion on national differences in COVID-19-related outcomes has centered on political and economic factors (Tisdell, 2020), and little is known about how national culture affects the speed of COVID-19 spread. Culture is defined as “the collective programming of the mind which distinguishes the members of one category of people from another” (Hofstede, 1984). National culture reflects the different patterns of beliefs and behaviors that vary across countries, and therefore culture may predict individuals’ or groups’ actions and reactions toward the pandemic. We argue that national culture serves as an important determinant of the between-country variations in the speed of COVID-19 spread since national culture has been shaping “collective actions and norms” during the pandemic (Guan et al., 2020). For instance, the United States Centers for Disease Control and Prevention (2022) suggests that the best way to prevent the disease is through vaccination and social distancing, which limits individuals’ exposure to the coronavirus. People in different cultures may react differently to social distancing, lockdown measures, and other COVID-19-related national health guidelines. Consistent with the cultural perspective that national culture has been a significant determinant of individuals’ actions during the pandemic and subsequent COVID-19-related outcomes, a recent cross-cultural study showed that people in the United States view their self-serving acts, such as social gatherings, as more acceptable, while such a self-interest bias is not found in China (Dong et al., 2021). Similarly, another recent study found that power distance is negatively related to COVID-19 morbidity and mortality (Kumar, 2021). Nevertheless, despite the importance of national culture, only a few studies have examined the effects of culture on the speed of COVID-19 spread. For example, one recent study suggested that relational mobility – the extent to which it is easy “to form new relationships and terminate current ones” – significantly predicts the speed of COVID-19 spread, such that countries with high relational mobility report a faster spread of the virus (Salvador et al., 2020). Another study showed that individualism and indulgence are positively related to the increase rate of total COVID-19 cases in European countries (Gokmen et al., 2021). However, other effects of national cultural dimensions remain largely unexplored.
We extend the cultural perspective by building on the path-breaking work of Hofstede (2001), (Hofstede & Bond, 1988), and Gelfand et al. (2011) to systematically examine the vital question of how differences in national culture explain the variations in the speed of COVID-19 spread across countries. We identify five important national cultural dimensions – power distance, uncertainty avoidance, humane orientation, in-group collectivism, and cultural tightness – that may influence the speed of COVID-19 spread and empirically test our hypotheses in three studies. Study 1 investigates how power distance and uncertainty avoidance relate to the speed of the virus spread using a sample of 60 countries. The results suggest that small power distance and strong uncertainty avoidance are related to faster disease spread at the beginning of the pandemic. Using a sample of 55 countries, Study 2 adopts the national cultural dimensions from House et al.’s (2004) GLOBE study and supports the findings of Study 1. Moreover, Study 2 reveals that countries with low humane orientation and high in-group collectivism report a faster spread of the coronavirus. Study 3 focuses on cultural tightness-looseness and shows that the speed of coronavirus spread is slower in countries with high cultural tightness.
Theory and Hypotheses
Power Distance
Power distance refers to the extent to which less powerful individuals “accept and expect that power is distributed unequally” (Hofstede et al., 2010). Similarly, the GLOBE project defined power distance as “the degree to which members of an organization or society expect and agree that power should be unequally shared” (House et al., 2002). Respect for authority and centralization are favored in large power distance countries, while independence and decentralization are valued in small power distance countries. People in large power distance societies believe that “whoever holds the power is right and good,” but people in small power distance cultures believe that “the use of power should be legitimate and follow criteria of good and evil” (Hofstede et al., 2010). These differences may translate into distinct opinions and actions toward the pandemic and subsequently affect the speed of coronavirus spread within a country.
Since people in large power distance countries are more likely to respect and follow COVID-19-related regulations such as social distancing and lockdown rules while people in small power distance countries value their individuality and independence over authorities, leading to the faster spread of the disease, we expect that power distance is negatively related to the speed of COVID-19 spread. Moreover, large power distance cultures generally have greater socioeconomic class inequalities (Carl et al., 2004). In such cultures, decision-making power is more likely to be centralized. During the pandemic, governments have used their authoritarian power expeditiously, and the people in large power distance cultures expected and accepted this since they are more tolerant of hierarchies. Furthermore, one study showed that in large power distance countries, patients report shorter medical consolation times since, compared to people in small power distance countries, they tend to respect and trust their doctors (Meeuwesen et al., 2009). When facing the COVID-19 pandemic, people in high power distance countries such as Russia, Vietnam and China are more likely to conform to authorities and follow the recommendations of medical experts, practicing social distancing, mask wearing, and hand washing, all of which dampen the spread of the disease. Supporting this view, a recent study suggested that Vietnam – a large power distance culture – effectively slowed down the spread of the coronavirus by implementing an early lockdown, introducing mask wearing regulations, and improving the “virality” of health information (Huynh, 2020). Therefore, we expect that people in large power distance countries are more likely to obey social distancing guidelines and to support national efforts to contain the pandemic, both of which contribute to the slower spread of COVID-19.
In contrast, in small power distance cultures, people tend to be suspicious of authority and hierarchy. People in such countries does not readily accept the social hierarchy or tolerate inequalities. They are more likely to question the legitimacy and power of authorities and experts. The relative preference for decentralization and individual freedom and the lack of respect for authority in small power distance countries such as France and Italy make it harder for authorities to convince citizens to follow new rules and regulations and take actions to prevent the spread of the disease. For example, more than 4000 people had been fined for violating the COVID-19 lockdown in France by March 2020 (McAuley, 2020), while the number reached 110,000 in Italy (Duncan, 2020); such rampant violations of COVID-19-related new regulations can speed up the spread of the disease. Taking these findings into account, we argue that people in small power distance countries are less likely to follow new safety guidelines and lockdown rules, thus potentially accelerating the spread of the disease.Hypothesis 1: Power distance is negatively related to the speed of COVID-19 spread.
Uncertainty Avoidance
Uncertainty avoidance is the extent to which the members of a society feel threatened by ambiguous and unknown situations (Hofstede et al., 2010). The GLOBE project defined uncertainty avoidance as “the extent to which members of an organization or society strive to avoid uncertainty by reliance on social norms, rituals, and bureaucratic practices to alleviate the unpredictability of future events” (House et al., 2002). In other words, uncertainty avoidance measures the extent to which a country manages ecosystem ambiguity and change through existing, known technological and organizational solutions (House et al., 2004) instead of engaging in risky behaviors (De Luque & Javidan, 2004). The COVID-19 pandemic has created highly ambiguous and unstructured situations. People in strong uncertainty avoidance countries report relatively higher stress and anxiety and believe that “the uncertainty inherent in life is a continuous threat that must be fought,” while people in weak uncertainty avoidance countries report lower stress and anxiety and believe that “uncertainty is a normal feature of life, and each day is accepted as it comes” (Hofstede et al., 2010).
People in uncertainty avoidance cultures rely on a mix of available resources, such as traditional knowledge, and tradable resources, such as modern technology, to avoid the costs of adverse change (Gupta et al., 2004). They tend to be less comfortable with environmental ambiguity and less tolerant of the uncertainty caused by the COVID-19 pandemic. Consequently, they are more likely to rely on traditional knowledge and known technologies for the initial management of pandemic, leading to over-confidence about their immunity. Prior research shows that knowledge deemed relevant to an issue breeds an overconfidence bias (Fabricius & Buttgen, 2013). Overconfidence based on known solutions that possibly worked in alternative situations makes them less concerned about the new virus and less likely to change their behaviors to follow new social distancing guidelines or lockdown rules, leading to the faster spread of COVID-19. Moreover, uncertainty avoidance is negatively related to individuals’ agreeableness (Hofstede & McCrae, 2004), suggesting that people may be less agreeable to new safety regulations and lockdown rules in strong uncertainty avoidance cultures, consequently accelerating the spread of COVID-19. Furthermore, uncertainty avoidance is positively related to willingness to justify unethical behaviors, such as cheating on taxes, avoiding public transportation fares, and claiming “unentitled government benefits” (Parboteeah et al., 2005). For example, one survey found that 60 percent of people in France – a strong uncertainty avoidance country – defied the lockdown rules and 43 percent invited others into their homes during the second national lockdown (Morrow, 2020). Therefore, we argue that violations of social distancing, masking wearing, and lockdown rules constitute new unethical behaviors amidst the pandemic and that such behaviors would be more prevalent in strong uncertainty avoidance cultures, resulting in a faster spread of the disease.
Conversely, uncertainty-accepting cultures are more tolerant of the chaos and uncertainties caused by the COVID-19 pandemic and more capable of coping with the unprecedented ambiguities. People in such cultures tend to have low confidence in the integrity of existing methods and machinery and seek to design new solutions through discovery-oriented planning. When facing substantial uncertainty caused by the COVID-19 pandemic, they are more likely to support new measures outlined in national contagion management guidelines to help curb the spread of the disease. They may engage in proactive social distancing and other preventive behaviors, such as using hand sanitizers, washing hands frequently, and not touching their faces, resulting in a slower spread of the disease. For example, Singapore – a weak uncertainty avoidance country – has taken effective measures to fight the pandemic and has successfully vaccinated more than 80 percent of the general population (Cortez et al., 2021).Hypothesis 2: Uncertainty avoidance is positively related to the speed of COVID-19 spread.
Humane Orientation
Humane orientation is defined as “the degree to which individuals in organizations or societies encourage and reward individuals for being fair, altruistic, friendly, generous, caring, and kind to others” (House et al., 2004), which is similar to Hofstede and Bond’s (1988) Kind Heartedness dimension. For example, Irish people living in Ireland, a country with a high humane orientation, have donated more than 2.5 million euros to native American tribes to help them cope with the devasting effects of COVID-19 (McGreevy, 2020). People in humane-oriented cultures tend to be concerned about the well-being of the socioeconomically weak members of the national community, take responsibility for others’ health and well-being, and benevolently provide social support to others and help solve their problems through personal engagement, even if they are strangers (Kabasakal & Bodur, 2004). Once they are aware of the gravity of the life-changing challenge faced by the socioeconomically vulnerable sections of their society, they tend to find channels to become personally involved in order to be a part of a solution. Since countries with a high humane orientation encourage, pursue, and reward generosity, kindness, and altruism (House et al., 2004), we expect that when facing the COVID-19 pandemic, people in such countries are more likely to be considerate and to altruistically participate in social distancing and other measures to help curtail the spread of the deadly disease since they tend to genuinely care about the health and well-being of others. Furthermore, research has shown that a humane orientation is positively related to religiosity – “the degree to which religion plays a central role in the lives of societal members” (Schlösser et al., 2013). Religiosity can be manifested as showing compassion to others and benefiting others altruistically (Wuthnow, 1991), both of which may help to dampen the spread of the disease because such personal characteristics may lead to prosocial behaviors such as mask wearing, practicing social distancing, avoiding crowds, taking care of the needy, and proactively helping vulnerable neighbors with grocery shopping during the pandemic. Consistent with this perspective, recent studies showed that religiosity is related to stronger prosocial COVID-19 responses across countries (Romano et al., 2021) and that moral considerations predict higher acceptance of societal disease-prevention regulations (Zhu et al., 2021) on dimension that matters most (staying at home). All of the above attitudes and behaviors may lead to a slower spread of COVID-19.
By contrast, people in low humane-oriented cultures believe that people are responsible for their own problems and that institutions are accountable for helping people to be responsible, and therefore they focus more on institutionalized human rights (Gupta et al., 2004). As a result, they are less likely to participate in staying at home and other disease contagion management initiatives proactively and selflessly, which subsequently leads to the faster spread of the disease. Moreover, people in low humane-oriented countries are more willing to justify their unethical behaviors (Parboteeah et al., 2005), such as not wearing masks and ignoring social distancing rules, which lead to the faster spread of the disease.Hypothesis 3: Humane orientation is negatively related to the speed of COVID-19 spread.
In-Group Collectivism
In-group collectivism refers to “the degree to which individuals express pride, loyalty, and cohesiveness in their organizations or families” (House et al., 2002). The GLOBE study assessed in-group collectivism by measuring the extent to which parents and children take pride in each other’s accomplishments, as well as the extent to which “aging parents generally live at home with their children” and “children generally live at home with their parents until they get married” within a society (House et al., 2004). Brewer and Venaik (2011) suggested that GLOBE’s in-group collectivism should be relabeled as “family collectivism” since the questions in the GLOBE study reflect family orientation and show a strong correlation with the “strength of family ties” and “respect for family and friends” dimensions reported in Gelfand et al. (2004).
We expect that people in high in-group collectivism cultures are more likely to have family gatherings and live in multigenerational households, both of which may significantly accelerate the spread of COVID-19. For example, people in family-oriented cultures are concerned about the well-being of the vulnerable members of their family and kinship group and the integrity of the whole family, even if that entails compromising due process (Gupta et al., 2004). The close family relationship may lead to higher occurrences of family gatherings and close social contacts within families, which may result in the faster spread of the disease. For example, in high in-group collectivism countries such as the Philippines and India, living in multigenerational households increases the probability of vulnerable family members such as elderly and those with comprised immune systems being exposed to the coronavirus, and the high occurrence of large family gatherings may accelerate the spread of COVID-19. In comparison, people in low in-group collectivism cultures such as New Zealand, Sweden, Denmark, and Czech Republic tend to focus on the integrity of the process that empowers them to be excellent, functional, and healthy. Compared with people in high in-group collectivism countries, they may be less likely to physically meet their close or extended family members during the pandemic, which ultimately reduces social gatherings and slows down the spread of the virus.Hypothesis 4: In-group collectivism is positively related to the speed of COVID-19 spread.
Cultural Tightness
Tight cultures establish clear social norms that are imposed on individuals (Pelto, 1968) suggests that in countries with high cultural tightness, “little deviation from normative behavior is tolerated, and severe sanctions are administered to those who deviate.” Gelfand et al. (2006) suggest that tightness-looseness comprises two dimensions: the strength of social norms and the level of tolerance for deviance from such norms. Particularly, compared with countries with a loose culture, countries with a tight culture report stronger norms and a lower tolerance of deviant behaviors (Gelfand et al., 2011). In short, in loose cultures, heterogeneity is common and deviations from the social norms are accepted. India, Malaysia, Singapore, and South Korea are examples of “high tightness” countries, while Brazil, New Zealand, Israel, and Venezuela are examples of “high looseness” countries.
When facing the COVID-19 pandemic, countries with high cultural tightness would report a slower spread of the disease because such nations are better at creating strong new social norms to dampen the spread of the disease, such as social distancing and mask wearing, punishing behavior that deviates from such norms, and enhancing social coordination to competently slow down the spread of the disease (Gelfand et al., 2011). Since people in tight cultures are “less willing to live near dissimilar others” and less tolerant of moral deviations (Uz, 2015), they may be less likely to become infected with COVID-19 through interacting with others and more likely to obey new safety rules and regulations and behave responsively to help stop the spread of the disease. Supporting this view, one recent study showed that countries with “high cultural looseness” reported 4.99 times the number of COVID-19 cases and 8.71 times the number of deaths than countries with “high cultural tightness” because nations with high cultural tightness are “more willing to abide by cooperative norms” which are essential for curbing the spread of the pandemic (Gelfand et al., 2021). Furthermore, people in tight cultures tend to have interdependent self-concepts, while those in loose cultures tend to hold independent self-concepts (Carpenter, 2000). The sense of interdependence in a tight culture may unite the nation to fight against the coronavirus as people understand the interdependence of individuals within the country and see themselves as an important part of the national pandemic control efforts, while the value of independence in loose cultures can lead to violations of COVID-19-related rules which may accelerate the spread of the disease. Therefore, we argue that the spread of the COVID-19 pandemic is slower in countries with tight cultures since individuals in such cultures tend to conform to new social norms due to the COVID-19 pandemic, such as social distancing, avoiding social gatherings, and mask wearing, and to be less tolerant of behaviors that deviate from these new norms, both of which slow the spread of the coronavirus.Hypothesis 5: Cultural tightness is negatively related to the speed of COVID-19 spread.
Overview of the Present Study
We used three studies to examine how cultural dimensions relate to the speed of COVID-19 spread. Study 1 used a sample of 60 countries to test Hypotheses 1 and 2 with Hofstede’s cultural dimensions. Although Hofstede’s data have a dominant place in the cultural studies (House et al., 2004), researchers (Minkov & Kaasa, 2021; Venaik & Brewer, 2010) have pointed out methodological problems with Hofstede’s work. GLOBE is an alternative framework that addresses some of these limitations (House et al., 2004). Hanges and Dickson (2004) report that Hofstede’s power distance and uncertainty avoidance indices are correlated 0.61 and −0.61 with the GLOBE’s power distance and uncertainty practices constructs, but only −0.03 and 0.32 with the value constructs. Therefore, Study 2 adopted GLOBE cultural practice dimensions to test the cultural impacts proposed in Hypotheses 1 to 4. Multi-dimensional cultural frameworks are subject to two limitations (see, e.g., Pattine et al., 2009): first, multiple dimensions introduce the possibility of inflated Type 1 error due to repeated testing of the latent cultural factor; second, results are never consistent across different patent dimensions, confounding the robustness of findings—why do the effect sizes and directions vary for different predictors. Gelfand et al. (2021) show that a unidimensional cultural construct—cultural tightness—alone predicts COVID case load and mortality. Therefore, Study 3 focused on the single cultural dimension of tightness-looseness and tested Hypothesis 5 using a sample of 31 countries. Our final samples for the three studies included all countries with valid national culture measures. In all three studies, we operationalized the speed of COVID-19 spread for a particular country in three ways: (1) the number of days taken for the number of COVID-19 cases to grow from x cases to y cases and (2) average daily cases and (3) average daily case growth rate (i.e., the number of new COVID-19 cases divided by the number of days used) during that period. Fewer number of days taken for the number of COVID-19 cases in a country to grow from x cases from y cases suggests faster spread of the disease. Higher average daily cases and higher average daily case growth rates also indicate the faster spread of the disease. Table 1 presents the phases and time intervals used in all three studies.Table 1. Summary of Phases Used in Regressions.
Regression models Time period
Entry phase: 0–4000 cases
Model 1 From the 1st reported cases to 1,000th cases
Model 2 From the 1,001th reported cases to 2,000th cases
Model 3 From the 2,001th reported cases to 3,000th cases
Model 4 From the 3,001th reported cases to 4,000th cases
Takeoff phase: 4,000–10,000 cases
Model 5 From the 4,001th reported cases to 5,000th cases
Model 6 From the 5,001th reported cases to 6,000th cases
Model 7 From the 6,001th reported cases to 7,000th cases
Model 8 From the 7,001th reported cases to 8,000th cases
Model 9 From the 8,001th reported cases to 9,000th cases
Model 10 From the 9,001th reported cases to 10,000th cases
Growth phase: 10,000–100,000 cases
Model 11 From the 10,001th reported cases to 100,000th cases
Maturity phase: 100,000–1,000,000 cases
Model 12 From the 100,001th reported cases to 1,000,000th cases
Proliferation phase: 1,000,0000–10,000,000 cases
Model 13 From the 1,000,001th reported cases to 10,000,000th cases
The data for this study came from three sources: COVID-19 data (updated to November 14, 2021) and national demographic and health-related characteristics control variables came from the Our World COVID-19 dataset provided by the European Centre for Disease Prevention and Control (ECDC). We also used World Bank data on GDP as measured in current U.S. dollars; health expenditure is shown as a percentage of GDP. Appendix Table 1 presents all variable sources and definitions. Appendix Table 2 tabulates the countries included in the three studies. Table 2–4 show the descriptive statistics, and Tables 5–11 present the regression results of the three studies.Table 2. Descriptive Statistics for Study 1 Hofstede Analyses.
Variables Mean Std. Dev 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
1. Power distance 58.33 20.66 1.00
2. Individualism 46.58 23.62 −0.66*** 1.00
3. Uncertainty avoidance 48.88 20.13 0.15 0.03 1.00
4. Masculinity 67.23 22.47 0.23* −0.22* 0.04 1.00
5. Long-term orientation 48.84 22.5 0.03 0.14 0.02 −0.01 1.00
6. Indulgence 48.01 22.38 −0.30** 0.14 0.09 −0.11 −0.53*** 1.00
7. Days useda 45.17 36.3 0.09 −0.15 −0.04 −0.10 −0.10 0.11 1.00
8. Average daily casesa 38.27 38.89 0.02 −0.08 0.05 −0.15 0.13 −0.12 −0.47*** 1.00
9. Average daily growth ratea 29.98 20.42 −0.20 0.10 −0.10 0.05 −0.04 0.03 −0.62*** 0.57*** 1.00
10. log GDP 26.67 1.55 −0.03 0.23* 0.31** −0.24* 0.17 0.09 −0.30** 0.32** 0.14 1.00
11. log population 16.9 1.74 0.34*** −0.20 0.29** −0.14 −0.01 −0.14 −0.20 0.32** 0.07 0.79*** 1.00
12. GDP per capita 30,563 18,659 −0.55*** 0.56*** −0.04 −0.30** 0.24* 0.27** −0.12 −0.03 0.16 0.13 −0.43*** 1.00
13. Percentage aged 65 older 14.53 5.78 −0.47*** 0.58*** −0.10 0.20 0.46*** −0.10 −0.09 −0.12 −0.04 −0.01 −0.46*** 0.50*** 1.00
14. Life expectancy 78.48 4.08 −0.54*** 0.55*** −0.04 −0.01 0.20 0.28** −0.15 0.04 0.18 0.17 −0.37*** 0.73*** 0.70*** 1.00
15. Health expenditure as % of GDP 7.59 2.75 −0.58*** 0.62*** −0.03 0.06 0.02 0.35*** −0.07 −0.04 0.10 0.27** −0.15 0.44*** 0.61*** 0.65*** 1.00
16. Human development index .84 .09 −0.58*** 0.68*** −0.10 −0.07 0.29** 0.24* −0.16 −0.04 0.15 0.12 −0.47*** 0.80*** 0.79*** 0.88*** 0.67*** 1.00
17. Diabetes prevalence 7.24 2.52 0.45*** −0.41*** 0.13 −0.07 −0.26** 0.10 0.08 0.14 0.02 0.07 0.27** −0.24* −0.45*** −0.32** −0.23* −0.35*** 1.00
18. Cardiovasc death rate 203 105 0.48*** −0.31** −0.02 0.05 0.10 −0.62*** 0.05 0.00 −0.16 −0.31** 0.11 −0.54*** −0.29** −0.74*** −0.50*** −0.58*** 0.15 1.00
***p < 0.01, **p < 0.05, *p < 0.1, N = 60.
a days used, average daily cases and average daily growth rate from the first case to 1000 cases.
Table 3. Descriptive Statistics for Study 2 Globe Analyses.
Variables Mean Std. Dev 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
1. Power distance 5.17 .42 1.00
2. Institutional collectivism 4.25 .43 −0.44*** 1.00
3. Uncertainty avoidance 4.13 .6 −0.52*** 0.46*** 1.00
4. Future orientation 3.82 .47 −0.52*** 0.49*** 0.75*** 1.00
5. Gender egalitarianism 3.38 .38 −0.27** −0.03 −0.05 −0.05 1.00
6. Assertiveness 4.12 .36 0.18 −0.44*** −0.13 0.03 −0.05 1.00
7. Performance orientation 4.06 .4 −0.38*** 0.50*** 0.61*** 0.63*** −0.30** 0.00 1.00
8. Humane orientation 4.11 .47 −0.15 0.45*** 0.08 0.13 −0.19 −0.39*** 0.32** 1.00
9. In-group collectivism 5.15 .72 0.58*** −0.21 −0.57*** −0.42*** −0.21 0.13 −0.14 0.25* 1.00
10. Days useda 45.31 26.64 0.14 0.00 −0.10 −0.09 0.08 −0.09 −0.03 0.23* 0.18 1.00
11. Average daily casesa 35.79 38.72 −0.11 0.10 0.19 0.04 −0.14 −0.04 0.13 −0.05 0.04 −0.54 1.00
12. Average daily growth ratea 28.04 19.34 −0.21 −0.05 0.10 0.11 −0.13 0.12 0.02 −0.22 −0.16 −0.69 −0.51 1.00
13. log GDP 26.68 1.67 −0.05 0.17 0.24* 0.26* −0.11 −0.03 0.22* −0.20 −0.28** −0.50*** 0.38*** 0.24* 1.00
14. log population 17.11 1.54 0.27** 0.02 −0.08 0.03 −0.26* 0.02 0.15 0.07 0.23* −0.26* 0.34** 0.04 0.73*** 1.00
15. GDP per capita 29,035 22,117 −0.39*** 0.31** 0.43*** 0.36*** 0.09 −0.07 0.12 −0.17 −0.51*** −0.26* 0.01 0.28** 0.30** −0.32** 1.00
16. Percentage aged 65 older 12.18 6.63 −0.20 0.03 0.27** 0.16 0.24* −0.02 0.00 −0.50*** −0.55*** −0.24* 0.06 0.13 0.39*** −0.11 0.35*** 1.00
17. Life expectancy 77 6.45 −0.27** 0.10 0.24* 0.15 0.09 −0.10 0.12 −0.35*** −0.45*** −0.41*** 0.13 0.27** 0.38*** −0.22* 0.61*** 0.74*** 1.00
Health expenditure as % of GDP 7.29 2.88 −0.30** −0.08 0.32** 0.24* 0.08 0.03 0.18 −0.39*** −0.69*** −0.18 −0.01 0.11 0.37*** −0.05 0.27** 0.69*** 0.50*** 1.00
18. Human development index .82 .11 −0.40*** 0.20 0.36*** 0.27** 0.24* −0.05 0.14 −0.38*** −0.61*** −0.38*** 0.08 0.29** 0.43*** −0.25* 0.70*** 0.81*** 0.90*** 0.60*** 1.00
19. Diabetes prevalence 7.61 3.42 −0.09 0.09 0.01 0.05 −0.14 −0.07 0.08 0.17 0.25* −0.09 0.08 0.10 0.05 0.04 0.24* −0.30** 0.11 −0.29** −0.01 1.00
20. Cardiovasc death rate 197 110 0.19 −0.01 −0.32** −0.28** 0.08 0.06 −0.13 0.30** 0.49*** 0.32** −0.05 −0.24* −0.29** 0.15 −0.45*** −0.41*** −0.50*** −0.50*** −0.48*** 0.16 1.00
***p < 0.01, **p < 0.05, *p < 0.1, N = 60.
a days used, average daily cases and average daily growth rate from the first case to 1000 cases.
Table 4. Descriptive Statistics for Study 3 Cultural Tightness.
Variables Mean Std. Dev 1 2 3 4 5 6 7 8 9 10 11 12 13
1. Cultural tightness 6.58 2.8 1.00
2. Days useda 35.67 16.59 0.09 1.00
3. Average daily casesa 44.67 49.37 0.10 −0.65*** 1.00
4. Average daily growth ratea 32.92 21.18 0.04 −0.75*** 0.43** 1.00
5. log GDP 27.38 1.53 0.19 0.08 0.28 −0.10 1.00
6. log population 17.37 1.81 0.30* 0.01 0.33* −0.05 0.82*** 1.00
7. GDP per capita 33,473 17,496 0.06 0.21 −0.20 −0.03 0.03 −0.47*** 1.00
8. Percentage aged 65 older 15.24 5.84 −0.32* −0.07 −0.15 −0.05 0.02 −0.33* 0.42** 1.00
9. Life expectancy 79.37 4.56 −0.18 −0.05 −0.07 0.04 0.06 −0.46** 0.74*** 0.72*** 1.00
10. Health expenditure as % of GDP 7.97 3.05 −0.38** −0.06 −0.17 −0.07 0.37** −0.06 0.47*** 0.66*** 0.56*** 1.00
11. Human development index .86 .1 −0.30* −0.01 −0.16 0.03 −0.01 −0.55*** 0.80*** 0.71*** 0.93*** 0.65*** 1.00
12. Diabetes prevalence 7.53 3.04 0.50*** 0.09 0.15 0.03 0.18 0.38** −0.18 −0.61*** −0.41** −0.38** −0.40** 1.00
13. Cardiovasc death rate 178 104 0.00 −0.10 0.12 −0.04 −0.26 0.24 −0.65*** −0.37** −0.82*** −0.44** −0.70*** 0.22 1.00
***p < 0.01, **p < 0.05, *p < 0.1, N = 31.
a days used, average daily cases and average daily growth rate from the first case to 1000 cases.
Table 5. Predictors of Days Used in Study 1.
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13)
Power distance 0.49 −0.02 −0.47 −0.04 −0.14 −0.05 −0.03 −0.23 0.01 0.02 −1.17 −2.64** −3.08
(0.39) (0.43) (0.44) (0.32) (0.12) (0.05) (0.08) (0.19) (0.13) (0.08) (1.37) (1.10) (1.90)
Individualism 0.02 0.39 0.15 −0.04 0.17 −0.04 0.01 −0.15 0.01 0.04 −1.77 −1.25 2.07
(0.36) (0.40) (0.40) (0.29) (0.11) (0.04) (0.07) (0.18) (0.12) (0.07) (1.23) (1.08) (1.64)
Masculinity 0.11 0.08 0.21 −0.01 0.16* 0.07** 0.02 0.04 −0.13 −0.06 −0.21 1.46** −0.22
(0.26) (0.29) (0.29) (0.21) (0.08) (0.03) (0.05) (0.13) (0.09) (0.05) (0.89) (0.72) (1.16)
Uncertainty avoidance −0.64** −0.73** −0.39 −0.06 0.10 −0.05 −0.04 −0.07 −0.04 −0.03 −1.06 0.10 4.00***
(0.28) (0.31) (0.31) (0.23) (0.09) (0.03) (0.06) (0.14) (0.09) (0.06) (0.95) (0.88) (1.06)
Long-term orientation 0.18 0.55 0.28 −0.25 0.02 −0.03 −0.11 −0.23 −0.10 −0.01 1.17 0.53 −1.10
(0.33) (0.36) (0.36) (0.26) (0.10) (0.04) (0.07) (0.16) (0.11) (0.07) (1.11) (0.90) (1.27)
Indulgence 0.13 −0.61 −0.04 −0.14 −0.02 −0.01 −0.02 −0.03 −0.07 −0.04 1.04 −0.60 1.24
(0.45) (0.50) (0.50) (0.36) (0.14) (0.05) (0.09) (0.22) (0.15) (0.09) (1.52) (1.24) (1.62)
log GDP −4.84 12.98 6.93 −7.71 −1.95 1.55 6.01 11.05 7.67 4.66 118.34 136.33* 78.85
(23.27) (25.38) (25.72) (18.64) (7.04) (2.77) (4.79) (11.24) (7.77) (4.73) (78.63) (68.42) (118.72)
log population −8.06 −33.15 −17.89 6.40 −1.29 −3.69 −7.44 −13.08 −8.56 −5.69 −127.75 −160.15** −8.30
(23.60) (25.73) (26.08) (18.90) (7.14) (2.81) (4.85) (11.40) (7.88) (4.81) (79.94) (67.95) (117.68)
GDP per capita −0.00 −0.00*** −0.00** 0.00 0.00 −0.00 −0.00 −0.00 −0.00 −0.00 0.00 −0.00 −0.00
(0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.01)
Percentage aged 65 older 1.74 −3.69 −2.13 2.04 0.05 0.08 −0.24 −0.02 −0.63 −0.49 −2.74 −5.66 −21.26*
(2.32) (2.53) (2.56) (1.86) (0.70) (0.28) (0.48) (1.12) (0.77) (0.48) (7.93) (6.28) (10.87)
Life expectancy −2.49 −0.47 3.66 −2.01 −0.47 −0.55 0.18 −0.32 0.70 0.47 29.50** 26.38** 25.22
(3.58) (3.90) (3.96) (2.87) (1.08) (0.43) (0.74) (1.73) (1.20) (0.73) (12.22) (10.41) (20.10)
Health expenditure as % of GDP 4.07 2.60 −1.41 −4.52* −1.64* −0.26 −0.68 −0.88 −0.57 −0.40 −9.11 −1.90 −28.13**
(3.16) (3.44) (3.49) (2.53) (0.96) (0.38) (0.65) (1.53) (1.05) (0.64) (10.67) (8.46) (10.25)
Human development index −197.08 70.95 59.56 44.63 −13.39 10.39 −18.31 −25.69 −17.13 2.74 −1510.83** −1280.50** 1173.17
(200.57) (218.75) (221.68) (160.69) (60.69) (23.88) (41.27) (96.93) (67.01) (40.47) (673.11) (564.75) (846.53)
Diabetes prevalence 0.11 −0.17 0.97 −0.88 0.45 0.59** 0.34 0.71 1.36* 1.23** 5.65 2.90 −12.11
(2.36) (2.57) (2.61) (1.89) (0.71) (0.28) (0.49) (1.14) (0.79) (0.49) (8.09) (6.72) (8.58)
Cardiovasc death rate −0.14 −0.12 0.01 −0.13 −0.01 −0.00 0.02 0.03 0.03 0.01 0.52 0.84*** 1.01*
(0.11) (0.12) (0.12) (0.09) (0.03) (0.01) (0.02) (0.05) (0.04) (0.02) (0.37) (0.29) (0.53)
Constant 646.47** 377.88 −88.91 293.11 124.87 68.41* −11.44 16.90 −77.66 −54.03 −1809.07* −1515.02 −4352.15**
(311.41) (339.65) (344.19) (249.50) (94.23) (37.08) (64.08) (150.50) (104.04) (63.95) (1063.79) (932.65) (2020.85)
Observations 60 60 60 60 60 60 60 60 60 59 59 55 33
R-squared 0.27 0.43 0.25 0.18 0.37 0.53 0.33 0.26 0.26 0.28 0.43 0.48 0.73
***p < 0.01, **p < 0.05, *p < 0.1. Standard errors in parentheses.
Table 6. Predictors of Average Daily Cases in Study 1.
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13)
Power distance −0.73* −0.75 −2.84 −5.70 −3.66 2.57 1.17 3.45 4.60 3.05 18.60 77.60* 205.03
(0.42) (1.34) (2.29) (5.23) (5.42) (4.25) (3.85) (5.30) (5.84) (5.58) (16.26) (42.72) (141.21)
Individualism −0.42 0.55 −0.24 −1.40 0.30 5.44 3.89 3.33 5.29 6.03 21.68 10.46 368.18***
(0.39) (1.23) (2.11) (4.83) (5.00) (3.91) (3.55) (4.89) (5.38) (5.02) (14.63) (41.88) (122.38)
Masculinity −0.06 −0.35 −1.70 −3.10 −3.01 −3.57 −0.68 0.18 −0.32 −1.08 −2.89 −4.20 −239.30**
(0.28) (0.90) (1.53) (3.50) (3.63) (2.84) (2.58) (3.55) (3.91) (3.63) (10.58) (27.83) (86.06)
Uncertainty avoidance 0.02 1.65* 3.27* 3.83 3.53 5.55* 4.53 2.20 4.22 5.36 14.51 −24.78 −87.95
(0.30) (0.96) (1.64) (3.74) (3.87) (3.03) (2.75) (3.79) (4.17) (3.86) (11.24) (34.14) (78.70)
Long-term orientation 0.25 0.05 1.96 2.01 2.70 5.09 4.25 0.63 1.48 2.27 −13.46 −60.03* 59.90
(0.35) (1.12) (1.91) (4.37) (4.52) (3.54) (3.21) (4.42) (4.87) (4.51) (13.16) (35.04) (94.46)
Indulgence −0.27 −1.68 −3.82 −9.06 −8.83 −5.83 −5.02 −10.37* −11.53* −10.12 −33.64* −85.56* −118.11
(0.48) (1.54) (2.64) (6.02) (6.24) (4.89) (4.43) (6.10) (6.72) (6.22) (18.11) (48.08) (120.52)
log GDP 26.29 45.57 231.71* 468.50 421.61 290.85 164.83 286.95 191.23 209.01 223.58 −376.10 −3342.81
(24.78) (78.90) (135.10) (308.65) (319.70) (250.37) (227.10) (312.69) (344.41) (320.58) (934.23) (2656.71) (8838.43)
log population −16.78 19.37 −108.13 −263.36 −213.78 −108.86 6.91 −98.67 −11.63 −0.86 242.33 2125.35 11,051.86
(25.13) (80.01) (137.00) (313.00) (324.20) (253.89) (230.30) (317.09) (349.26) (325.93) (949.80) (2638.27) (8760.89)
GDP per capita −0.00 0.00 −0.00 −0.01 −0.01 0.00 0.01 0.01 0.01 0.01 0.02 0.02 −1.05*
(0.00) (0.00) (0.00) (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) (0.03) (0.11) (0.50)
Percentage aged 65 older −3.94 −18.15** −29.87** −37.24 −38.29 −27.53 −14.73 −17.80 −23.37 −26.13 −33.21 22.40 866.52
(2.47) (7.87) (13.47) (30.77) (31.87) (24.96) (22.64) (31.17) (34.33) (32.33) (94.23) (243.79) (809.60)
Life expectancy 7.13* 18.05 33.37 59.04 52.15 20.02 −23.98 −48.41 −65.92 −82.75 −329.35** −1259.51*** −3202.43**
(3.81) (12.14) (20.78) (47.47) (49.17) (38.51) (34.93) (48.09) (52.97) (49.82) (145.17) (404.23) (1496.67)
Health expenditure as % of −1.14 13.80 3.41 −5.45 28.42 68.22* 72.78** 65.58 77.52 127.45*** 352.09*** 1075.55*** 2356.51***
GDP (3.36) (10.71) (18.33) (41.88) (43.38) (33.98) (30.82) (42.43) (46.74) (43.52) (126.82) (328.46) (762.88)
Human development index −22.34 197.11 −192.24 −1325.73 −1483.21 −2564.24 −1335.29 −717.89 3.90 −207.36 3528.52 35,551.16 145,081.57**
(213.64) (680.16) (1164.58) (2660.61) (2755.90) (2158.20) (1957.64) (2695.41) (2968.85) (2744.51) (7997.92) (21,927.63) (63,022.69)
Diabetes prevalence 2.54 9.28 26.99* 56.21* 67.27** 36.33 36.78 35.72 23.16 46.39 82.28 60.15 2180.43***
(2.51) (8.00) (13.70) (31.31) (32.43) (25.40) (23.04) (31.72) (34.94) (32.98) (96.11) (260.84) (638.49)
Cardiovasc death rate 0.17 0.30 0.34 −0.12 −0.35 −1.16 −1.43 −2.35 −3.34** −2.31 −6.48 −23.77** −89.02**
(0.12) (0.37) (0.63) (1.44) (1.49) (1.17) (1.06) (1.46) (1.61) (1.50) (4.36) (11.41) (39.50)
Constant −815.71** −2908.04*** −6104.37*** −9947.52** −9404.16** −5533.46 −1968.62 −1194.26 731.65 696.58 11,503.26 46,178.07 24,702.36
(331.70) (1056.05) (1808.17) (4130.98) (4278.93) (3350.92) (3039.51) (4185.02) (4609.57) (4337.45) (12,640.00) (36,212.20) (150,448.67)
Observations 60 60 60 60 60 60 60 60 60 59 59 55 33
R-squared 0.28 0.58 0.61 0.47 0.50 0.63 0.65 0.53 0.51 0.62 0.54 0.62 0.87
***p < 0.01, **p < 0.05, *p < 0.1. Standard errors in parentheses.
Table 7. Predictors of Average Daily Case Growth Rates in Study 1.
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13)
Power distance −0.45** −0.08 −0.13 −0.20 −0.14 0.04 0.03 0.04 0.07 0.04 0.05 0.02** 0.01**
(0.22) (0.12) (0.11) (0.21) (0.22) (0.13) (0.07) (0.08) (0.08) (0.07) (0.05) (0.01) (0.00)
Individualism −0.00 0.04 −0.00 −0.05 −0.03 0.10 0.08 0.04 0.07 0.08 0.06 0.00 0.00
(0.20) (0.11) (0.10) (0.20) (0.20) (0.12) (0.06) (0.07) (0.07) (0.06) (0.04) (0.01) (0.00)
Masculinity −0.13 −0.01 −0.07 −0.12 −0.11 −0.10 −0.02 0.00 −0.00 −0.01 −0.01 −0.00 −0.00
(0.14) (0.08) (0.07) (0.14) (0.15) (0.09) (0.05) (0.05) (0.05) (0.04) (0.03) (0.01) (0.00)
Uncertainty avoidance 0.38** 0.13 0.15* 0.15 0.10 0.12 0.09* 0.03 0.05 0.07 0.04 −0.01 −0.01***
(0.15) (0.09) (0.08) (0.15) (0.15) (0.09) (0.05) (0.06) (0.06) (0.05) (0.03) (0.01) (0.00)
Long-term orientation 0.03 0.00 0.10 0.07 0.05 0.10 0.08 0.01 0.01 0.02 −0.04 −0.02* 0.00
(0.18) (0.10) (0.09) (0.18) (0.18) (0.11) (0.06) (0.07) (0.07) (0.06) (0.04) (0.01) (0.00)
Indulgence −0.01 −0.16 −0.18 −0.34 −0.31 −0.17 −0.09 −0.16* −0.18* −0.14* −0.10* −0.03* −0.00
(0.25) (0.14) (0.12) (0.25) (0.25) (0.15) (0.08) (0.09) (0.09) (0.08) (0.05) (0.01) (0.00)
log GDP −14.30 5.76 9.98 18.28 17.77 11.53 3.45 5.16 3.91 2.81 0.89 0.01 −0.26
(12.76) (7.15) (6.21) (12.64) (12.77) (7.53) (3.99) (4.77) (4.75) (3.92) (2.65) (0.72) (0.23)
log population 18.01 −0.39 −4.43 −10.84 −11.14 −6.84 −0.50 −2.25 −1.47 −0.37 0.43 0.41 0.40*
(12.94) (7.25) (6.29) (12.82) (12.95) (7.63) (4.05) (4.83) (4.82) (3.98) (2.69) (0.72) (0.22)
GDP per capita 0.00 0.00 −0.00 −0.00 −0.00 −0.00 0.00 0.00 0.00 0.00 0.00 0.00 −0.00
(0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00)
Percentage aged 65 older −1.98 −1.63** −1.44** −1.23 −1.15 −0.73 −0.31 −0.31 −0.39 −0.34 −0.10 −0.00 0.04
(1.27) (0.71) (0.62) (1.26) (1.27) (0.75) (0.40) (0.48) (0.47) (0.40) (0.27) (0.07) (0.02)
Life expectancy 1.58 1.81 1.55 2.36 2.44 1.23 −0.41 −0.64 −0.95 −1.09* −0.96** −0.34*** −0.09**
(1.96) (1.10) (0.95) (1.94) (1.96) (1.16) (0.61) (0.73) (0.73) (0.61) (0.41) (0.11) (0.04)
Health expenditure as % of −1.09 1.30 0.46 −0.31 0.27 1.19 1.38** 0.99 1.27* 1.61*** 1.07*** 0.30*** 0.07***
GDP (1.73) (0.97) (0.84) (1.72) (1.73) (1.02) (0.54) (0.65) (0.65) (0.53) (0.36) (0.09) (0.02)
Human development index 184.97* 4.58 −5.38 −61.54 −75.92 −85.58 −28.13 −14.48 −5.73 −3.08 7.84 7.95 4.32**
(109.96) (61.65) (53.50) (108.99) (110.11) (64.89) (34.39) (41.08) (40.98) (33.55) (22.67) (5.94) (1.61)
Diabetes prevalence 0.91 0.88 1.26* 2.01 2.14 1.05 0.63 0.62 0.41 0.58 0.25 0.01 0.05***
(1.29) (0.73) (0.63) (1.28) (1.30) (0.76) (0.40) (0.48) (0.48) (0.40) (0.27) (0.07) (0.02)
Cardiovasc death rate 0.04 0.03 0.02 −0.01 −0.00 −0.02 −0.03 −0.03 −0.05** −0.03* −0.02 −0.01** −0.00***
(0.06) (0.03) (0.03) (0.06) (0.06) (0.04) (0.02) (0.02) (0.02) (0.02) (0.01) (0.00) (0.00)
Constant −157.93 −272.41*** −278.97*** −379.85** −366.22** −209.82** −37.71 −30.86 6.50 14.06 32.98 14.02 3.25
(170.73) (95.72) (83.06) (169.22) (170.97) (100.76) (53.39) (63.78) (63.62) (53.02) (35.84) (9.82) (3.84)
Observations 60 60 60 60 60 60 60 60 60 59 59 55 33
R-squared 0.31 0.56 0.61 0.41 0.38 0.52 0.65 0.54 0.52 0.62 0.55 0.61 0.85
***p < 0.01, **p < 0.05, *p < 0.1. Standard errors in parentheses.
Table 8. Predictors of Days Used in Study 2.
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13)
Power distance 21.40 23.41 32.28* 14.96 9.29* 4.17*** 4.11 5.02 11.87** 8.78** 4.52 −108.82 −208.73
(12.78) (15.24) (17.32) (14.89) (5.20) (1.40) (3.66) (7.47) (5.34) (3.26) (67.45) (70.88) (203.62)
Institutional collectivism −0.67 4.34 −9.22 −29.14* −6.35 −0.48 −0.90 1.33 −0.30 1.41 161.56** −69.83 −190.65
(13.27) (15.83) (17.98) (15.46) (5.40) (1.45) (3.80) (7.75) (5.54) (3.21) (66.47) (66.89) (136.43)
Uncertainty avoidance −8.50 −8.45 −14.82 −3.11 −8.01* −0.31 0.81 3.58 1.57 −0.14 76.78 −35.14 126.15
(11.13) (13.28) (15.08) (12.97) (4.53) (1.22) (3.19) (6.50) (4.65) (2.74) (56.60) (64.32) (115.83)
Future orientation 5.74 −40.05** −45.08** −9.50 −6.26 −0.23 1.30 3.87 7.93 1.84 −6.88 66.31 164.24
(12.72) (15.18) (17.24) (14.82) (5.18) (1.39) (3.64) (7.44) (5.32) (3.43) (70.95) (72.88) (166.28)
Gender egalitarianism 10.32 −0.19 −2.10 7.12 5.47 1.22 −0.00 −6.74 −0.72 1.05 −30.08 −130.41** 164.60
(11.22) (13.38) (15.20) (13.07) (4.56) (1.23) (3.21) (6.56) (4.69) (2.72) (56.28) (55.70) (114.40)
Assertiveness −12.69 −16.97 −24.20 −7.90 3.62 1.09 1.77 0.96 −1.60 0.00 11.48 15.16 −272.48
(12.92) (15.41) (17.51) (15.05) (5.26) (1.41) (3.70) (7.55) (5.40) (3.23) (66.81) (66.17) (221.95)
Performance orientation 22.48 58.07*** 77.10*** 14.83 15.92** 2.41 0.25 −6.33 −0.35 3.98 39.29 −118.75 −127.29
(16.32) (19.47) (22.12) (19.01) (6.64) (1.79) (4.67) (9.54) (6.82) (4.75) (98.28) (99.07) (235.57)
Humane orientation 4.72 1.34 −0.61 28.31** 3.99 1.77 4.71 3.49 3.94 3.40 −73.82 60.39 −267.72*
(10.90) (13.01) (14.78) (12.70) (4.44) (1.19) (3.12) (6.37) (4.56) (2.65) (54.76) (54.71) (128.12)
In-group collectivism −3.71 −31.56** −53.71*** −15.78 −12.53*** −3.81*** −5.29* −7.22 −5.72 −6.29** 50.70 51.37 419.68*
(10.44) (12.46) (14.15) (12.17) (4.25) (1.14) (2.99) (6.10) (4.36) (3.08) (63.70) (67.95) (211.40)
log GDP 2.30 −12.07 −6.15 −13.21 −0.59 −0.80 −0.04 −2.19 −1.23 −0.03 117.55 120.74 −49.78
(14.26) (17.01) (19.32) (16.61) (5.80) (1.56) (4.08) (8.33) (5.96) (3.46) (71.48) (73.39) (175.35)
log population −11.21 1.72 −1.52 13.74 −1.49 −0.24 −0.84 0.48 −0.79 −1.52 −140.87* −152.20* 175.32
(14.88) (17.76) (20.17) (17.34) (6.06) (1.63) (4.26) (8.70) (6.22) (3.61) (74.58) (74.92) (172.27)
GDP per capita 0.00 −0.00 −0.00 −0.00 −0.00 −0.00* −0.00 −0.00 −0.00 −0.00** −0.00 −0.00 0.01
(0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.01)
Percentage aged 65 older 1.85 1.95 1.93 2.33 1.39** 0.23 0.00 −0.08 −0.02 0.07 1.05 0.40 −26.36
(1.32) (1.57) (1.78) (1.53) (0.54) (0.14) (0.38) (0.77) (0.55) (0.32) (6.65) (6.62) (15.45)
Life expectancy −1.65 −3.96** −4.02* −1.67 −1.21* −0.27 −0.01 0.15 −0.39 −0.50 11.86 5.74 −9.21
(1.63) (1.94) (2.20) (1.90) (0.66) (0.18) (0.47) (0.95) (0.68) (0.41) (8.51) (8.84) (20.21)
Health expenditure as % of GDP 1.25 −3.71 −8.10** −6.70** −2.61*** −0.80*** −0.88 −0.96 −0.97 −1.01 −4.22 −2.74 −5.40
(2.30) (2.75) (3.12) (2.69) (0.94) (0.25) (0.66) (1.35) (0.96) (0.60) (12.44) (12.42) (16.08)
Human development index −143.96 244.58 252.44 176.56 25.86 19.82 21.14 37.58 46.01 45.10 −1789.06** −789.08 2921.50
(146.94) (175.29) (199.15) (171.20) (59.79) (16.08) (42.06) (85.88) (61.39) (36.26) (750.21) (781.84) (2076.90)
Diabetes prevalence 0.78 3.62** 4.23** 1.13 1.27** 0.46*** 0.12 −0.41 0.73 1.05*** −9.56 −1.92 −38.48*
(1.41) (1.69) (1.92) (1.65) (0.58) (0.15) (0.40) (0.83) (0.59) (0.37) (7.58) (7.66) (20.23)
Cardiovasc death rate 0.06 −0.05 −0.07 −0.12* −0.01 −0.00 0.00 0.01 0.00 −0.01 0.07 0.55** −0.61
(0.05) (0.06) (0.07) (0.06) (0.02) (0.01) (0.01) (0.03) (0.02) (0.01) (0.25) (0.25) (0.49)
Constant 198.42 450.43* 518.75* 154.99 102.19 13.77 −13.99 34.04 −21.69 −19.90 −965.01 888.49 −1994.47
(194.79) (232.37) (264.00) (226.95) (79.26) (21.32) (55.76) (113.84) (81.38) (49.35) (1020.98) (1059.07) (2861.58)
Observations 55 55 55 55 55 55 55 55 55 54 54 53 26
R-squared 0.46 0.51 0.52 0.34 0.45 0.56 0.24 0.20 0.32 0.42 0.49 0.47 0.80
***p < 0.01, **p < 0.05, *p < 0.1. Standard errors in parentheses.
Table 9. Predictors of Average Daily Cases in Study 2.
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13)
Power distance −47.76** −16.94 −133.19 −352.15 −329.43 −141.07 −78.25 −254.06 −309.18 −210.40 196.71 −1373.67 −991.01
(19.34) (62.43) (120.77) (248.35) (275.64) (261.22) (235.45) (233.97) (224.49) (235.11) (736.81) (2283.31) (18,007.66)
Institutional collectivism 10.70 −24.64 132.39 79.86 16.92 −185.31 −317.05 −412.85* −431.39* −388.92 −941.65 368.32 −13,044.04
(20.08) (64.83) (125.39) (257.86) (286.20) (271.23) (244.46) (242.94) (233.08) (231.69) (726.09) (2154.83) (12,065.56)
Uncertainty avoidance 45.37** −12.94 209.72* 518.54** 403.99 210.85 35.35 168.47 197.70 249.01 −381.57 69.46 −9188.37
(16.85) (54.38) (105.19) (216.32) (240.09) (227.53) (205.08) (203.80) (195.54) (197.28) (618.25) (2071.99) (10,243.84)
Future orientation −29.16 46.52 −136.97 −380.39 −203.05 −116.85 69.01 −135.15 −248.04 −265.20 −737.55 −1705.95 13,252.76
(19.26) (62.17) (120.25) (247.29) (274.47) (260.11) (234.45) (232.98) (223.54) (247.31) (775.05) (2347.82) (14,705.99)
Gender egalitarianism −14.01 −73.75 −196.95* −347.92 −358.63 −203.02 −119.87 −82.45 −90.55 9.37 453.11 4163.55** −9951.09
(16.98) (54.81) (106.02) (218.02) (241.98) (229.32) (206.69) (205.40) (197.07) (196.16) (614.75) (1794.52) (10,117.13)
Assertiveness 5.07 0.17 95.61 68.25 46.41 −39.44 −77.02 −149.92 −134.38 63.58 690.43 −137.69 −32,750.84
(19.55) (63.12) (122.08) (251.06) (278.65) (264.07) (238.02) (236.53) (226.94) (232.87) (729.80) (2131.60) (19,629.43)
Performance orientation −30.41 2.75 −154.39 −182.38 −232.19 −53.13 34.60 91.10 59.28 −74.07 59.00 −9.14 −997.49
(24.70) (79.74) (154.24) (317.18) (352.04) (333.62) (300.70) (298.82) (286.71) (342.56) (1073.55) (3191.71) (20,833.33)
Humane orientation −17.50 −54.29 −145.54 −409.43* −315.09 −226.73 −111.92 −172.99 −218.28 −109.46 209.37 −584.85 −15,749.80
(16.51) (53.28) (103.07) (211.95) (235.25) (222.94) (200.94) (199.69) (191.59) (190.87) (598.18) (1762.43) (11,330.90)
In-group collectivism 34.17** 36.73 211.27** 427.20** 490.49** 320.71 324.79 444.74** 376.13** 407.74* 163.07 2233.25 1344.66
(15.81) (51.04) (98.72) (203.01) (225.32) (213.53) (192.46) (191.26) (183.50) (222.04) (695.87) (2189.18) (18,696.02)
log GDP 19.32 72.72 61.55 106.43 97.13 −5.04 33.22 158.65 47.92 −30.47 −699.19 −1779.36 −6310.62
(21.58) (69.68) (134.78) (277.15) (307.62) (291.52) (262.76) (261.11) (250.53) (249.16) (780.84) (2364.41) (15,507.59)
log population −6.96 −13.47 38.84 81.38 99.44 182.30 130.63 52.85 179.48 297.43 1334.45 4371.13* 15,335.32
(22.53) (72.74) (140.69) (289.32) (321.12) (304.32) (274.29) (272.58) (261.52) (259.97) (814.74) (2413.46) (15,235.38)
GDP per capita −0.00 0.00 −0.00 0.00 0.00 0.01 0.01 0.02** 0.02** 0.02*** 0.04* 0.09 0.89
(0.00) (0.00) (0.00) (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) (0.02) (0.06) (0.94)
Percentage aged 65 older −2.06 −8.60 −10.87 −1.17 −14.39 −20.61 −16.49 3.61 10.23 −1.64 −0.78 −266.73 −2076.68
(1.99) (6.43) (12.44) (25.58) (28.39) (26.90) (24.25) (24.10) (23.12) (23.17) (72.60) (213.24) (1366.34)
Life expectancy 3.18 7.60 13.26 17.79 26.64 29.80 10.93 −14.12 −18.31 −22.41 −145.91 61.79 −222.50
(2.46) (7.95) (15.38) (31.62) (35.09) (33.26) (29.98) (29.79) (28.58) (29.67) (92.98) (284.77) (1786.93)
Health expenditure as % of GDP 0.43 8.20 25.20 26.05 72.63 101.11** 106.59** 80.38* 81.11* 151.87*** 405.37*** 1268.19*** 1045.41
(3.49) (11.26) (21.78) (44.79) (49.72) (47.12) (42.47) (42.20) (40.49) (43.34) (135.84) (399.99) (1422.15)
Human development index −114.48 −45.80 146.87 −701.32 −610.14 −558.88 −27.45 −302.10 109.46 636.10 11,229.67 10,050.80 69,679.64
(222.43) (718.03) (1388.88) (2856.11) (3170.03) (3004.17) (2707.77) (2690.83) (2581.71) (2614.94) (8195.03) (25,187.32) (183,680.12)
Diabetes prevalence −2.52 −9.17 −13.90 −15.66 −15.67 −23.45 −16.12 −3.05 6.23 17.53 123.86 80.94 58.01
(2.14) (6.91) (13.37) (27.49) (30.52) (28.92) (26.07) (25.90) (24.85) (26.43) (82.84) (246.79) (1789.32)
Cardiovasc death rate 0.03 0.17 0.06 0.06 −0.01 −0.10 0.21 0.52 0.30 0.37 0.60 −2.47 5.30
(0.07) (0.24) (0.46) (0.96) (1.06) (1.01) (0.91) (0.90) (0.86) (0.87) (2.72) (8.08) (43.65)
Constant −289.95 −1639.90* −2877.17 −2777.32 −4170.33 −3308.73 −3417.26 −2437.29 −506.92 −3140.17 −4823.06 −55,697.32 142,263.39
(294.87) (951.86) (1841.17) (3786.22) (4202.37) (3982.51) (3589.58) (3567.12) (3422.47) (3558.73) (11,152.79) (34,118.49) (253,076.93)
Observations 55 55 55 55 55 55 55 55 55 54 54 53 26
R-squared 0.42 0.61 0.56 0.54 0.51 0.52 0.56 0.63 0.67 0.74 0.68 0.65 0.87
***p < 0.01, **p < 0.05, *p < 0.1. Standard errors in parentheses.
Table 10. Predictors of Average Daily Case Growth Rates in Study 2.
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13)
Power distance −21.72** −2.41 −5.91 −12.53 −13.23 −5.78 −0.65 −3.82 −3.88 −2.03 0.69 −0.30 0.16
(10.26) (5.85) (5.79) (10.25) (10.86) (7.09) (4.50) (3.67) (3.17) (2.90) (2.11) (0.62) (0.51)
Institutional collectivism −7.77 −4.49 5.16 1.83 1.52 −2.57 −5.81 −6.06 −5.54 −4.54 −2.55 0.14 −0.19
(10.66) (6.07) (6.01) (10.64) (11.27) (7.36) (4.67) (3.81) (3.30) (2.86) (2.08) (0.59) (0.34)
Uncertainty avoidance 1.80 −0.45 8.57* 20.14** 15.93 8.29 0.76 3.05 3.00 2.76 −1.06 −0.05 −0.24
(8.94) (5.09) (5.04) (8.93) (9.46) (6.18) (3.92) (3.20) (2.76) (2.44) (1.77) (0.57) (0.29)
Future orientation −0.02 3.75 −5.58 −16.35 −10.95 −6.75 0.77 −2.19 −3.67 −3.24 −2.19 −0.39 −0.07
(10.22) (5.82) (5.76) (10.21) (10.81) (7.06) (4.48) (3.65) (3.16) (3.05) (2.22) (0.64) (0.41)
Gender egalitarianism −17.73* −7.49 −10.27* −13.98 −13.18 −6.25 −1.80 −1.25 −0.92 0.23 1.35 1.16** −0.33
(9.01) (5.13) (5.08) (9.00) (9.53) (6.23) (3.95) (3.22) (2.79) (2.42) (1.76) (0.49) (0.28)
Assertiveness 2.73 −0.53 4.92 3.20 1.73 0.08 −0.91 −1.94 −1.08 1.27 2.13 −0.03 −0.58
(10.37) (5.91) (5.85) (10.36) (10.97) (7.17) (4.55) (3.71) (3.21) (2.87) (2.09) (0.58) (0.55)
Performance orientation −12.28 1.11 −6.10 −5.50 −6.97 −1.83 0.75 1.05 0.89 −0.78 0.25 0.26 0.44
(13.11) (7.47) (7.39) (13.09) (13.86) (9.06) (5.74) (4.69) (4.05) (4.23) (3.08) (0.87) (0.59)
Humane orientation −8.07 −4.77 −6.33 −16.02* −12.87 −7.58 −1.13 −2.14 −2.41 −0.95 0.75 −0.24 −0.08
(8.76) (4.99) (4.94) (8.75) (9.26) (6.05) (3.84) (3.13) (2.71) (2.36) (1.71) (0.48) (0.32)
In-group collectivism 1.72 3.85 8.88* 15.71* 18.09** 10.20* 5.40 7.14** 5.46** 4.55 0.41 0.63 −0.35
(8.39) (4.78) (4.73) (8.38) (8.87) (5.80) (3.68) (3.00) (2.59) (2.74) (1.99) (0.60) (0.53)
log GDP −11.57 7.05 3.33 3.28 4.71 1.41 0.27 2.75 1.14 −0.54 −1.85 −0.41 −0.15
(11.45) (6.53) (6.46) (11.44) (12.11) (7.91) (5.02) (4.10) (3.54) (3.08) (2.24) (0.64) (0.44)
log population 14.71 −2.14 1.14 3.39 1.43 3.11 2.59 0.52 2.08 3.76 3.67 1.05 0.28
(11.96) (6.81) (6.74) (11.94) (12.65) (8.26) (5.24) (4.28) (3.70) (3.21) (2.34) (0.66) (0.43)
GDP per capita 0.00 0.00 −0.00 0.00 0.00 0.00 0.00 0.00** 0.00** 0.00*** 0.00** 0.00 0.00
(0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00)
Percentage aged 65 older −0.92 −0.71 −0.56 0.10 −0.24 −0.40 −0.35 0.02 0.11 −0.01 −0.01 −0.07 −0.01
(1.06) (0.60) (0.60) (1.06) (1.12) (0.73) (0.46) (0.38) (0.33) (0.29) (0.21) (0.06) (0.04)
Life expectancy −0.31 0.61 0.55 0.56 0.91 0.84 0.25 −0.17 −0.24 −0.30 −0.41 0.02 −0.02
(1.31) (0.74) (0.74) (1.31) (1.38) (0.90) (0.57) (0.47) (0.40) (0.37) (0.27) (0.08) (0.05)
Health expenditure as % of GDP −1.42 0.77 1.14 0.82 1.86 2.20* 2.01** 1.29* 1.34** 1.88*** 1.21*** 0.34*** 0.00
(1.85) (1.05) (1.04) (1.85) (1.96) (1.28) (0.81) (0.66) (0.57) (0.54) (0.39) (0.11) (0.04)
Human development index 198.60 −3.79 15.24 −18.79 −38.11 −28.76 0.83 −6.62 −3.09 9.02 30.03 1.47 1.83
(118.03) (67.26) (66.58) (117.89) (124.84) (81.56) (51.71) (42.21) (36.50) (32.28) (23.49) (6.87) (5.17)
Diabetes prevalence −0.47 −0.80 −0.66 −0.53 −0.54 −0.58 −0.31 −0.07 0.12 0.26 0.36 0.02 0.04
(1.14) (0.65) (0.64) (1.13) (1.20) (0.79) (0.50) (0.41) (0.35) (0.33) (0.24) (0.07) (0.05)
Cardiovasc death rate −0.04 0.02 0.00 −0.00 0.00 −0.00 0.00 0.01 0.01 0.00 0.00 −0.00 −0.00
(0.04) (0.02) (0.02) (0.04) (0.04) (0.03) (0.02) (0.01) (0.01) (0.01) (0.01) (0.00) (0.00)
Constant 234.80 −131.78 −130.60 −85.46 −119.09 −90.15 −69.67 −49.53 −25.13 −41.34 −17.31 −15.30 4.63
(156.46) (89.17) (88.26) (156.28) (165.50) (108.12) (68.55) (55.95) (48.39) (43.94) (31.97) (9.30) (7.12)
Observations 55 55 55 55 55 55 55 55 55 54 54 53 26
R-squared 0.34 0.57 0.55 0.50 0.43 0.49 0.53 0.62 0.68 0.74 0.68 0.63 0.82
***p < 0.01, **p < 0.05, *p < 0.1. Standard errors in parentheses.
Table 11. Predictors of Days Used in Study 3.
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13)
Cultural tightness −1.07 0.40 −3.25 −0.53 0.64 2.82* 5.67* 0.95 4.20** 2.63*** −0.59 26.92** −18.76
(1.84) (4.95) (5.94) (1.10) (1.52) (1.37) (3.03) (1.04) (1.55) (0.87) (13.68) (9.79) (28.09)
log GDP 10.89 6.07 −9.49 −0.94 7.27 13.36 26.93 5.01 13.30 10.16 235.45* 207.96** −126.56
(15.24) (40.95) (49.08) (9.13) (12.54) (11.34) (25.08) (8.62) (12.84) (7.17) (112.59) (94.80) (337.01)
log population −9.49 −25.01 −5.45 −2.22 −11.02 −19.69* −38.32 −4.82 −15.55 −11.72 −179.84 −257.94*** 176.71
(14.63) (39.33) (47.14) (8.76) (12.04) (10.89) (24.09) (8.27) (12.33) (6.88) (108.03) (86.42) (315.69)
GDP per capita 0.00 −0.00 −0.00 −0.00 −0.00 −0.00* −0.00 −0.00 −0.00 −0.00* 0.00 −0.01*** 0.00
(0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.01)
Percentage aged 65 older 1.46 −2.70 −1.58 0.70 1.09 −1.61* −4.35** −1.37* −1.50 −1.01* −13.32 1.62 4.80
(1.23) (3.31) (3.97) (0.74) (1.01) (0.92) (2.03) (0.70) (1.04) (0.58) (9.15) (6.78) (17.91)
Life expectancy −3.66 2.76 3.95 −0.78 −2.80 1.50 4.63 1.54 0.23 0.44 64.18** 13.48 −11.34
(3.11) (8.37) (10.03) (1.86) (2.56) (2.32) (5.12) (1.76) (2.62) (1.47) (23.10) (18.08) (43.31)
Health expenditure as % of GDP −2.33 4.78 3.40 −0.70 −0.90 2.01 4.80 0.21 0.93 0.49 −32.50** 6.11 −21.83
(2.02) (5.42) (6.50) (1.21) (1.66) (1.50) (3.32) (1.14) (1.70) (0.96) (15.02) (11.66) (36.36)
Human development index −62.43 −55.11 −44.57 25.60 67.25 −42.04 −103.90 26.85 76.35 51.77 −1927.63 −1628.75* 1751.49
(152.08) (408.70) (489.84) (91.08) (125.10) (113.14) (250.33) (85.99) (128.15) (71.54) (1122.83) (823.83) (2170.38)
Diabetes prevalence 0.88 1.11 2.87 0.39 0.75 −0.47 −2.19 −1.03 0.01 0.40 −7.20 −3.15 −4.89
(1.57) (4.21) (5.05) (0.94) (1.29) (1.17) (2.58) (0.89) (1.32) (0.75) (11.71) (7.98) (16.90)
Cardiovasc death rate −0.06 0.05 −0.01 −0.01 0.00 0.08 0.18 0.06 0.10 0.07* 2.10*** 0.52 −0.35
(0.09) (0.24) (0.29) (0.05) (0.07) (0.07) (0.15) (0.05) (0.07) (0.04) (0.65) (0.58) (1.08)
Constant 238.18 137.19 119.83 115.81 159.54 −98.16 −321.87 −178.38 −185.14 −156.02 −6536.09*** −694.78 249.95
(245.78) (660.51) (791.63) (147.20) (202.18) (182.86) (404.56) (138.97) (207.10) (115.62) (1814.78) (1719.95) (3748.41)
Observations 31 31 31 31 31 31 31 31 31 30 30 28 19
R-squared 0.25 0.37 0.23 0.37 0.33 0.51 0.47 0.24 0.43 0.53 0.58 0.64 0.42
***p < 0.01, **p < 0.05, *p < 0.1. Standard errors in parentheses.
To rule out alternative explanations, we controlled for several important national socioeconomic, demographic, and health-related characteristics. GDP and GDP per capita are key economic indicators related to the number of COVID-19 cases (e.g., Pardhan & Drydakis, 2021). In this paper, GDP is the 2019 GDP measured in current U.S. dollars. GDP per capita is the 2019 GDP per capita. We also controlled for Human Development Index which is “a summary measure of average achievement in key dimensions of human development: a long and healthy life, being knowledgeable and have a decent standard of living,” measured by the “the geometric mean of normalized indices for each of three dimensions” (United Nations Development Programme, 2020) since it may influence the speed of the pandemic spread. We controlled for population which is the total number of people in the country since research shows that population is related to the number of COIVD-19 cases (e.g., Khan et al., 2021). We used log-transformed GDP and population in the regressions. We controlled for the percentage aged 65 older which refers to the percentage of people who are aged 65 years or older, since seniors are more suspectable to the virus and countries with higher proportion of seniors may experience faster spread of the pandemic. Lastly, we include key indicators of nations’ medical capability and relevant health-related variables which can influence how fast the virus spread. Life expectancy is the average age that people can expect to live, measured by the average age at which people die in a country. Health expenditure percentage measures the percentage of GDP a country spends on health. We controlled fro diabetes prevalence – the percentage of the population who have diabetes – since diabetes is shown to be positively related to COVID-19 infection (e.g., Singh et al., 2020). We also controlled for cardiovascular death rate which is the rate of death due to cardiovascular disease since it is positively related to the severity of COVID-19 (e.g., Kang et al., 2020).
Study 1
Method
Study 1 investigated how two of Hofstede’s national cultural dimensions – power distance and uncertainty avoidance – relate to the speed of COVID-19 spread. The scores on the six national cultural dimensions were obtained from Hofstede et al. (2010). All cultural dimension indices were measured on a 0-100 scale, with higher numbers representing great power distance, individualism, masculinity, uncertainty avoidance, long-term orientation, and indulgence. Each cultural dimension index was divided by 100 in this study.
Results
Tables 5–7 present the multivariate analyses results for our three dependent variables in Study 1: (1) number of days taken for the number of COVID-19 cases to grow from x cases to y cases and (2) average daily cases and (3) average daily case growth rate for that country during that period. In these analyses, we found limited support for Hypothesis 1 (model 1 in Table 6 and model 1 in Table 7), suggesting that power distance is negatively related to the speed of COVID-19 spread only at the very beginning of the COVID-19 pandemic (i.e., the period taken to reach 1000 cases in a particular country) and not in the later stages of the pandemic. The coefficients in other models were mostly in the hypothesized direction during the entry phase.
We also found support for Hypothesis 2, which posited that uncertainty avoidance is positively related to the speed of COVID-19 spread (models 1 and 2 in Table 5; models 2, 3, and 6 in Table 6; models 1, 3, and 7 in Table 7). While uncertainty avoidance did not have an across-the-board impact, it did play a noticeable role during the earlier phase of the COVID-19 lifecycle in a nation. Uncertainty avoidance was key to the rapid entry of COVID-19: It was negatively related to the average number of days from the first to the 2,000th case of COVID-19. It had a positive impact on average daily cases during the initial entry period and during the periods between the 1,001st and the 3,000th case and between the 5,001st and the 6,000th case. It also had a positive impact on the average daily case growth rates during the periods between (a) the first and the 1,000th case, (b) the 2,001st and the 3,000th case, and (c) the 6,001st and the 7,000th case. The coefficients in other models were largely in the predicted direction during the entry, take-off, and growth phases. Taken together, the results suggest that power distance only affects the speed of COVID-19 spread in the entry phrase, while uncertainty avoidance has more enduring effects which last through the entry, take-off, and growth phases of the pandemic.
Study 2
Method
In Study 2, we examined how four cultural practice dimensions studied in the GLOBE project – power distance, uncertainly avoidance, humane orientation, and in-group collectivism – relate to the speed of COVID-19 spread. The national culture data came from House et al. (2004). The GLOBE project expanded the five Hofstede dimensions into nine more refined categories. It retained Hofstede’s labels for power distance and uncertainty avoidance, changed Hofstede’s long-term orientation to future orientation, split Hofstede’s collectivism into institutional collectivism and in-group collectivism, divided Hofstede’s masculinity into gender egalitarianism and assertiveness, and added humane orientation and performance orientation as two distinct dimensions. The GLOBE practices measure a country’s culture “as it is” exhibited. Population weighted indices were used for Germany, South Africa, and Switzerland. In Study 2, the final sample consisted of 55 countries, of which 39 were the same as those used in Study 1 and 16 were different countries. The dependent variables and control variables were the same as in Study 1.
Results
Tables 8–10 present the multivariate regression results for Study 2. First, consistent with the findings of Study 1, we found support for Hypothesis 1, which posited that power distance is negatively related to the speed of COVID-19 spread (models 3, 5, 6, 9, and 10 in Table 8; model 1 in Table 9; model 1 in Table 10). Noticeably, power distance had positive impacts on the number of days used for five periods taken for the number of cases to grow from 2001 to 10,000 cases, considering the interval of every 1000 cases. As for the average daily cases and average daily case growth rates, power distance had a significant negative effect during the period of the first 1000 cases. The coefficients in other models were generally in the hypothesized direction during the entry, take-off, and growth phases.
Second, also consistent with the findings of Study 1, the results supported Hypothesis 2, which posited that uncertainty avoidance is significantly positively related to the speed of COVID-19 spread (model 5 in Table 8; models 1, 3, and 4 in Table 9; models 3 and 4 in Table 10). Uncertainty avoidance had a negative effect on the number of days taken for the number of cases to grow from 4001 to 5000 and a positive impact on average daily cases and average case growth rates during the early phase when COVID-19 cases were between 0 and 4000. The coefficients in other models were mostly in the predicted direction during the entry, take-off, and growth phases. Since Hofstede’s Uncertainty Avoidance Index is negatively correlated with GLOBE’s Uncertainty Avoidance practice scores (−0.61, p < 0.01: Hanges & Dickson, 2004, p. 140), these findings suggest that the results are robust regardless of the construct measurement variations.
Third, we found limited support for Hypothesis 3, which posited that humane orientation is negatively related to the speed of COVID-19 spread (model 4 in Table 8; model 4 in Table 9; model 4 in Table 10). Humane orientation was positively related to (a) the number of days taken for the number of cases to grow from 3001 to 4000 and (b) average daily cases and (c) average case growth rate during that period. The coefficients in the other models were largely in the expected direction during the entry, take-off, and growth phases.
Lastly, we found strong support for Hypothesis 4, which posited that in-group collectivism is positively related to the speed of COVID-19 spread (models 2, 3, 5, 6, 7, and 10 in Table 8; models 1, 3, 4, 5, 8, 9, and 10 in Table 9; models 3, 4, 5, 6, 8, and 9 in Table 8), suggesting that in-group collectivism has more lasting effects (i.e., up to the first 10,000 cases) than the other hypothesized cultural effects. In-group collectivism had a significant negative impact on the days taken for the number of COVID-19 cases to rise from 1001 to 10,000. For all three measures of the speed of COVID-19 spread, six out of the 10 coefficients, considering the interval of every 1000 cases, were significant. Most of the other coefficients of in-group collectivism were also in the expected direction.
Study 3
Method
Study 3 examined whether cultural tightness is negatively related to the speed of COVID 19 spread. We adopted the cultural tightness index from Gelfand et al. (2011) which systematically measures cultural tightness (as opposed to looseness). Sample items include “There are many social norms that people are supposed to abide by in this country”; “In this country, if someone acts in an inappropriate way, others will strongly disapprove”; and “People in this country almost always comply with social norms” (Gelfand et al., 2011). Our final sample consisted of 31 countries. A population weighted tightness index was used for Germany. The dependent variables and control variables were the same as in Study 1.
Results
Tables 11–13 present the multivariate regression results for Study 3. We found strong support for Hypothesis 5, which posited that cultural tightness is negatively related to the speed of COVID-19 spread (models 6, 7, 9, 10, and 12 in Table 11; models 8 and 12 in Table 12; models 8 and 12 in Table 13). Cultural tightness had a significant positive impact on the days taken for the number of COVID-19 cases to grow from 5001 to 10,000 and from 100,001 to 1,000,000. It had a significantly negative effect on average daily cases and average daily case growth rate during the take-off phase from 7001 to 8000 reported cases and during the maturity phase from 100,001 to 1,000,000 reported cases. Overall, Hypothesis 5 was supported during the take-off phase and strongly supported during the maturity phase.Table 12. Predictors of Average Daily Cases in Study 3.
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13)
Cultural tightness 1.73 0.39 −3.24 −28.93 −50.97 −61.37 −69.82 −107.83* −108.10 −96.93 −156.33 −1074.92** 1136.43
(4.97) (12.80) (29.48) (67.68) (70.31) (58.03) (48.53) (59.61) (65.69) (66.29) (204.93) (489.60) (2328.36)
log GDP 71.01* 3.29 138.71 244.09 107.03 −264.52 −300.64 −116.88 −204.86 −172.32 −1195.52 −7978.22 −22,135.15
(41.14) (105.86) (243.76) (559.68) (581.37) (479.87) (401.33) (492.89) (543.17) (545.42) (1686.15) (4742.54) (27,932.49)
log population −52.08 72.90 21.74 44.69 197.89 486.61 497.60 374.44 459.41 428.48 1537.75 9577.54** 30,845.05
(39.50) (101.66) (234.10) (537.49) (558.31) (460.84) (385.42) (473.35) (521.63) (523.33) (1617.84) (4323.45) (26,165.38)
GDP per capita −0.00 −0.00 −0.00 −0.00 0.01 0.01 0.02 0.03* 0.03* 0.03 0.04 0.25* −0.18
(0.00) (0.00) (0.01) (0.02) (0.02) (0.02) (0.01) (0.02) (0.02) (0.02) (0.05) (0.13) (0.93)
Percentage aged 65 older −4.02 −17.33* −20.75 −14.19 −13.10 4.64 24.11 34.79 46.36 35.61 61.23 −25.26 −1966.17
(3.32) (8.56) (19.70) (45.24) (46.99) (38.79) (32.44) (39.84) (43.90) (44.34) (137.07) (339.15) (1484.43)
Life expectancy 16.51* 37.46* 91.71* 191.82 165.50 74.75 −2.58 23.29 −2.27 −49.25 −336.89 −1137.32 820.03
(8.40) (21.63) (49.80) (114.34) (118.77) (98.03) (81.99) (100.69) (110.96) (111.91) (345.97) (904.54) (3589.90)
Health expenditure as % of −5.39 10.08 −10.39 −36.90 −13.27 49.27 54.05 27.21 30.22 104.78 488.03** 1147.80* 3803.55
GDP (5.45) (14.02) (32.27) (74.10) (76.97) (63.53) (53.13) (65.26) (71.91) (72.78) (224.99) (583.19) (3013.66)
Human development index −361.93 231.51 −587.08 −4120.02 −3277.40 −2185.47 −1193.29 −3578.35 −2968.88 −1821.76 5482.69 46,324.07 228,317.33
(410.52) (1056.46) (2432.75) (5585.58) (5801.99) (4789.00) (4005.27) (4919.02) (5420.75) (5439.25) (16,815.15) (41,213.66) (179,889.38)
Diabetes prevalence −0.42 −7.02 3.18 28.80 45.79 21.22 42.65 79.87 65.88 84.56 219.88 493.59 −539.37
(4.23) (10.90) (25.09) (57.60) (59.83) (49.39) (41.31) (50.73) (55.90) (56.71) (175.30) (399.14) (1400.85)
Cardiovasc death rate 0.58** 0.43 1.72 3.16 2.13 −1.40 −2.12 −0.91 −1.85 −2.02 −10.95 −34.32 −36.46
(0.24) (0.61) (1.42) (3.25) (3.38) (2.79) (2.33) (2.86) (3.15) (3.16) (9.78) (28.85) (89.35)
Constant −1947.34*** −4047.52** −10,298.31** −18,554.59* −16,196.16* −4941.04 586.88 −2644.27 −132.38 1762.99 25,064.79 99,998.12 −176,205.23
(663.45) (1707.37) (3931.61) (9026.97) (9376.71) (7739.60) (6472.99) (7949.72) (8760.57) (8791.23) (27,177.63) (86,043.98) (310,682.06)
Observations 31 31 31 31 31 31 31 31 31 30 30 28 19
R-squared 0.39 0.63 0.53 0.42 0.45 0.52 0.59 0.57 0.53 0.64 0.56 0.72 0.78
***p < 0.01, **p < 0.05, *p < 0.1. Standard errors in parentheses.
Table 13. Predictors of Average Daily Case Growth Rates in Study 3.
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13)
Cultural tightness −0.27 −0.32 −0.23 −1.40 −2.23 −1.77 −1.10 −1.60* −1.46 −1.11 −0.43 −0.32** 0.02
(2.49) (1.17) (1.39) (2.86) (2.87) (1.74) (0.89) (0.92) (0.92) (0.82) (0.58) (0.13) (0.06)
log GDP −36.14* 1.27 6.29 8.13 6.80 −1.81 −4.71 −0.11 −1.24 −2.16 −2.97 −1.99 0.11
(20.58) (9.70) (11.53) (23.64) (23.72) (14.39) (7.35) (7.64) (7.63) (6.77) (4.80) (1.29) (0.71)
log population 34.22* 5.20 1.03 2.32 3.24 8.05 8.12 4.29 4.97 5.18 3.96 2.46* 0.16
(19.76) (9.32) (11.07) (22.70) (22.78) (13.82) (7.06) (7.34) (7.33) (6.50) (4.60) (1.18) (0.66)
GDP per capita 0.00 −0.00 −0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00** −0.00
(0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00)
Percentage aged 65 older −0.69 −1.46* −1.09 −0.26 −0.41 −0.07 0.36 0.45 0.53 0.40 0.15 −0.03 −0.04
(1.66) (0.78) (0.93) (1.91) (1.92) (1.16) (0.59) (0.62) (0.62) (0.55) (0.39) (0.09) (0.04)
Life expectancy −1.00 3.86* 4.31* 7.40 7.57 3.89 −0.05 0.60 0.09 −0.69 −0.95 −0.24 0.04
(4.20) (1.98) (2.36) (4.83) (4.85) (2.94) (1.50) (1.56) (1.56) (1.39) (0.98) (0.25) (0.09)
Health expenditure as % of GDP 0.47 0.89 −0.31 −1.43 −1.05 0.65 1.11 0.34 0.60 1.40 1.46** 0.28* 0.03
(2.72) (1.28) (1.53) (3.13) (3.14) (1.91) (0.97) (1.01) (1.01) (0.90) (0.64) (0.16) (0.08)
Human development index 260.52 −7.63 −20.36 −167.14 −188.35 −115.76 −18.82 −61.36 −48.04 −17.86 14.07 9.09 2.24
(205.37) (96.81) (115.08) (235.92) (236.70) (143.63) (73.33) (76.24) (76.19) (67.56) (47.84) (11.21) (4.55)
Diabetes prevalence −0.79 −0.23 0.13 1.26 1.87 0.83 0.67 1.26 1.00 1.02 0.63 0.13 0.00
(2.12) (1.00) (1.19) (2.43) (2.44) (1.48) (0.76) (0.79) (0.79) (0.70) (0.50) (0.11) (0.04)
Cardiovasc death rate −0.12 0.05 0.08 0.11 0.11 0.01 −0.04 −0.00 −0.01 −0.03 −0.03 −0.01 0.00
(0.12) (0.06) (0.07) (0.14) (0.14) (0.08) (0.04) (0.04) (0.04) (0.04) (0.03) (0.01) (0.00)
Constant 308.36 −387.90** −482.04** −688.03* −666.55* −290.88 3.79 −75.30 −28.13 25.91 66.64 21.64 −9.93
(331.90) (156.45) (185.98) (381.28) (382.53) (232.12) (118.51) (123.22) (123.13) (109.19) (77.33) (23.40) (7.87)
Observations 31 31 31 31 31 31 31 31 31 30 30 28 19
R-squared 0.17 0.61 0.52 0.37 0.36 0.44 0.58 0.56 0.55 0.63 0.58 0.71 0.73
***p < 0.01, **p < 0.05, *p < 0.1. Standard errors in parentheses.
Discussion
General Discussion
Consistent with Hypothesis 1, we found that during both the entry and take-off phases, power distance practices slowed down the entry of COVID-19 up to the first 10,000 cases. Under novel conditions, people wait and watch for more information, paralyzed by the conceived uncertainty about the need for divergent behaviors. The personal authority of leaders helps to fill the consciousness void. People are open to following guidance from leaders when there is a risk to their life and health. Nevertheless, we found that nations where the societal culture supported the exercise of power by leaders and deference to the power of leaders were more effective in slowing the spread of COVID-19. Nations where the societal culture promoted a suspicious view of leaders witnessed a rapid spread of COVID-19, possibly because the leaders in these countries were more reluctant to exercise their power and the followers were also reluctant to adhere to the guidance issued by their leaders.
Supporting Hypothesis 2, we found that during the entry phase (i.e., up to the first 4000 cases), uncertainty avoidance practices were a catalyst of the higher caseload overall and the average cases per day. Uncertainty avoidance also led to the rapid entry and spread of COVID-19 up to the first 2000 cases. The nations where the societal culture supported the mitigation of uncertainty using appropriate techniques and technologies witnessed a rapid early spread of COVID-19. In such nations, people seek social validation of the data, information, and knowledge from trusted formal channels. Few trustworthy channels were available during the early phase of the spread of COVID-19. Formal networks tend to be liberal and to promote life-as-usual without any safeguards until sufficient data are available that signal the need for a divergent behavior. In contrast, the nations where the societal culture embraced uncertainty witnessed a slower spread of COVID-19 in the early phase. In such nations, people tend to rely on informal networks for making sense of novel conditions and managing the conceived uncertainty about the need for divergent behaviors. Informal networks “quickly gather high-resolution information and data. Neighborhood needs are rapidly assessed, support and failure points are known, and local knowledge is quickly disseminated” (Brugh et al., 2019).
Consistent with Hypothesis 4, we found that during the entry and take-off phases (i.e., up to the first 10,000 cases), in-group collectivism practices were a catalyst of the rapid entry and the average daily caseload. In the nations with societal cultures grounded in cohesive family and group ties, people were more vulnerable to a rapid spread of COVID-19 during both the entry and take-off phases. In contrast, in the nations with societal cultures grounded in weak family and group ties, people were less susceptible to a rapid spread of COVID-19 and to becoming super-spreaders during the entry and take-off phases.
Some recent studies have reported a positive impact of individualism on COVID-19 spread (Lu et al., 2021; Maaravi et al., 2021). Individualism refers to the extent to which tertiary societal institutions are weak, and people focus on their individual freedoms that may engender collective well-being (House et al., 2004) like in Eastern Europe (Gupta et al., 2004). In regions like Southern Asia, where tertiary institutions such as the non-government organizations and the media are strong, the people’s choices and freedom tend to be integrated with the spirit for collective good (Gelfand et al., 2004). Therefore, as a follow-up, we investigated the effect of individualism using Hofstede’s Individualism index and GLOBE’s institutional collectivism practices scale in Tables 5–10 respectively. Institutional collectivism practices mitigated the average daily case load during the take-off phase (effects are negative in all five models of take-off phase and significant in two). Additionally, Individualism catalyzed the average daily case load during the proliferation phase of more than a million cases.
Supporting Hypothesis 5, we found that during the take-off phase (i.e., the period from the 4,001st to the 10,000th case), cultural tightness dampened COVID-19 penetration in a nation. During the maturity phase (i.e., the period from the 100,001st case to the 1,000,000th case), cultural tightness slowed the spread of COVID-19 and mitigated the average daily caseload. In nations with tight cultures, people formalize institutional codes over time and enforce those codes at all levels. Thus, during the maturity phase, cultural tightness significantly attenuated the spread of COVID-19. Figure 1 summarizes the effects over the spread phase across the five hypotheses. Table 14 presents the summary of the significant effects in different phases of COVID spread across the three dependent variables and three studies.Figure 1. Effects over spread phase.
Table 14. Summary of the Significant Effects in Different Phases of COVID Spread Across the Three Dependent Variables and Three Studies.
Variables P1 P2 P3 P4 P5 P6 P7 P8 P9 P10 P11 P12 P13 Predicted effect P13 effect reversed?
Scale 1 1.25 1.5 1.75 2 2.17 2.33 2.50 2.67 2.83 3.00 4.00 5.00
Power distance −4 −1 −1 −1 −1 −1 3 1 <0 Y
Individualism 1 1 1 1 1 >0 Y
Uncertainty avoidance 1 2 4 2 −1 1 1 −2 >0 Y
Future orientation −1 −2 −1 -
Assertiveness -
Gender egalitarianism −1 −1 1 1 3 1 -
Performance orientation 1 −1 −1 -
Humane orientation −3 1 <0 Y
Family orientation 1 1 1 2 3 2 1 2 2 1 −1 >0 Y
Indulgence −2 −2 −1 −2 −2 <0 n.s
Tightness −1 −1 −2 −1 −1 −3 <0 n.s
In summary, national culture played an important role in terms of galvanizing leaders into taking decisive actions and encouraging people to adopt “COVID-apt” behaviors during the entry and take-off phases that constituted the first wave of COVID-19. In large power distance cultures, leaders are likely to offer their guidance and followers tend follow that guidance to mitigate the effects of the novel conditions. In strong uncertainty avoidance cultures, leaders tend to offer their guidance until sufficient data were available and followers may not follow the guidance until a sufficient number of people had experienced the effects of the novel conditions themselves. The cultural factor also influenced the institutional entrepreneurship of people in becoming the infecting super-spreaders or the infected victims during the first wave of COVID-19. Family and in-group collectivism encouraged people to continue enjoying their freedom to interact with different groups. In-group interactions limited the power of people to self-manage the risks of contagion during the entry and take-off phases. Overall, in absence of leader-centric risk-mitigation and COVID appropriate behavioral clarification and enforcement, group orientation and cultural looseness accelerated the COVID-19 spread during the early stages.
Cultural factors did not play a prominent role during the growth phase. By this time, sufficient data were available for people to make conscious decisions and to follow COVID-apt behaviors, even in low power distance nations. Further, even in weak uncertainty avoidance nations, adequate knowledge existed in the informal networks to lead people to exercise caution, whether they trusted the data or not. Additionally, sufficient time was available for social networks to become digital as people worried about the life and health of their loved ones even in in-group collectivism cultures. Cultural factors again became salient during the maturity phase, where some nations experienced their second wave of COVID-19. Cultural tightness played a key role in societies formalizing the norms of COVID-apt behaviors and in universalizing them religiously through everyone’s efforts.
This study has important practical implications. Following earlier research suggesting that there are significant cultural differences among various racial and ethnic groups within a nation (e.g., Coon & Kemmelmeier, 2001), this study helps to explain why people of different racial and ethnic groups are affected differently by the COVID-19 pandemic in multicultural societies such as the United States. For instance, APM Research Lab (2021) showed that the COVID-19 death rate for White Americans is 40.4 deaths per 100,000; the corresponding rates for African Americans, Hispanic Americans, and Asian Americans are 88.4, 54.4, and 36.4, respectively in the beginning of the pandemic. Culture may explain the racial differences beyond economic and political reasons, since people from different cultures may act differently in terms of social distancing, avoiding social gatherings, and wearing masks. For example, high in-group collectivism might explain the faster spread of the coronavirus among Hispanics, while high power distance may explain the slower spread of COVID-19 among Asian Americans. Policy makers should consider the cultural differences among countries and among diverse racial and ethnical groups in multicultural societies when designing and implementing COVID-19 contagion management policies and practices.
Limitations and Future Research
We recognize the limitations of this research. We adopted a phased approach, dividing the data into 13 periods, making the results suspectable to Type 1 error. To control for Type I error (see Edwards, 2001), we divided the nominal p value of .05 by the number of models examined (i.e., 13), yielding a critical p value of .003,846. Results remain robust at this critical p value. Further, while we show that cultural dimensions measured earlier (Gelfand et al., 2011; Hofstede, 2010; House et al., 2004) predict pandemic management during the 2019–2021 period, there is a further need to examine the causal mechanisms through which cultural dimensions generate these effects using multisource, multi-wave research designs. There is a need to measure and evaluate the paths through which the cultural effects materialize, as without knowledge of these effects, decision makers are likely to persist with culturally correlated biases and may not be conscious of the cultural factors that are shaping and binding their rationality. Moreover, as many studies have noted the dynamic nature of national cultures (e.g., Beugelsdijk & Welzel, 2018; Inglehart, 1990, 1997; Inglehart & Welzel, 2005), we recommend that future studies reexamine the effects of national culture on the speed of COVID-19 spread with updated cultural indexes. Research over the past 30 years has highlighted how power is becoming more decentralized (e.g., Triesman, 2007); people are relying more on data, information, and knowledge; and the world is becoming increasingly socially networked for managing the flow of knowledge and decentralized powers (e.g., Kushlev et al., 2017). Our findings suggest that these global cultural factors may have been a key catalyst for the rapid entry and take-off of COVID-19 across the world. Decentralized power, data orientation, and social networks make people more sensitive to emerging situations. The negative factors may get amplified before sufficient data are available about their negativity. Thus, the benefits of data-based knowledge may be attenuated. Further, when sufficient data are available, not all societies may be prepared to translate them into norms to be universally followed. For instance, for some groups, the imperative to safeguard their livelihoods in the face of weak social security may override the norms of COVID-apt behaviors. Additionally, even when societies translate the knowledge into normative codes, people may become lax after a while. Therefore, further research is needed to understand these behavioral tendencies and develop appropriate educational methods for promoting healthy behaviors.
There is also a need to further understand the cultural dimensions of the proliferation phase. Table 5.6 summarizes the significant effects across the three dependent variables and the three studies. This table suggests that national culture influenced the speed of the spread of COVID-19 in expected direction during the early stages; the direction reversed during the proliferation phase (above 1 million cases) for power distance, uncertainty avoidance, humane orientation, in-group collectivism, as well as individualism. In case of tightness, the effects were not significant during the proliferation phase. Additionally, although not hypothesized, future orientation had a negative impact on the speed of spread during the entry and growth phases, possibly due to proactive planning. Gender egalitarianism had a negative impact during the entry phase, but positive during the takeoff, maturity, and proliferation phases possibly because of the stronger risk of contagion when both men and women are equally active in diverse spheres. Indulgence had a negative impact during the takeoff, growth, and maturity phases. Indulgence is related to enjoying life and being happy Hofstede et al., (2010). People in the Indulgence cultures like Latin America indulge themselves with the emergent social situations, enjoying that as a virtual medium of entertainment for watching how others fall victim to the unexpected situations and being happy that they have gained the knowledge to save themselves from the same, adverse fate. There is a need for studies that consider these additional dimensions as well.
Conclusions
Why does COVID-19 spread faster in certain nations than in others? Although the political and economic differences among countries and the differential government interventions can significantly influence how fast the coronavirus spreads, we show that national cultural dimensions – power distance, uncertainty avoidance, humane orientation, in-group collectivism, and tightness – are significant predictors of the speed of COVID-19 spread at the beginning of the pandemic and should not be overlooked. In this research, we systematically examined the effects of national culture across 78 countries in three studies. The results collectively indicate that national culture significantly influences the speed of COVID-19 spread at the beginning of the pandemic but not in the later stages. We find that power distance, humane orientation, and cultural tightness are negatively related to the speed of COVID-19 spread, while uncertainty avoidance and in-group collectivism are positively related to speed of the disease spread at the beginning of the pandemic. We also show that compared with other cultural dimensions, cultural tightness has more lasting effects since it is significantly related to the speed of coronavirus spread up to the first 1,000,000 cases in a country.
Author Biographies
Dr. Xiaoyu Huang is an Associate Professor of Management at the Jack H. Brown College of Business and Public Administration at California State University, San Bernardino. Her research interests include strategic human resource management, international human resource management, leadership, and cross-cultural management. Her research has been published in Human Resource Management, Human Resource Management Journal, International Journal of Human Resource Management, Applied Psychology, and Journal of Organizational Change Management, etc.
Dr. Vipin Gupta (Ph.D., Wharton School; www.vipingupta.net) is a Professor of Management, and Co-director of the Center for Global Management, at the Jack H. Brown College of Business and Public Administration of California State University San Bernardino. Professor Gupta has published thirty influential books, including the co-edited GLOBE (Global Leadership and Organizational Behavior Effectiveness Program) book “Culture, Leadership, and Organizations – The GLOBE Study of 62 societies” (Sage Publications, 2004), eleven books on regional models of family business under CASE (Culturally-sensitive Assessment Systems and Education) Project, and twelve self-authored books in 2021-2022 under the project Vastly Integrated Processes Inside Nature,” on the metaphysics of everything, everybody, and everyone.
Dr. Cailing Feng is a professor and doctoral supervisor in the College of Public Administration in Nanjing Agricultural University. Her research interests focus on human resource management, organizational behavior, and leadership.
Dr. Fu Yang is a Professor at the School of Business Administration, Southwestern University of Finance and Economics. His research interests focus on leadership, career development, proactive behavior, and teams. His work has been published in several journals such as Human Resource Management, Journal of Business Ethics, European Journal of Work and Organizational Psychology, Applied Psychology: An International Review, Human Resource Management Journal, and International Journal of Human Resource Management.
Dr. Lihua Zhang is a Professor of Human Resource Management at the School of Labor and Human Resources at Renmin University of China. Her research interests are in the areas of transformational leadership, cross-culture management, organizational change, and human resource management in China. Her research has been published in the field's top journals such as Organization Science and Organizational Dynamics.
Jiaming Zheng is a PhD student in Human Resource Management and Organizational Behavior at the School of Labor and Human Resources at Renmin University of China. Her research interests focus on psychological contract, human resource management, and cross-cultural management.
Dr. Montgomery Van Wart is a professor of public administration at CSUSB, as well as a university administrator. He is the author of over 100 publications and has books on leadership, human resource management, ethics, and government-business relations, among others.
ORCID iDs
Xiaoyu Huang https://orcid.org/0000-0003-0487-7814
Fu Yang https://orcid.org/0000-0003-4385-2011
Montgomery Van Wart https://orcid.org/0000-0001-9243-4479
Appendix Table A1. Variable Definitions.
Variables Source Variable definitions
Dependent variables
Number of days used ECDC* The number of days taken for the number of COVID-19 cases to grow from x cases to y cases
Average daily cases ECDC* The average number cases per day from x COVID-19 cases to y cases
Average daily case growth rates ECDC* The average case growth rate (i.e., the number of new COVID-19 cases divided by the number of days used) from x COVID-19 cases to y cases times 100
Independent variables
Hofstede cultural dimensions Hofstede et al., 2010 Each cultural dimension index divided by 100
Globe cultural dimensions House et al., 2004 GLOBE cultural dimension index. Population weighted indices are used for Germany, South Africa, and Switzerland
Cultural tightness Gelfand et al. (2011) Tightness scores are used
Control variables
Gross domestic product (GDP) World bank The 2019 GDP measured in current U.S. dollars
GDP per capita ECDC* 2019 GDP per person
Population The total number of people in the country in 2019
Percentage aged 65 or older ECDC* The percentage of people who are 65 years old or older in 2019
Life expectancy ECDC* The average age that people can expect to live, measured by the average age people die in the country in 2019
Health expenditure (% of GDP) World bank The percentage of GDP a country spent on health in 2019
Human development index ECDC* This index is “a summary measure of average achievement in key dimensions of human development: a Long and healthy life, being knowledgeable and have a decent standard of living”, measured by the “the geometric mean of normalized indices for each of three dimensions” (united nations development programme) in 2019
Diabetes prevalence ECDC* The percentage of the population who had diabetes in 2019
Cardiovascular death rate ECDC* The rate of death due to cardiovascular disease in 2019
*Our World COVID-19 dataset, provided by the European Centre for Disease Prevention and Control (ECDC).
Table A2. List of Included Countries.
Studies Countries No. of countries
Study 1 Argentina, Australia, Austria, Bangladesh, Belgium, Brazil, Bulgaria, Canada, Chile, China, Colombia, Croatia, Czech, Denmark, El Salvador, Estonia, Finland, France, Germany, Greece, Hungary, India, Indonesia, Iran, Ireland, Italy, Japan, Latvia, Lithuania, Luxembourg, Malaysia, Malta, Mexico, Morocco, Netherlands, New Zealand, Norway, Pakistan, Peru, Philippines, Poland, Portugal, Romania, Russia, Serbia, Singapore, Slovakia, Slovenia, South Korea, Spain, Sweden, Switzerland, Thailand, trinidad, Turkey, United Kingdom, United States, Uruguay, Venezuela, vietnam 60
Study 2 Albania, Argentina, Australia, Austria, Bolivia, Brazil, Canada, China, Colombia, Costa Rica, Denmark, Ecuador, Egypt, El Salvador, Finland, France, Georgia, Germany*, Greece, Guatemala, Hungary, India, Indonesia, Ireland, Israel, Italy, Japan, Kazakhstan, Kuwait, Malaysia, Mexico, Morocco, Namibia, Netherlands, New Zealand, Nigeria, Philippines, Poland, Portugal, Qatar, Russia, Singapore, Slovenia, South Africa*, South Korea, Spain, Sweden, Switzerland*, Thailand, Turkey, United Kingdom, United States, Venezuela, Zambia, Zimbabwe 55
Study 3 Australia, Austria, Belgium, Brazil, China, Estonia, France, Germany*, Greece, Hungary, Iceland, India, Israel, Italy, Japan, Malaysia, Mexico, Netherlands, New Zealand, Norway, Pakistan, Poland, Portugal, Singapore, South Korea, Spain, Turkey, Ukraine, United Kingdom, United States, Venezuela 31
Total number of countries 78
*Population weighted measures are used.
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding: The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Social Science Foundation of China (No.22BGL140).
==== Refs
References
APM Research Lab. (2021). The color of coronavirus: COVID-19 deaths by race and ethnicity in the United States. https://www.apmresearchlab.org/covid/deaths-by-race
Beugelsdijk S. Welzel C. (2018). Dimensions and dynamics of national culture: Synthesizing hofstede with inglehart. Journal of Cross-Cultural Psychology, 49 (10 ), 1469–1505. 10.1177/002202211879850 30369633
Brewer P. Venaik S. (2011). Individualism-collectivism in Hofstede and GLOBE. Journal of International Business Studies, 42 (3 ), 436–445. 10.1057/jibs.2010.62
Brugh W. Sorokin G. Scott G. R. (2019). Combining formal and informal structures in crisis response. In Frontiers of engineering: Reports on leading-edge engineering from the 2018 symposium. National Academies Press.
Carl D. Gupta V. Javidan M. (2004). Power distance. In House R. J. Hanges P. J. Javidan M. Dorfman P. W. Gupta V. (Eds.). Culture, leadership, and organizations: The globe study of 62 societies (pp. 513–563). Sage
Carpenter S. (2000). Effects of cultural tightness and collectivism on self-concept and causal attributions. Cross-Cultural Research, 34 (1 ), 38–56. 10.1177/106939710003400103
Coon H. M. Kemmelmeier M. (2001). Cultural orientations in the United States. Journal of Cross-Cultural Psychology, 32 (3 ), 348–364. 10.1177/0022022101032003006
Cortez M. F. Mokhtar F. Wei L. D. (2021). Singapore confronts the division and fear that come from living with covid. Bloomberg. https://www.bloomberg.com/news/articles/2021-10-14/singapore-confronts-division-and-fear-bred-by-living-with-covid
Dizikes P. (2020). When culture clashes with Covid-19. MIT News. https://news.mit.edu/2020/when-culture-clashes-covid-19-0625
Dong M. Spadaro G. Yuan S. Song Y. Ye Z. Ren X. (2021). Self-interest bias in the covid-19 pandemic: A cross-cultural comparison between the United States and China. Journal of Cross-Cultural Psychology, 52 (7 ), 663–679. 10.1177/00220221211025739
Duncan C. (2020). Coronavirus: Italy charges nearly 110,000 people for breaking lockdown rules as tougher sanctions introduced. Independent. https://www.independent.co.uk/news/world/europe/coronavirus-italy-lockdown-police-charges-fines-quarantine-lombardy-a9427046.html
Edwards J. R. (2001) Multidimensional constructs in organizational behavior research: An integrative analytical framework. Organizational Research Methods, 4 (2 ), 144–192. 10.1177/109442810142004
Fabricius G. Buttgen M. (2013). The influence of knowledge on overconfidence: Consequences for management and project planning. International Journal of Business and Management, 8 (11 ), 1–12. 10.5539/ijbm.v8n11p1
Gelfand M. J. Bhawuk D. P. Nishii L. H. Bechtold D. J. (2004). Individualism and collectivism. In House R. J. Hanges P. J. Javidan M. Dorfman P. W. Gupta V. (Eds.). Culture, leadership, and organizations: The globe study of 62 societies (pp. 437–512). Sage.
Gelfand M. J. , Jackson J. C. Pan X. Nau D. Pieper D. Denison E. Dagher M. Van Lange P. Chiu C. Y. Wang M. (2021). The relationship between cultural tightness-looseness and COVID-19 cases and deaths: A global analysis. The Lancet Planetary Health, 5 (3 ), e135–e144. 10.1016/S2542-5196(20)30301-6 33524310
Gelfand M. J. Nishii L. H. Raver J. L. (2006). On the nature and importance of cultural tightness-looseness. Journal of Applied Psychology, 91 (6 ), 1225–1244. 10.1037/0021-9010.91.6.1225 17100480
Gelfand M. J. Raver J. L. Nishii L. Leslie L. M. Lun J. Lim B. C. Duan S. Almaliach A. Ang S. Arnadottir J. Aycan Z. Boehnke K. Boski P. Cabecinhas R. Chan D. Chhokar J. D D’Amato A. Ferrer M. Yamaguchi S. 2011). Differences between tight and loose cultures: A 33-nation study. Science, 332 (6033 ), 1100–1104. 10.1126/science.1197754 21617077
Gokmen Y. Baskici C. Ercil Y. (2021). The impact of national culture on the increase of COVID-19: A cross-country analysis of European countries. International Journal of Intercultural Relations, 81 , 1–8. 10.1016/j.ijintrel.2020.12.006 33518841
Guan Y. Deng H. Zhou X. (2020). Understanding the impact of the COVID-19 pandemic on career development: Insights from cultural psychology. Journal of Vocational Behavior, 119 , 103438–103438. 10.1016/j.jvb.2020.103438 32382162
Gupta V. De Luque M. S. House R. (2004). Multisource construct validity of globe scales. In House R. J. Hanges P. J. Javidan M. Dorfman P. W. Gupta V. (Eds.). Culture, leadership, and organizations: The GLOBE study of 62 societies (pp. 152–177). Sage.
Hanges P. J. Dickson M. W. (2004). The development and validation of the GLOBE culture and leadership scales. In House R. J. Hanges P. J. Javidan M. Dorfman P. W. Gupta V. (Eds.). Culture, leadership, and organizations: The GLOBE study of 62 societies (pp. 122–151). Sage.
Hofstede G. (2001). Culture’s consequences: Comparing values, behaviors, institutions and organizations across nations. Sage publications.
Hofstede G. McCrae R. R. (2004). Personality and culture revisited: Linking traits and dimensions of culture. Cross-Cultural Research, 38 (1 ), 52–88. 10.1177/106939710325944
Hofstede G. Bond M. H. (1988). The Confucius connection: From cultural roots to economic growth. Organizational dynamics, 16 (4 ), 5–21.
Hofstede G. H. Hofstede G. J. Minkov M. (2010). Cultures and organizations: Software of the mind. McGraw-Hill.
House R. Javidan M. Hanges P. Dorfman P. (2002). Understanding cultures and implicit leadership theories across the globe: An introduction to project globe. Journal of World Business, 37 (1 ), 3–10. 10.1016/S1090-9516(01)00069-4
House R. J. Hanges P. J. Javidan M. Dorfman P. W. Gupta V. (2004). Culture, leadership, and organizations: The GLOBE study of 62 societies. Sage.
Huynh T. L. D. (2020). The COVID-19 containment in Vietnam: What are we doing? Journal of Global Health, 10 (1 ), 010338. 10.7189/jogh.10.010338 32373318
Inglehart R. (1990). Culture shift in advanced industrial society. Princeton University Press.
Inglehart R. (1997). Modernization and postmodernization: Cultural, economic, and political change in 43 societies. Princeton University Press.
Inglehart R. Welzel C. (2005). Modernization, cultural change and democracy: The human development sequence. Cambridge University Press.
Kabasakal H. Bodur M. (2004). Humane orientation in societies, organizations, and leader attributes. In House R. J. Hanges P. J. Javidan M. Dorfman P. W. Gupta V. (Eds.). Culture, leadership, and organizations: The globe study of 62 societies (pp. 564–601). Sage.
Kang Y. Chen T. Mui D. Ferrari V. Jagasia D. Scherrer-Crosbie M. Chen Y. Han (2020). Cardiovascular manifestations and treatment considerations in COVID-19. Heart, 106 (15 ), 1132–1141. 10.1136/heartjnl-2020-317056 32354800
Khan W. Hussain A. Khan S. A. Al-Jumailey M. Nawaz R. Liatsis P. (2021). Analysing the impact of global demographic characteristics over the COVID-19 spread using class rule mining and pattern matching. Royal Society Open Science, 8 (1 ), 201823. 10.1098/rsos.201823 33614100
Kumar R. (2021). Impact of societal culture on covid-19 morbidity and mortality across countries. Journal of Cross-Cultural Psychology, 52 (7 ), 643–662. 10.1177/0022022121102510
Kushlev K. Proulx J. D. Dunn E. W. (2017). Digitally connected, socially disconnected: The effects of relying on technology rather than other people. Computers in. Human Behavior, 76 , 68–74.
Lu J. G. Jin P. English A. S. (2021). Collectivism predicts mask use during COVID-19. Proceedings of the National Academy of Sciences, 118 (23 ), e2021793118. 10.1073/pnas.2021793118
Maaravi Y. Levy A. Gur T. Confino D. Segal S. (2021). ‘‘The tragedy of the commons”: How individualism and collectivism affected the spread of the COVID-19 pandemic. Frontiers in Public Health, 9 , 37. 10.3389/fpubh.2021.627559
Mayer J. D. Schintler L. A. Bledsoe S. (2020). Culture, freedom, and the spread of covid‐19: Do some societies and political systems have national anti‐bodies? World Medical & Health Policy, 12 (4 ), 498–511. 10.1002/wmh3.377
McAuley C. (2020). Much of Europe is now on lockdown. But can authorities actually enforce those rules? Washington Post. https://www.washingtonpost.com/world/much-of-europe-is-now-on-lockdown-but-can-authorities-actually-enforce-thoserules/2020/03/21/c4f04bae-6abd-11ea-b199-3a9799c54512_story.html
McGreevy C. R. (2020). Irish people donate €2.5m to Native American tribe devastated by coronavirus. The Irish Times. https://www.irishtimes.com/news/ireland/irish-news/irish-people-donate-2-5m-to-native-american-tribe-devastated-by-coronavirus-1.4414963
Meeuwesen L. Van den Brink-Muinen A. Hofstede G. (2009). Can dimensions of national culture predict cross-national differences in medical communication? Patient Education and Counseling, 75 (1 ), 58–66. 10.1016/j.pec.2008.09.015 19013047
Minkov M. Kaasa A. (2021). A test of Hofstede’s model of culture following his own approach. Cross Cultural & Strategic Management, 28 (2 ), 384–406. 10.1108/CCSM-05-2020-0120
Morrow A. (2020). 60 percent of French people admit to violating lockdown rules. RFI. https://www.rfi.fr/en/france/20201112-60-percent-of-french-admit-to-violating-lockdown-rules-confinement-ifop-survey
Parboteeah K. P. Bronson J. W. Cullen J. B. (2005). Does national culture affect willingness to justify ethically suspect behaviors? International Journal of Cross Cultural Management, 5 (2 ), 123–138. 10.1177/1470595805054489
Pardhan S. Drydakis N. (2021). Associating the change in new COVID-19 cases to GDP per capita in 38 European countries in the first wave of the pandemic. Frontiers in Public Health, 8 , 582140. 10.3389/fpubh.2020.582140 33553086
Pattin K. A. White B. C. Barney N. Gui J Nelson H. H. Kelsey K. R. Andrew A. S. Karagas M. R. Moore J. H. (2009). A computationally efficient Hypothesis testing method for epistasis analysis using multifactor dimensionality reduction. Genetic Epidemiology, 33 (1 ): 87–94. 10.1002/gepi.20360 18671250
Pelto P. J. (1968). The differences between “tight” and “loose” societies. Society, 5 (5 ), 37–40. 10.1126/science.1197754
Romano A. Spadaro G. Balliet D. Joireman J. Van Lissa C. Jin S. Agostini N. P. Bélanger J. J. Gützkow B. Kreienkamp J. Leander (2021). Cooperation and trust across societies during the COVID-19 pandemic. Journal of Cross-Cultural Psychology, 52 (7 ), 622–642. 10.1177/00220221209889
Salvador C. E. Berg M. K. Yu Q. San Martin A. Kitayama S. (2020). Relational mobility predicts faster spread of COVID-19: A 39-country study. Psychological Science, 31 (10 ), 1236–1244. 10.1177/0956797620958118 32915703
Schlösser O. Frese M. Heintze A. M. Al-Najjar M. Arciszewski T. Besevegis E. Zhang K. (2013). Humane orientation as a new cultural dimension of the globe project: A validation study of the globe scale and out-group humane orientation in 25 countries. Journal of Cross-Cultural Psychology, 44 (4 ), 535–551. 10.1177/0022022112465671
Singh A. K. Gupta R. Ghosh A. Misra A. (2020). Diabetes in COVID-19: Prevalence, pathophysiology, prognosis and practical considerations. Diabetes & Metabolic Syndrome: Clinical Research & Reviews, 14 (4 ), 303–310. 10.1016/j.dsx.2020.04.004
Tisdell C. A. (2020). Economic, social and political issues raised by the COVID-19 pandemic. Economic Analysis and Policy, 68 , 17–28. 10.1016/j.eap.2020.08.002 32843816
Triandis H. C. (1989). The self and social behavior in differing cultural contexts. Psychological Review, 96 (3 ), 506–520. 10.1037/0033-295X.96.3.506
Triesman D. (2007). What have we learned about the causes of corruption from ten years of cross-national empirical research? Annual Review of Political Science, 10 (1 ), 211–244. 10.1146/annurev.polisci.10.081205.095418
United Nations Development Programme. (2020). Human development reports. http://hdr.undp.org/en/content/human-development-index-hdi
United States Centers for Disease Control and Prevention. (2022). How to protect yourself & others . https://www.cdc.gov/coronavirus/2019-ncov/prevent-getting-sick/prevention.html
Uz I. (2015). The index of cultural tightness and looseness among 68 countries. Journal of Cross-Cultural Psychology, 46 (3 ), 319–335. 10.1177/002202211456361
Venaik S. Brewer P. (2010). Avoiding uncertainty in hofstede and globe. Journal of International Business Studies, 41 (8 ), 1294–1315. 10.1057/jibs.2009.96
Wuthnow R. (1991). Acts of compassion: Caring for ourselves and helping others . Princeton University Press
Zhu N. Smetana J. G. Chang L. (2021). Acceptance of society-level and individual-level preventive measures during the COVID-19 pandemic among college students in three societies. Journal of Cross-Cultural Psychology, 52 (7 ), 606–621. 10.1177/0022022121995971
| 0 | PMC9703026 | NO-CC CODE | 2022-11-29 23:21:06 | no | Cross Cult Res. 2022 Nov 25;:10693971221141478 | utf-8 | Cross Cult Res | 2,022 | 10.1177/10693971221141478 | oa_other |
==== Front
Curr Cardiol Rep
Curr Cardiol Rep
Current Cardiology Reports
1523-3782
1534-3170
Springer US New York
36441403
1801
10.1007/s11886-022-01801-6
Pericardial Disease (AL Klein and CL Jellis, Section Editors)
COVID-19 Vaccine–Related Myocardial and Pericardial Inflammation
Furqan Muhammad 1
Chawla Sanchit 1
Majid Muhammad 2
Mazumdar Samia 2
Mahalwar Gauranga 2
Harmon Evan 2
http://orcid.org/0000-0001-9240-8369
Klein Allan [email protected]
2
1 grid.239578.2 0000 0001 0675 4725 Department of Internal Medicine, Cleveland Clinic Foundation, Fairview Hospital, Cleveland, OH USA
2 grid.239578.2 0000 0001 0675 4725 Center for the Diagnosis and Treatment of Pericardial Diseases, Section of Cardiovascular Imaging, Department of Cardiovascular Medicine, Heart, Vascular, and Thoracic Institute, Cleveland Clinic, J1-4, 9500 Euclid Avenue, Cleveland, OH USA
28 11 2022
2022
24 12 20312041
7 10 2022
© The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2022, Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Purpose of Review
To review myocarditis and pericarditis developing after COVID-19 vaccinations and identify the management strategies.
Recent Findings
COVID-19 mRNA vaccines are safe and effective. Systemic side effects of the vaccines are usually mild and transient. The incidence of acute myocarditis/pericarditis following COVID-19 vaccination is extremely low and ranges 2–20 per 100,000. The absolute number of myocarditis events is 1–10 per million after COVID-19 vaccination as compared to 40 per million after a COVID-19 infection. Higher rates are reported for pericarditis and myocarditis in COVID-19 infection as compared to COVID-19 vaccines.
Summary
COVID-19 vaccine–related inflammatory heart conditions are transient and self-limiting in most cases. Patients present with chest pain, shortness of breath, and fever. Most patients have elevated cardiac enzymes and diffuse ST-segment elevation on electrocardiogram. Presence of myocardial edema on T2 mapping and evidence of late gadolinium enhancement on cardiac magnetic resonance imaging are also helpful additional findings. Patients were treated with non-steroidal anti-inflammatory drugs and colchicine with corticosteroids reserved for refractory cases. At least 3–6 months of exercise abstinence is recommended in athletes diagnosed with vaccine-related myocarditis. COVID-19 vaccination is recommended in all age groups for the overall benefits of preventing hospitalizations and severe COVID-19 infection sequela.
Keywords
COVID-19 vaccination
Pericarditis
Myocarditis
SARS-CoV-2
Myopericarditis
issue-copyright-statement© Springer Science+Business Media, LLC, part of Springer Nature 2022
==== Body
pmcIntroduction
The incidence of myocarditis and pericarditis in the general population ranges from 2 to 20 individuals per 100,000 per year and is more common among males [1–3]. A higher incidence of myocarditis, up to 36.5 per 100,000, is also reported in some population-based studies [4]. The most common etiology of these two conditions in the developed world is viral or idiopathic, and adenoviruses, enteroviruses, parvovirus B19, herpesviridae, Influenza, HIV, hepatitis C, and HIV are commonly identified viruses [1]. Coronaviruses are also found to be associated with inflammatory diseases of the heart. Recently, SARS-CoV-2 was identified to cause myocarditis and pericarditis. Additionally, cases of COVID19 vaccine-associated myocarditis (COVID19VAM) and COVID19 vaccine-related pericarditis (COVID19VAP) are also reported. Interestingly, inflammatory heart conditions have also been previously reported as an adverse reaction to influenza and hepatitis B vaccinations [5]. This review focuses on the inflammatory heart diseased related to COVID19 vaccinations with a brief summary of available cases.
COVID-19 Vaccine–Related Inflammatory Heart Diseases
COVID-19 mRNA vaccines are safe and effective as several large-scale clinical trials confirmed that these vaccines improve patient outcomes and disease severity [6, 7]. Systemic side effects of the vaccines are usually mild, transient, and more commonly reported in younger populations [8]. Although autoimmune reactions and inflammatory heart diseases are rare, there is growing literature describing these infrequent side effects arising from COVID-19 vaccines, particularly mRNA vaccines.
The adverse events regarding vaccinations in the USA are documented in the Vaccine Adverse Event Reporting System (VAERS), a passive surveillance system for early reporting of potential adverse effects [9]. Oster et al. analyzed VAERS with the data of approximately 350 million mRNA-based COVID-19 vaccine administrations. They identified 1991 cases, of which 1626 met the case definition for myocarditis of which only 391 had pericarditis. Pericarditis without myocardial involvement was reported in 684 patients. The rates were highest after the second dose of the COVID-19 vaccine and predominantly young males between the age of 12 and 24 years were affected. The reporting rates were lower in females than males across all age strata aged younger than 50 years [10]. The percentage of males and the age group of the recipients presenting with myopericarditis were similar to pre-COVID 2019 era documentation in the VAERS associated with other vaccinations. However, VAERS is susceptible to reporting bias, affected by various factors such as general awareness of the adverse effects [11]. CDC also reported myocarditis associated with COVID 2019 mRNA vaccines from Pfizer-BioNTech and Moderna. The data from CDC showed an increased risk of myopericarditis among recipients between 12 and 39 years of age within 7 days of receiving the COVID-19 vaccine compared to unvaccinated individuals or those who received non-mRNA COVID-19 vaccines on the same days. The findings were consistent with the CDC dataset and several other epidemiological studies [12–15]. Witberg et al. found the rate of vaccine-related myocarditis almost five times more prevalent in 16–29-year-old population (10.69 per 100,000) compared to the general population (2.13 per 100,000) [12]. However, the rate of COVID-19 vaccine myopericarditis in the male population between 12 and 39 years was reported to be 1.8 per 100,000 by Husby et al. and up to 19% in adolescents aged between 12 and 17 years by Chua et al. [13, 15]. A study conducted in the UK included around 10 million vaccinated individuals and showed that myocarditis was more common in males less than 40 years of age who received two doses of the Moderna vaccine (113 cases per million) compared to 3 doses of the Pfizer vaccine (28 cases per million) [14].
Proposed Pathophysiologic Mechanisms of COVID-19 Vaccine–Related Pericardial and Myocardial Inflammation
The Pfizer-BioNTech and Moderna vaccines use modified mRNA packaged in a lipid nanoparticle and injected intramuscularly into the human body. Upon attaching to the host cell, the nanoparticle inserts mRNA into the cytoplasm that travels to the ribosomes to synthesize viral spike proteins (translation) [16]. These newly synthesized proteins are degraded by proteasomes into antigenic peptides. These antigenic peptides are expressed on the cellular membrane through the major histocompatibility complex (MHC) class I to interact with CD8+ cytotoxic T cells. The translated proteins also gain entry into the antigen-presenting cells (APCs) and are expressed on the cellular surface by MHC class II. T-cell receptor (TCR) membrane protein and CD4 proteins of CD4+ T cells interact with the MHC class II to produce cytokines such as IL-2, IL-4, and IL5 and activate cellular immunity. This interaction also activates humoral immune response by triggering the differentiation of the B cells that in turn release a significant amount of antibodies against the viral spike proteins [17]. Additionally, innate immune response is also activated when RNA in mRNA vaccines binds to Toll-like receptor (TLR) and produces type-I interferon.
The AstraZeneca and Johnson & Johnson vaccines have a similar mechanism of action. These vaccines use a modified chimpanzee DNA adenovirus. This does not create an immune response to adenovirus but only to the viral protein encoded in the host DNA. The DNA vector encodes a protein similar to viral s-protein and migrates to cell nucleus and utilizes host enzymes to convert to mRNA. Host cell ribosomes interact with this mRNA resulting into translated proteins that are expressed on cell membranes by MHC. The mechanism of DNA vaccines are similar to RNA vaccines from this point. The interaction of T cells with MHC as described above leads to the activation of B cells, T cells, and plasma cells to form antibodies against the viral proteins [16].
Various mechanisms are proposed for vaccine-induced myocarditis and pericarditis, including hyper-activation of the immune system, molecular mimicry, and differences of sex hormones. Figure 1 describes the possible pathophysiological mechanisms including hyper-inflammation, molecular mimicry, and hormonal differences.Fig. 1 Mechanism of action of COVID vaccines and development of myopericarditis
Hyper-inflammation
SARS-CoV-2 mRNA vaccines contain nucleoside-modified mRNA that encodes the viral spike glycoprotein of the virus. The vaccine does not contain live viruses or DNA [18]. Nucleoside modifications of mRNA have been shown to reduce innate immunogenicity. Usually, the tendency of dendritic cells or Toll-like receptor (TLR) expressing cells to express and activate cytokines markers is markedly less when exposed to mRNA with nucleoside modifications versus when treated with unmodified RNA. However, in selected individuals with genetic tendencies, the immune response to mRNA may still accelerate the activation of both innate and acquired immune responses [19].
The dendritic cells or TLR expressing cells exposed to ribonucleic acid (RNA) can excessively express and activate cytokines markers in some individuals with genetic inclination. The immune system may therefore detect the mRNA in the vaccine as an antigen, resulting in hyper-activation of inflammatory cascades leading to the development of myocardial and pericardial inflammation [19].
Bystander Activation
Virus-specific CD8 + T cells migrate to the target tissue infected with the viral infection and activate perforin and granzyme-mediated cytotoxicity. Reactive oxygen species (ROS) and nitric oxide that are released by adjacent macrophages also incur damage to the surrounding tissues. Additionally, pro-inflammatory cytokines that are released by CD4 + cells also enhance the phagocytic function of macrophages [20]. This bystander activation, dysregulated lymphocyte proliferation, and ineffective clearance of killed cells expose autoantigens and contribute to the generation of autoreactive cells [21]. In severe COVID-19 cases, lymphocytes are accumulated at the site of viral infection and activate bystander killing of adjacent non-infected cells by releasing pro-inflammatory cytokines and ROS. Autopsy reports of COVID-19 patients showed lymphocytic infiltration in the lungs, heart, kidney, and liver suggesting bystander activation [22]. Similar mechanism may be implicated in COVID-19 vaccines. It has been shown that CD8 + T cells response is mostly apparent only after the second dose of vaccination in addition to the increased quantity of CD4 + cells in infection-naïve individuals [23]. This corresponds to a higher number of inflammatory heart diseases primarily identified after the second dose in the epidemiological studies, but data exploring pathophysiological basis of these rare COVID-19 vaccine side effects is lacking.
Molecular Mimicry
Molecular mimicry between the spike protein of SARS-CoV-2 and self-antigens is another potential mechanism that may result in myocarditis following vaccination [6]. Antibodies against spike glycoproteins have been shown to cross-react with similar human peptide proteins including α-myosin. The COVID-19 vaccine does not appear to commence de novo immune-mediated adverse events. Instead, certain individuals with genetic susceptibility may have dysregulated immune pathways at baseline which may be aggravated following vaccination against COVID-19. This may result in polyclonal B cell expansion, immune complex formation, and further inflammation [19].
Autoantibodies and Anti-idiotype Antibodies
According to the Network Hypothesis, antibody response against an antigen induces downstream antibody response against the antigen-specific antibody. In other words, initial antibody induced by the antigen (Ab1) has immunogenic regions called idiotopes that may trigger antibodies against Ab1 antibody. These antibodies, known as Ab2 antibodies, may structurally mimic Ab1 antibodies or the original viral antigen. Neutralization of Ab1 antibodies by forming immune complexes, up/downregulate the ACE2 receptors directly, and mimicking the viral particles are various mechanisms by which these autoantibodies may affect the affected or normal cells [24]. This type of anti-idiotype response is previously implicated in autoimmunity arising after viral infection and preclinical studies of Ab2-induced autoimmune myocarditis [25]. However, the exact role of anti-idiotype antibody response is not investigated in the context of COVID-19 vaccines and needs further exploration.
Sex Hormones
Male prevalence in myopericarditis cases has been illustrated before, and the reasons remain unknown. Majority of the reported cases of COVID19 vaccine related inflammatory heart conditions were male (Table 1), and the underlying sex hormonal differences may be responsible for this gender distribution. Testosterone has been shown to play a role by commitment to a Th1-type immune response [26]. Estrogen’s inhibitory effects on pro-inflammatory T cells result in diminished cell-mediated immune responses, and the incidence of pericarditis is higher during the postmenopausal period in women [27]. Stimulating effect of testosterone and inhibitory effect of estrogen on the inflammatory cells in combination may explain an increased risk of developing myo-pericardial inflammation secondary to COVID19 vaccination.
Clinical Manifestations and Diagnostic Findings
The clinical manifestations and course of myocarditis and pericarditis are variable. Myocarditis may range from subclinical disease to fatigue, chest pain, heart failure, ventricular arrhythmias, cardiogenic shock, and rarely death [28–30]. Acute myocarditis is defined as the development of symptoms of heart failure (dyspnea, orthopnea, and lower extremity edema) over 3 months or less, while chronic myocarditis is when these symptoms persist over 3 months [31]. The clinical diagnosis of acute pericarditis requires two of the following criteria: (1) sharp, pleuritic positional chest pain; (2) pericardial friction rub; (3) new diffuse ST elevation or PR depression; and (4) pericardial effusion [32].
Table 1 summarizes 88 patients published in the literature as of April 2022. Out of these cases, 67 patients ≤ 18 years, and 81 (92%) were males. All the cases of vaccine-related pericarditis and myocarditis were after mRNA vaccination. Only 17 (19%) cases happened after receiving the first dose, and the rest occurred after the second dose of the vaccine. The most common symptoms included chest pain, shortness of breath, and fever. The onset of symptoms varied extensively among all cases, ranging from hours to as long as 39 days after vaccine administration. In the published cases, all but one patient did not have evidence of elevated cardiac enzymes. The most common electrocardiogram (ECG) finding was diffuse ST-segment elevation in 53 patients (60%), followed by diffuse non-specific ST-segment changes in five patients (6%), and other changes included T-wave inversions, ST-segment depression, junctional rhythm, and left bundle branch block (Table 2). A total of 17 patients (22%) had normal ECGs. Most patients underwent further evaluation with dedicated transthoracic echocardiography (TTE) and cardiac magnetic resonance imaging (CMR). TTE findings were normal in two-thirds of the patients. Among the rest of the cases, there was heterogeneity in the echocardiographic findings. Pericardial effusions were found in only six patients (7%). There were eight (9%) patients noted to have a reduced left ventricle (LV) function, of which seven were ≤ 18 years. CMR was available in 63 patients (72%), and the presence of myocardial edema on T2 mapping was seen in 55 patients. Evidence of late gadolinium enhancement (LGE) was only present in three patients. Additionally, five cases showed evidence of myocardial fibrosis. However, there were five (6%) patients who had no evidence of acute myocarditis and or pericarditis upon imaging (Table 2).Table 1 Studies reporting myocardial or pericardial inflammation in individuals receiving COVID-19 vaccine
Study
(Last name of first author followed by year published) No. of patients Age (avg) Gender (x male, y female) Symptoms Vaccine Dose Days after vaccine (avg)
Ambati 2021 [33] 2 17 2 M Chest pain, fatigue, fever Pfizer-BioNTech 2nd 2
Fleming-Nouri 2021 [34] 7 21 7 M Chest pain Pfizer-BioNTech 2nd 3
Sakaguchi 2021 [35] 1 49 1 M Fever, cough, orthopnea Pfizer-BioNTech 2nd 2
Chen 2021 [36] 1 16 1 M fever, chest pain, myalgia Pfizer-BioNTech 1st 4
Das 2021 [37] 25 15 24 M, 1 F Chest pain, dyspnea, fatigue, chills Pfizer-BioNTech 2nd 3
Nygaard 2022 [38] 15 17 13 M, 2 F Chest pain Pfizer-BioNTech 2nd 2
Bartlett 2021 [39] 1 46 1 M Fever, chest pain Pfizer-BioNTech 2nd 1
Badshah 2021 [40] 1 22 1 F Chest pain, chills Moderna 2nd 1
Hung 2022 [41] 1 23 1 M Fever, sore throat, myalgia AstraZeneca/Oxford 1st 7
McLean 2021 [42] 1 16 1 M Chest pain Pfizer-BioNTech 2nd 3
Schauer 2021 [43] 13 15 13 M Chest pain, fever, headache, myalgia Pfizer-BioNTech 2nd 3
Farooq 2022 [44] 1 63 1 M Dyspnea, chest tightness AstraZeneca/Oxford 2nd 35
Umei 2021 [45] 1 20 1 M Fever, myalgia Moderna 2nd 2
Kim 2021 [46] 1 29 1 M Fever, chest pain Pfizer-BioNTech 2nd 4
Ameratunga 2022 [47] 1 57 1 F Lethargy, dyspnea Pfizer-BioNTech 1st 3
Ashaari 2021 [48] 1 66 1 M Chest pain, fatigue Pfizer-BioNTech 1st 7
D’Angelo 2021 [49] 1 30 1 M Chest pain, shortness of breath Pfizer-BioNTech 2nd 2
Marshall 2021 [50] 7 17 7 M Chest pain Pfizer-BioNTech 2nd 3
Facetti 2021 [51] 1 20 1 M Chest pain Pfizer-BioNTech 2nd 3
King 2021 [52] 4 25 4 M Chest pain, pleuritic Moderna 2nd 4
Hasnie 2021 [53] 1 22 1 M Chest pain Moderna 1st 3
Patrignani 2021 [54] 1 56 1 M Abdominal and chest pain Pfizer-BioNTech 1st 4
Table 2 Laboratory and multimodality imaging findings in COVID-19 vaccine–related myopericarditis
Study
(Last name of first author followed by year published) EKG changes Elevation of cardiac enzymes Echocardiography findings MRI findings Management Outcomes
Ambati 2021 [33] Diffuse STE Y Normal N/A NSAIDs Resolved
Fleming-Nouri 2021 [34] STE Y mildly decreased left ventricular systolic function Generalized edema, hyperemia, fibrosis NSAIDs, IVIG Resolved
Sakaguchi 2021 [35] Normal Y Normal Diffuse myocardial edema Aspirin, diuretics Resolved
Chen 2021 [36] Poor R wave progression Y N/A Myocardial fibrosis and edema NSAIDs Resolved
Das 2021 [37] STE Y Mild LV systolic dysfunction, EF 48% Myocardial edema NSAIDS Resolved
Nygaard 2022 [38] STE Y Normal Myocarditis Ketorolac, colchicine, NSAIDs Resolved
Bartlett 2021 [39] STE Diffuse Y Normal, some reduced LVEF None Colchicine, NSAIDs Resolved
Badshah 2021 [40] Normal Y LVEF 45% Edema, pericarditis Steroid, colchicine, aspirin Resolved
Hung 2022 [41] STE Y Normal Myopericarditis IVIG, NSAIDs Resolved
McLean 2021 [42] STE Y Pericardial effusion Myocardial fibrosis, hyperemia, small pericardial effusion NSAIDs Resolved
Schauer 2021 [43] STE Y Normal Edema Supportive only Resolved
Farooq 2022 [44] Normal Y N/A Mild fibrosis of the basal septum and inferior and lateral walls. No myocardial inflammation or infarction IV and oral steroids Resolved
Umei 2021 [45] T-wave inversion Y Regional hypokinesis Myopericarditis None Resolved
Kim 2021 [46] STE Y Normal, small pericardial effusion Myopericarditis NSAIDs Resolved
Ameratunga 2022 [47] Normal Y N/A N/A None Died
Ashaari 2021 [48] None Y N/A No findings NSAIDs Resolved
D’Angelo 2021 [49] STE Y Pericardial effusion Myopericarditis NSAIDs, IVIG, steroids Resolved
Marshall, 2021 [50] STE V2-V4 Y Mild pericardial effusion, wall motion abnormality Myocardial necrosis, diffuse fibrosis NSAIDs Resolved
Facetti, 2021 [51] STE Y N/A Myopericarditis NSAIDs Resolved
King 2021 [52] TWI lateral waves, STE Y Normal N/A None Resolved
Hasnie 2021 [53] Diffuse STE Y Normal Myopericarditis Aspirin, colchicine Resolved
Patrignani 2021 [54] Normal Y Hypokinetic mid to apical anterior segments Myocarditis None Resolved
Management of Vaccine-Induced Pericarditis and Myocarditis
A higher degree of suspicion is warranted to clinically identify these patients as they may present without typical cardiac symptoms and signs of chest pain or dyspnea. The diagnosis of myocarditis should be suspected in patients with elevated cardiac biomarkers (e.g., troponin, BNP, or NT-proBNP) and new ST-segment changes suggestive of myocardial injury on the ECG or new onset impaired systolic left ventricle systolic function in the absence of underlying ischemic changes [55]. Pericardial effusion may be an initial clue on the point of care ultrasound examination and should raise suspicion for pericarditis in an appropriate clinical setting. Patients with pericardial and myocardial inflammation have overlapping features and may present as a spectrum of myopericarditis or perimyocarditis [56].
COVID-19 vaccine–related myopericarditis is transient and self-limiting in most cases. Upon review of the 88 patients, the onset of symptoms after exposure was shorter for COVID19VAM compared to a typical viral illness. These cases were typically diagnosed within a few days of receiving the vaccination, and in one particular case, the symptoms occurred within a few hours of vaccine administration. Majority of the cases were treated with non-steroidal anti-inflammatory drugs (NSAIDs) alone. Some patients did require dual therapy with either an NSAID with colchicine or an NSAID with glucocorticoids. In a minority of patients (16%), intravenous immunoglobulin (IVIG) was also utilized in addition to NSAIDs and steroids, particularly in younger patients. In cases where follow-up was available, a complete clinical resolution was documented in 87 (99%) patients during the follow-up. There was only one recorded death among all reviewed cases.
Overall, patients presenting with the symptoms concerning myopericarditis require cautious triaging whether they are seen in the ambulatory or emergency department setting. The American Heart Association recommends that patients seek immediate medical advice if they experience sudden, sharp, stabbing chest pain, shortness of breath, or loss of consciousness [57]. The usual workup involves a careful history and physical exam, ECG, cardiac enzymes, and PCR for common viral etiologies when indicated. Even though elevated troponin may indicate acute cardiac injury, it is not specific to myocarditis. This is important as it may mimic other cardiac disorders and therefore requires a high level of clinical suspicion. The utility of CMR in this patient population remains unclear, but it may offer a further cardiac evaluation to confirm the diagnosis or better characterize progression or resolution [58]. There are no guidelines available for managing COVID19VAM or COVID19VAP. However, the treatment usually involves NSAIDs, colchicine, and guideline-directed therapy for heart failure with a reduced ejection fraction. Steroids and IVIG have immunomodulatory properties, theoretically reducing the specific immune response triggered by the administration of the COVID-19 vaccine. However, steroids and IVIG use should be limited to refractory cases as only limited data is available.
Majority of patients with myocardial and pericardial inflammation after the COVID-19 vaccine show complete resolution. A three-month follow-up of five patients with myocarditis after COVID19 vaccination showed normalization of ejection fraction and resolution of myocardial edema in all patients [58]. However, at this point, there remains a scarcity of long-term data for COVID-19 vaccine–related inflammatory heart conditions.
Activity Restriction and Return to Play for Athletes
COVID-19 Infection–Related Myocarditis
The prevalence of COVID-19 infection–related myocarditis is up to 3% in young athletes (Big Ten COVID-19 Registry). There was no adverse cardiac event in short-term surveillance of COVID-19-positive athletes with myocarditis [59]. American College of Cardiology Guideline recommends at least 3–6 months of exercise abstinence in athletes diagnosed with myocarditis after a COVID-19 infection. An expert evaluation and testing is indicated in these athletes prior to resuming the training and exercise. In selected cases of rapid resolution of symptoms, earlier re-evaluation may also be sought prior to 3 months (not sooner than 1 month) [60•]. There is a need for more studies in this arena to understand the disease progression and resolution of myocarditis with long-term effects.
COVID-19 Vaccine–Related Myocarditis
The burden of COVID-19 vaccine–related cardiac inflammatory conditions is not specifically reported for athletes. The data from the US Military Health System showed only 23 cases (median age 25 years) after 2.8 million doses of mRNA vaccinations [61]. A case of myocarditis in an athlete who presented with squeezing chest pain 10 days after the second dose of mRNA vaccine was identified. He was managed with ibuprofen and beta blocker and athletic activity was restricted for 3–6 months [62]. Screening for myocarditis is not indicated in asymptomatic athletes who received COVID-19 vaccination. Activity restriction along with repeat testing in select cases is recommended in athletes developing this extremely rare vaccine-related myocarditis [60•].
COVID-19 Infection and COVID-19 Vaccine–Related Pericarditis
There is paucity of data for physical activity in patients developing pericardial inflammation secondary to COVID-19 infection or COVID-19 vaccine. In general, patients with pericarditis are recommended to limit the physical activity as activity has anecdotally been related to increased severity and recurrence of pericarditis [63, 64].
Recommendations for COVID-19 Vaccination
The absolute number of myocarditis events is 1–10 per million vaccinated persons following the COVID-19 vaccine. However, the risk of myocarditis is significantly high after a COVID-19 infection (40 per million) [65]. Additionally, COVID-19 infection–related myocarditis is found in around 2 to 4 patients hospitalized for COVID-19 per 1000 hospitalizations as compared to 10 cases of vaccine-related myocarditis per 100,000 vaccinations [66]. Similarly, higher rates are reported for pericarditis and myopericarditis in COVID-19 infection as compared to COVID-19 vaccines [65, 67••]. There was no increase in the recurrence of pericarditis after COVID19 vaccination in patients on rilonacept therapy [68].
The most common age group affected by COVID19VAM and COVID19VAP is adolescent boys (between 12 and 17 years). A study estimated substantially low incidence rates in milder complications and improved outcomes for mortality and hospitalizations within this age group [69]. The current data suggests that the risk of inflammatory cardiac conditions, although rare overall, is significantly higher in COVID-19 infection as compared to COVID-19 vaccines. The benefit of vaccination clearly outweighs the risk of these rare side effects. Therefore, COVID-19 vaccination is still recommended in all age groups for the benefits of preventing hospitalizations and severe COVID-19 infection sequela.
Conclusions and Future Directions
Pericardial and myocardial inflammation is a rare side effect of the COVID-19 vaccine that resolves with anti-inflammatory medications. Mechanisms of vaccine-induced inflammatory heart conditions and the long-term effect of these inflammatory conditions need exploration. Moreover, further studies are required to comparatively assess the COVID-19 infection and COVID-19 vaccine–related inflammatory heart conditions.
Declarations
Conflict of Interest
Dr. Klein has received a research grant from and is a scientific advisory board member for Kiniksa Pharmaceuticals; and is a scientific advisory board member for Swedish Orphan Biovitrum, Sweden, Pfizer, USA, and Cardiol Therapeutics; and he has a leadership or fiduciary role on the National Board of Echocardiography. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
Human and Animal Rights and Informed Consent
This article does not contain any studies with human or animal subjects performed by any of the authors.
This article is part of the Topical Collection on Pericardial Disease
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
Papers of particular interest, published recently, have been highlighted as: • Of importance •• Of major importance
1. Tschöpe C, Ammirati E, Bozkurt B, Caforio ALP, Cooper LT, Felix SB, et al. Myocarditis and inflammatory cardiomyopathy: current evidence and future directions. Nat Rev Cardiol. 2020 183 [Internet]. Nature Publishing Group; 2020 [cited 2022 Aug 29];18:169–93. Available from: https://www.nature.com/articles/s41569-020-00435-x.
2. Fairweather DL, Cooper LT, Blauwet LA. Sex and gender differences in myocarditis and dilated cardiomyopathy. Curr Probl Cardiol [Internet]. Curr Probl Cardiol. 2013 [cited 2022 Aug 29];38:7–46. Available from: https://pubmed.ncbi.nlm.nih.gov/23158412/.
3. Prepoudis A, Koechlin L, Nestelberger T, Boeddinghaus J, Lopez-Ayala P, Wussler D, et al. Incidence, clinical presentation, management, and outcome of acute pericarditis and myopericarditis. Eur Hear Journal Acute Cardiovasc Care [Internet]. Oxford Academic; 2022 [cited 2022 Aug 29];11:137–47. Available from: https://academic.oup.com/ehjacc/article/11/2/137/6445206.
4. Lota AS, Halliday B, Tayal U, Salmi S, Shakur R, Hammersley D, et al. Epidemiological trends and outcomes of acute myocarditis in the National Health Service of England | Circulation. AHA [Internet]. 2019 [cited 2022 Sep 2];140. Available from: https://www.ahajournals.org/doi/10.1161/circ.140.suppl_1.11463.
5. Su JR, McNeil MM, Welsh KJ, Marquez PL, Ng C, Yan M, et al. Myopericarditis after vaccination, Vaccine Adverse Event Reporting System (VAERS), 1990–2018. Vaccine [Internet]. Vaccine. 2021 [cited 2022 Aug 29];39:839–45. Available from: https://pubmed.ncbi.nlm.nih.gov/33422381/.
6. Polack FP, Thomas SJ, Kitchin N, Absalon J, Gurtman A, Lockhart S, et al. Safety and efficacy of the BNT162b2 mRNA Covid-19 vaccine. N Engl J Med [Internet]. Massachusetts Medical Society; 2020 [cited 2022 Aug 29];383:2603–15. Available from: https://www.nejm.org/doi/full/10.1056/nejmoa2034577.
7. Baden LR, El Sahly HM, Essink B, Kotloff K, Frey S, Novak R, et al. Efficacy and safety of the mRNA-1273 SARS-CoV-2 vaccine. N Engl J Med [Internet]. Massachusetts Medical Society; 2021 [cited 2022 Sep 2];384:403–16. Available from: https://www.nejm.org/doi/full/10.1056/nejmoa2035389.
8. Bozkurt B, Kamat I, Hotez PJ. Myocarditis with COVID-19 mRNA vaccines. Circulation [Internet]. Lippincott Williams & WilkinsHagerstown, MD; 2021 [cited 2022 Aug 29];144:471–84. Available from: https://www.ahajournals.org/doi/abs/10.1161/CIRCULATIONAHA.121.056135.
9. Gubernot D, Jazwa A, Niu M, Baumblatt J, Gee J, Moro P, et al. U.S. population-based background incidence rates of medical conditions for use in safety assessment of COVID-19 vaccines. Vaccine [Internet]. Vaccine. 2021 [cited 2022 Aug 29];39:3666–77. Available from: https://pubmed.ncbi.nlm.nih.gov/34088506/.
10. Oster ME, Shay DK, Su JR, Gee J, Creech CB, Broder KR, et al. Myocarditis cases reported after mRNA-based COVID-19 vaccination in the US from December 2020 to August 2021. JAMA [Internet]. American Medical Association; 2022 [cited 2022 Aug 29];327:331–40. Available from: https://jamanetwork.com/journals/jama/fullarticle/2788346.
11. Shimabukuro TT, Nguyen M, Martin D, DeStefano F. Safety monitoring in the Vaccine Adverse Event Reporting System (VAERS). Vaccine [Internet]. NIH Public Access; 2015 [cited 2022 Aug 29];33:4398. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4632204/.
12. Witberg G, Barda N, Hoss S, Richter I, Wiessman M, Aviv Y, et al. Myocarditis after Covid-19 vaccination in a large health care organization. N Engl J Med [Internet]. Massachusetts Medical Society; 2021 [cited 2022 Aug 29];385:2132–9. Available from: https://www.nejm.org/doi/10.1056/NEJMoa2110737.
13. Husby A, Hansen JV, Fosbøl E, Thiesson EM, Madsen M, Thomsen RW, et al. SARS-CoV-2 vaccination and myocarditis or myopericarditis: population based cohort study. BMJ. BMJ Publishing Group; 2021;375.
14. Patone M, Mei XW, Handunnetthi L, Dixon S, Zaccardi F, Shankar-Hari M, et al. Risk of myocarditis after sequential doses of COVID-19 vaccine and SARS-CoV-2 infection by age and sex. Circulation [Internet]. Lippincott Williams & Wilkins Hagerstown, MD ; 2022 [cited 2022 Aug 29]; Available from: http://www.ncbi.nlm.nih.gov/pubmed/35993236.
15. Chua GT, Kwan MYW, Chui CSL, Smith RD, Cheung EC-L, Ma T, et al. Epidemiology of acute myocarditis/pericarditis in Hong Kong adolescents following comirnaty vaccination. Clin Infect Dis. Oxford University Press (OUP); 2021.
16. Mascellino MT, Di Timoteo F, De Angelis M, Oliva A. Overview of the main anti-SARS-CoV-2 vaccines: mechanism of action, efficacy and safety. Infect Drug Resist [Internet]. Dove Press; 2021 [cited 2022 Sep 30];14:3459. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8418359/.
17. Cox RJ, Brokstad KA. Not just antibodies: B cells and T cells mediate immunity to COVID-19. Nat Rev Immunol 2020 2010 [Internet]. Nature Publishing Group; 2020 [cited 2022 Sep 30];20:581–2. Available from: https://www.nature.com/articles/s41577-020-00436-4.
18. Cox RJ, Brokstad KA. Not just antibodies: B cells and T cells mediate immunity to COVID-19. Nat Rev Immunol 2020 2010 [Internet]. Nature Publishing Group; 2020 [cited 2022 Aug 29];20:581–2. Available from: https://www.nature.com/articles/s41577-020-00436-4.
19. Sette A, Crotty S. Adaptive immunity to SARS-CoV-2 and COVID-19. Cell [Internet]. Elsevier; 2021 [cited 2022 Aug 29];184:861. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7803150/.
20. Shah S, Danda D, Kavadichanda C, Das S, Adarsh MB, Negi VS. Autoimmune and rheumatic musculoskeletal diseases as a consequence of SARS-CoV-2 infection and its treatment. Rheumatol Int [Internet]. Springer; 2020 [cited 2022 Sep 19];40:1539–54. Available from: https://link.springer.com/article/10.1007/s00296-020-04639-9.
21. Kim JY, Ko JH, Kim Y, Kim YJ, Kim JM, Chung YS, et al. Viral load kinetics of SARS-CoV-2 infection in first two patients in Korea. J Korean Med Sci [Internet]. Korean Academy of Medical Science; 2019 [cited 2022 Sep 19];35. 10.3346/jkms.2020.35.e86.
22. Ehrenfeld M, Tincani A, Andreoli L, Cattalini M, Greenbaum A, Kanduc D, et al. Covid-19 and autoimmunity. Autoimmun Rev [Internet]. Elsevier; 2020 [cited 2022 Sep 19];19:102597. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7289100/.
23. Neidleman J, Luo X, McGregor M, Xie G, Murray V, Greene WC, et al. Mrna vaccine-induced t cells respond identically to sars-cov-2 variants of concern but differ in longevity and homing properties depending on prior infection status. Elife. eLife Sciences Publications Ltd; 2021;10.
24. Murphy WJ, Longo DL. A possible role for anti-idiotype antibodies in SARS-CoV-2 infection and vaccination. N Engl J Med [Internet]. Massachusetts Medical Society; 2022 [cited 2022 Sep 19];386:394–6. Available from: https://www.nejm.org/doi/full/10.1056/NEJMcibr2113694.
25. Paque RE, Miller R. Autoanti-idiotypes exhibit mimicry of myocyte antigens in virus-induced myocarditis. J Virol [Internet]. American Society for Microbiology (ASM); 1991 [cited 2022 Sep 19];65:16. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC240484/?report=abstract.
26. Huber SA, Pfaeffle B. Differential Th1 and Th2 cell responses in male and female BALB/c mice infected with coxsackievirus group B type 3. J Virol [Internet]. J Virol; 1994 [cited 2022 Aug 29];68:5126–32. Available from: https://pubmed.ncbi.nlm.nih.gov/8035512/.
27. Kytö V, Sipilä J, Rautava P. Clinical profile and influences on outcomes in patients hospitalized for acute pericarditis. Circulation [Internet]. Lippincott Williams and Wilkins; 2014 [cited 2021 Feb 2];130:1601–6. Available from: https://www.ahajournals.org/doi/10.1161/CIRCULATIONAHA.114.010376.
28. Ammirati E, Cipriani M, Moro C, Raineri C, Pini D, Sormani P, et al. Clinical presentation and outcome in a contemporary cohort of patients with acute myocarditis: multicenter Lombardy Registry. Circulation [Internet]. Circulation; 2018 [cited 2022 Aug 29];138:1088–99. Available from: https://pubmed.ncbi.nlm.nih.gov/29764898/.
29. Kociol RD, Cooper LT, Fang JC, Moslehi JJ, Pang PS, Sabe MA, et al. Recognition and initial management of fulminant myocarditis. Circulation [Internet]. Lippincott Williams & WilkinsHagerstown, MD; 2020 [cited 2022 Aug 29];141:E69–92. Available from: https://www.ahajournals.org/doi/abs/10.1161/CIR.0000000000000745.
30. Corrado D, Basso C, Thiene G. Sudden cardiac death in young people with apparently normal heart. Cardiovasc Res [Internet]. Cardiovasc Res; 2001 [cited 2022 Aug 29];50:399–408. Available from: https://pubmed.ncbi.nlm.nih.gov/11334844/.
31. Cooper LT, Baughman KL, Feldman AM, Frustaci A, Jessup M, Kuhl U, et al. The role of endomyocardial biopsy in the management of cardiovascular disease: a scientific statement from the American Heart Association, the American College of Cardiology, and the European Society of Cardiology. Circulation [Internet]. Circulation; 2007 [cited 2022 Aug 29];116:2216–33. Available from: https://pubmed.ncbi.nlm.nih.gov/17959655/.
32. Adler Y, Charron P, Imazio M, Badano L, Barón-Esquivias G, Bogaert J, et al. 2015 ESC Guidelines for the diagnosis and management of pericardial diseases. Eur. Heart J. Oxford University Press; 2015. p. 2921–64.
33. Ambati S, Colon M, Mihic M, Sanchez J, Bakar A. Acute myopericarditis after COVID-19 vaccine in teenagers. Case reports Cardiol [Internet]. Case Rep Cardiol; 2021 [cited 2022 Aug 29];2021:1–5. Available from: https://pubmed.ncbi.nlm.nih.gov/34589238/.
34. Fleming-Nouri A, Haimovich AD, Yang D, Schulz WL, Coppi A, Taylor RA. Myopericarditis in young adults presenting to the emergency department after receiving a second COVID‐19 mRNA vaccine. Acad Emerg Med [Internet]. Wiley-Blackwell; 2021 [cited 2022 Aug 29];28:802. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8441914/
35. Sakaguchi S, Fujimoto N, Ichikawa K, Izumi D, Katsuta K, Takafuji M, et al. Myopericarditis after COVID-19 mRNA vaccination. Circ J [Internet]. Circ J; 2022 [cited 2022 Aug 29];86:472. Available from: https://pubmed.ncbi.nlm.nih.gov/34588400/.
36. Chen DH, Arefin AR, Joshi A, Khanji MY. Myopericarditis in a teenager following first mRNA COVID vaccine dose: the role of multi-parametric cardiovascular magnetic resonance. Eur Hear J Case Reports [Internet]. Oxford University Press; 2021 [cited 2022 Aug 29];5. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8493007/.
37. Das BB, Kohli U, Ramachandran P, Nguyen HH, Greil G, Hussain T, et al. Myopericarditis after messenger RNA coronavirus disease 2019 vaccination in adolescents 12 to 18 years of age. J Pediatr [Internet]. Elsevier Inc.; 2021 [cited 2022 Aug 29];238:26–32.e1. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8321962.
38. Nygaard U, Holm M, Bohnstedt C, Chai Q, Schmidt LS, Hartling UB, et al. Population-based incidence of myopericarditis after COVID-19 vaccination in Danish adolescents. Pediatr Infect Dis J [Internet]. Pediatr Infect Dis J; 2022 [cited 2022 Aug 29];41:E25–8. Available from: https://pubmed.ncbi.nlm.nih.gov/34889875/.
39. Bartlett VL, Thomas A, Hur DJ, Malm B. Myopericarditis with significant left ventricular dysfunction following COVID-19 vaccination: a case report. Am J Case Rep [Internet]. Am J Case Rep; 2021 [cited 2022 Aug 29];22. Available from: https://pubmed.ncbi.nlm.nih.gov/34795198/.
40. Badshah M, Shriver J, Rynders B, Sjovold A, Shaukat MHS, Rajpurohit N. MODERNA mRNA-1273 vaccine-associated myopericarditis in a patient with a subclinical autoimmune predisposition. J Cardiol Cases [Internet]. Japanese College of Cardiology; 2021 [cited 2022 Aug 29];24:227. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8617476/.
41. Hung YP, Sun KS. A case of myopericarditis with pleuritis following AstraZeneca Covid-19 vaccination. QJM [Internet]. QJM; 2022 [cited 2022 Aug 29];114:879–81. Available from: https://pubmed.ncbi.nlm.nih.gov/34741522/.
42. McLean K, Johnson TJ. Myopericarditis in a previously healthy adolescent male following COVID-19 vaccination: a case report. Acad Emerg Med [Internet]. Acad Emerg Med; 2021 [cited 2022 Aug 29];28:918–21. Available from: https://pubmed.ncbi.nlm.nih.gov/34133825/.
43. Schauer J, Buddhe S, Colyer J, Sagiv E, Law Y, Mallenahalli Chikkabyrappa S, et al. Myopericarditis after the Pfizer messenger ribonucleic acid coronavirus disease vaccine in adolescents. J Pediatr [Internet]. Elsevier; 2021 [cited 2022 Aug 29];238:317. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8253718/.
44. Farooq M, Mohammed Y, Zafar M, Dharmasena D, Rana UI, Kankam O. COVID-19 vaccine-induced pneumonitis, myositis and myopericarditis. Cureus [Internet]. Cureus; 2022 [cited 2022 Aug 29];14. Available from: https://www.cureus.com/articles/82331-covid-19-vaccine-induced-pneumonitis-myositis-and-myopericarditis.
45. Umei TC, Kishino Y, Watanabe K, Shiraishi Y, Inohara T, Yuasa S, et al. Recurrence of myopericarditis following mRNA COVID-19 vaccination in a male adolescent. CJC open [Internet]. CJC Open; 2022 [cited 2022 Aug 29];4:350–2. Available from: https://pubmed.ncbi.nlm.nih.gov/34904134/.
46. Kim D, Choi JH, Jang JY, So O, Cho EJ, Choi H, et al. A case report for myopericarditis after BNT162b2 COVID-19 mRNA vaccination in a Korean young male. J Korean Med Sci [Internet]. J Korean Med Sci; 2021 [cited 2022 Aug 29];36:1–7. Available from: https://pubmed.ncbi.nlm.nih.gov/34636504/.
47. Ameratunga R, Woon ST, Sheppard MN, Garland J, Ondruschka B, Wong CX, et al. First identified case of fatal fulminant necrotizing eosinophilic myocarditis following the initial dose of the Pfizer-BioNTech mRNA COVID-19 vaccine (BNT162b2, Comirnaty): an extremely rare idiosyncratic hypersensitivity reaction. J Clin Immunol [Internet]. Nature Publishing Group; 2022 [cited 2022 Aug 29];42:441. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8720536/.
48. Ashaari S, Sohaib HA, Bolger K. A case report: symptomatic pericarditis post-COVID-19 vaccination. Eur Hear journal Case reports [Internet]. Eur Heart J Case Rep; 2021 [cited 2022 Aug 29];5. Available from: https://pubmed.ncbi.nlm.nih.gov/34693198/.
49. D’Angelo T, Cattafi A, Carerj ML, Booz C, Ascenti G, Cicero G, et al. Myocarditis after SARS-CoV-2 vaccination: a vaccine-induced reaction? Can J Cardiol [Internet]. Elsevier; 2021 [cited 2022 Aug 29];37:1665. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8187737/.
50. Marshall M, Ferguson ID, Lewis P, Jaggi P, Gagliardo C, Collins JS, et al. Symptomatic acute myocarditis in 7 adolescents after pfizer-biontech covid-19 vaccination. Pediatrics [Internet]. American Academy of Pediatrics; 2021 [cited 2022 Aug 29];148. Available from: https://publications.aap.org/pediatrics/article/148/3/e2021052478/179728/Symptomatic-Acute-Myocarditis-in-7-Adolescents.
51. Facetti S, Giraldi M, Vecchi AL, Rogiani S, Nassiacos D. [Acute myocarditis in a young adult two days after Pfizer vaccination]. G Ital Cardiol (Rome) [Internet]. G Ital Cardiol (Rome); 2021 [cited 2022 Aug 29];22:891–3. Available from: https://pubmed.ncbi.nlm.nih.gov/34709227/.
52. King WW, Petersen MR, Matar RM, Budweg JB, Pardo LC, Petersen JW. Myocarditis following mRNA vaccination against SARS-CoV-2, a case series. Am Hear J plus [Internet]. Elsevier; 2021 [cited 2022 Aug 29];8:100042. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8349733/.
53. Hasnie AA, Hasnie UA, Patel N, Aziz MU, Xie M, Lloyd SG, et al. Perimyocarditis following first dose of the mRNA-1273 SARS-CoV-2 (Moderna) vaccine in a healthy young male: a case report. BMC Cardiovasc Disord [Internet]. BioMed Central Ltd; 2021 [cited 2022 Aug 29];21:1–6. Available from: https://bmccardiovascdisord.biomedcentral.com/articles/10.1186/s12872-021-02183-3.
54. Patrignani A, Schicchi N, Calcagnoli F, Falchetti E, Ciampani N, Argalia G, et al. Acute myocarditis following Comirnaty vaccination in a healthy man with previous SARS-CoV-2 infection. Radiol Case Reports [Internet]. Elsevier; 2021 [cited 2022 Aug 29];16:3321. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8326008/.
55. Howlett JG, McKelvie RS, Arnold JMO, Costigan J, Dorian P, Ducharme A, et al. Canadian Cardiovascular Society Consensus Conference guidelines on heart failure, update 2009: diagnosis and management of right-sided heart failure, myocarditis, device therapy and recent important clinical trials. Can J Cardiol [Internet]. Can J Cardiol; 2009 [cited 2022 Aug 29];25:85–105. Available from: https://pubmed.ncbi.nlm.nih.gov/19214293/.
56. Imazio M Brucato A Barbieri A Ferroni F Maestroni S Ligabue G Good prognosis for pericarditis with and without myocardial involvement: results from a multicenter, prospective cohort study Circulation 2013 128 42 49 10.1161/CIRCULATIONAHA.113.001531 23709669
57. AHA. COVID-19 vaccine benefits still outweigh risks, despite possible rare heart complications | American Heart Association [Internet]. AHA Statement. 2021 [cited 2022 Sep 2]. Available from: https://newsroom.heart.org/news/covid-19-vaccine-benefits-still-outweigh-risks-despite-possible-rare-heart-complications.
58. Cavalcante JL Shaw KA Gössl M Cardiac magnetic resonance imaging midterm follow up of COVID-19 vaccine–associated myocarditis JACC Cardiovasc. Imaging 2022 15 10 1821 1824 10.1016/j.jcmg.2022.01.008 36202461
59. Moulson N, Petek BJ, Drezner JA, Harmon KG, Kliethermes SA, Patel MR, et al. SARS-CoV-2 cardiac involvement in young competitive athletes. Circulation [Internet]. Lippincott Williams & WilkinsHagerstown, MD; 2021 [cited 2022 Sep 15];144:256–66. Available from: https://www.ahajournals.org/doi/abs/10.1161/CIRCULATIONAHA.121.054824.
60. • Gluckman TJ, Bhave NM, Allen LA, Chung EH, Spatz ES, Ammirati E, et al. 2022 ACC expert consensus decision pathway on cardiovascular sequelae of COVID-19 in adults: myocarditis and other myocardial involvement, post-acste Sequelae of SARS-CoV-2 infection, and return to play: a report of the American College of Cardiology Solution Set Oversight Committee. J Am Coll Cardiol [Internet]. Elsevier; 2022 [cited 2022 Sep 15];79:1717. Available from: /pmc/articles/PMC8926109/. American College of Cardiology issued a consensus report on recommendations regarding return to play in athelets with myocarditis after COVID-19 infection. This study is a detailed expert opinion on young active patient population including athletes affected with myocarditis secondary to COVID-19 infection.
61. Ryan M, Montgomery J, Engler R, Hoffman D, McClenathan B, Collins L, et al. Myocarditis following immunization with mRNA COVID-19 vaccines in members of the US military. JAMA Cardiol [Internet]. American Medical Association; 2021 [cited 2022 Sep 15];6:1202–6. Available from: https://jamanetwork.com/journals/jamacardiology/fullarticle/2781601.
62. Kauth M, Kovacs RJ. Recognizing and managing myocarditis following COVID-19 vaccination: mitigating risk of sudden cardiac death in athletes. J Am Coll Cardiol [Internet]. Elsevier; 2022 [cited 2022 Sep 15];79:2397. Available from: /pmc/articles/PMC8972386/.
63. Shah NP Verma BR Ala CK Khayata M Phelan D Imazio M Exercise is good for the heart but not for the inflamed pericardium? JACC Cardiovasc Imaging. 2019 12 1880 1 10.1016/j.jcmg.2019.01.022 30878417
64. Berglund F, Klein AL. Is exercise restriction necessary in patients with pericarditis? Cleve Clin J Med [Internet]. Cleveland Clinic Journal of Medicine; 2022 [cited 2022 Sep 15];89:437–41. Available from: https://www.ccjm.org/content/89/8/437.
65. Patone M, Mei XW, Handunnetthi L, Dixon S, Zaccardi F, Shankar-Hari M, et al. Risks of myocarditis, pericarditis, and cardiac arrhythmias associated with COVID-19 vaccination or SARS-CoV-2 infection. Nat Med 2021 282 [Internet]. Nature Publishing Group; 2021 [cited 2022 Sep 15];28:410–22. Available from: https://www.nature.com/articles/s41591-021-01630-0.
66. Ammirati E, Lupi L, Palazzini M, Hendren NS, Grodin JL, Cannistraci C V., et al. Prevalence, characteristics, and outcomes of COVID-19-associated acute myocarditis. Circulation [Internet]. Circulation; 2022 [cited 2022 Sep 15];145:1123–39. Available from: https://pubmed.ncbi.nlm.nih.gov/35404682/.
67. •• Chou OHI, Zhou J, Lee TTL, Kot T, Lee S, Wai AKC, et al. Comparisons of the risk of myopericarditis between COVID-19 patients and individuals receiving COVID-19 vaccines: a population-based study. Clin Res Cardiol [Internet]. Clin Res Cardiol; 2022 [cited 2022 Sep 15]; Available from: https://pubmed.ncbi.nlm.nih.gov/35333945/. This is an important study comparing the risk of myocarditis with COVID-19 infection and vaccination and findings from this study suggest that the rate of myocarditis after COVID-19 infection is 326 per million as compared to 5.5 per million after COVID-19 vaccination.
68. Brucato A, Trotta L, Arad M, Cremer PC, Gaddam E, Insalaco A, et al. Abstract 12057: Absence of pericarditis recurrence in rilonacept-treated patients with COVID-19 and mRNA vaccinations: Experience from phase 3 rhapsody long-term extension. Circulation. 2022;146:A12057. Available from: https://www.ahajournals.org/doi/abs/10.1161/circ.146.suppl_1.12057.
69. Gurdasani D, Bhatt S, Costello A, Denaxas S, Flaxman S, Greenhalgh T, et al. Vaccinating adolescents against SARS-CoV-2 in England: a risk-benefit analysis. J R Soc Med [Internet]. J R Soc Med; 2021 [cited 2022 Sep 15];114:513–24. Available from: https://pubmed.ncbi.nlm.nih.gov/34723680/.
| 36441403 | PMC9703393 | NO-CC CODE | 2022-12-14 23:47:19 | no | Curr Cardiol Rep. 2022 Nov 28; 24(12):2031-2041 | utf-8 | Curr Cardiol Rep | 2,022 | 10.1007/s11886-022-01801-6 | oa_other |
==== Front
J Clin Immunol
J Clin Immunol
Journal of Clinical Immunology
0271-9142
1573-2592
Springer US New York
36441289
1408
10.1007/s10875-022-01408-0
Letter to Editor
Macrophage Activation Syndrome Complicated by Toxic Epidermal Necrolysis Following SARS-CoV-2 mRNA Vaccination
http://orcid.org/0000-0002-4502-5733
Franzblau Lauren E. [email protected]
1
Mauskar Melissa [email protected]
2
Wysocki Christian A. [email protected]
1
1 grid.267313.2 0000 0000 9482 7121 Department of Internal Medicine, Division of Allergy and Immunology, University of Texas Southwestern, 5323 Harry Hines Boulevard, Building F, Room F04100B, Dallas, TX 75390 USA
2 grid.267313.2 0000 0000 9482 7121 Department of Dermatology, University of Texas Southwestern, Dallas, TX USA
28 11 2022
14
28 5 2022
14 11 2022
© The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2022
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Keywords
Macrophage activation syndrome
Steven Johnson syndrome
toxic epidermal necrolysis
vaccine adverse effects
SARS-CoV-2 vaccine
==== Body
pmcTo the Editor:
SARS-CoV-2 vaccines have been rapidly deployed worldwide. They have proved to be generally safe and effective at preventing severe disease by eliciting a strong immune response with frequent, albeit usually mild, inflammatory side effects (e.g., fever, myalgias). More serious inflammatory complications have been reported infrequently, such as vaccine-associated myocarditis. Here, we report a case of macrophage activation syndrome (MAS) following SARS-CoV-2 vaccination, which was complicated by toxic epidermal necrolysis (TEN), and provide a tabular review of similar cases of MAS and/or adult-onset Still’s disease (AOSD) occurring in association with SARS-CoV-2 vaccination.
A 23-year-old previously healthy woman developed daily fevers and cervical lymphadenopathy 1 week after receiving her second dose of the Moderna SARS-CoV-2 mRNA vaccine. She was initially seen at an urgent care center and prescribed azithromycin and corticosteroids for a presumed upper respiratory infection. Her symptoms did not abate, and, over the next 10 days, she developed rash, nausea, vomiting, and abdominal pain, leading her to present to the emergency department. On admission (18 days after vaccination), her examination was notable for temperature of 39.1 °C, sinus tachycardia (140 bpm), hypotension (88/66 mmHg), enlarged cervical lymph nodes, and a faint blanchable morbilliform eruption on all extremities.
Initial evaluation demonstrated leukocytosis (18.19 × 109/L [4–11 × 109/L]) with 93% neutrophils; hemoglobin, 11.4 g/dL (12–15 g/dL); platelets, 93 × 109/L (150–450 × 109/L), elevated liver enzymes (AST, 130 U/L [10–50 U/L]; ALT, 43 U/L [10–35 U/L]); and CRP, 186.1 mg/L (≤ 5 mg/L). Computed tomography of her chest, abdomen, and pelvis revealed splenomegaly and diffuse lymphadenopathy. Due to concern for infection, blood cultures were obtained, and broad-spectrum antibiotics were initiated. Her negative cultures and a lack of improvement after multiple courses of antibiotics raised concern for malignancy or autoimmune disease. She underwent excisional lymph node biopsy as well as bone marrow biopsy. Allopurinol was started given concern for lymphoma and risk of tumor lysis syndrome and later discontinued when biopsy pathology and flow cytometry did not show signs of monoclonal expansion/proliferation, therefore ruling out hematolymphoid malignancy. Her biopsies showed no hemophagocytosis and negative bacterial, fungal, and mycobacterial staining and cultures. EBV PCR, IgM, and IgG as well as SARS-CoV-2 PCR and nuclear capsid IgG testing were also negative, arguing against prior infection. Autoimmune serologies were all within normal limits. She met 5 of 8 diagnostic criteria for hemophagocytic lymphohistiocytosis (HLH), with fevers; splenomegaly; ferritin, 56,131 ng/mL (5–204 ng/mL); triglycerides, 811 mg/dL (< 150 mg/dL); and elevated soluble interleukin (IL) 2-receptor, 6158.5 pg/mL (175.3–858.2 pg/mL). Interestingly, she also met Yamaguchi criteria for AOSD and had marked elevations in both IL-18 (181,803 pg/mL [89–540 pg/mL]) and CXCL9 (548,075 pg/mL [< 647 pg/mL]), supporting the diagnosis of MAS [1].
Her illness course, inflammatory markers, and therapies are shown in Fig. 1A. Dexamethasone and anakinra were initiated on day 36 after vaccination along with prophylactic atovaquone and pantoprazole. Her symptoms and labs improved despite a brief lapse in anakinra, following hospital discharge, until day 64 after vaccination, when she developed a painful, desquamating, and dusky Nikolsky positive eruption, involving 70% of her total body surface area and mucosae consistent with TEN (Fig. 1B). Skin biopsy later confirmed subepidermal vesicular dermatitis with full-thickness epidermal necrosis (Fig. 1C). Potential culprit drugs were held (pantoprazole, atovaquone, and cephalosporins), and she received intravenous immunoglobulin (IVIG) 500 mg/kg for 3 days and cyclosporine (CsA) 150 mg BID. Although she no longer had symptoms fulfilling Yamaguchi criteria for AOSD, her IL-18 remained elevated, and, therefore, anakinra was restarted, and steroids were continued to address her underlying MAS. Her rash quickly stopped progressing, and her inflammatory markers trended down. She required prolonged hospitalization for wound care and rehabilitation. Following discharge, she transitioned from anakinra 100 mg daily to canakinumab 300 mg every 28 days, and prednisone was tapered off. She has continued to do well on canakinumab 300 mg every 4 weeks, without recurrence of MAS or TEN symptoms. Testing for HLA-B*5801, which is associated with allopurinol-related Steven Johnson syndrome (SJS)-TEN, was negative. Informed consent was obtained from the patient for publication of this case.Fig. 1 Time course of illness, clinical photographs, and histopathology
This patient without history of autoimmune disease developed AOSD with MAS in close proximity to SARS-CoV-2 vaccination. Her course was later complicated by TEN. MAS was diagnosed based on markedly elevated IL-18 and fulfillment of Yamaguchi criteria for AOSD. Marked elevations in IL-18 are a distinguishing feature of both AOSD and MAS. Total IL-18 greater than 24,000 pg/mL differentiates MAS from other types of familial and secondary HLH (sensitivity, 83%; specificity, 94%). Although IL-18 alone cannot differentiate between AOSD and MAS, it tends to be an order of magnitude higher in MAS [1]. Our patient also had unusually high CXCL9, which is a Th1 chemokine induced by interferon gamma production. CXCL9 elevations are seen with MAS, as well as HLH, though typically on the order of 1000–10,000 pg/mL [1].
Both new-onset AOSD and exacerbations of preexisting disease have been reported, following various types of SARS-CoV-2 vaccination (See Supplemental Table 1 and accompanying references). Neither IL-18 nor CXCL9 levels were reported in these publications. Most cases occurred within 1–2 weeks of vaccination, and, as in many other autoimmune conditions, there appears to be a female predominance. Similar to infection, vaccination is a possible trigger for AOSD with or without MAS in predisposed individuals. Both conditions are marked by widespread innate immune activation. Treatment involves IL-1 blockade as well as corticosteroids [2]. Anakinra and canakinumab were used successfully in this case as well as others reported in the literature [3]. Our patient has avoided additional SARS-CoV-2 boosters and, instead, was given tixagevimab and cilgavimab for COVID prophylaxis.
Although our patient responded appropriately to MAS treatment, she was exposed to multiple medications during the early, uncontrolled phase of MAS, which may have increased her risk of developing TEN. The presence of inflammatory disorders, such as autoimmune disease, increases the risk of TEN, particularly in the early stages before the disease is controlled [4]. Inflammation promotes antigen presentation and activation of T cells and NK cells, leading to drug-specific CD8 + T cell and NK cell-mediated epidermal damage [5]. In this case, the lapse in anakinra due to delayed insurance coverage following discharge coincided with a resurgence of her inflammatory markers and development of TEN. It is possible that her immunosuppression had masked or prevented TEN and that this gap allowed it to fully manifest. Additionally, certain medications and HLA types are associated with higher risk of SJS-TEN, including allopurinol. The cause of our patient’s TEN remains unknown. Although HLA-B*5801 testing was negative, this does not rule out allopurinol. Treatment of SJS-TEN involves stopping suspected culprit drugs (in this case, allopurinol, atovaquone, pantoprazole, and cephalosporins), wound care, and, typically, immunosuppression. Recent studies have supported the use of cyclosporine and the combination of IVIG plus corticosteroids [5].
The SARS-CoV-2 pandemic led to the rapid development and dissemination of multiple life-saving vaccines. Our report should not dissuade the continued use of these vaccines but provide insight into the presentation of rare inflammatory complications including AOSD and MAS. This case and other reports suggest that these sequelae are responsive to typical immunosuppression, namely IL-1 blockade and corticosteroids. In at least one case, the patient was able to safely receive her 2nd dose once the disease was controlled [3]. Finally, in patients with uncontrolled inflammatory disease, there is an increased risk of SJS/TEN, and development of new mucocutaneous symptoms deserves prompt evaluation and treatment.
Supplementary Information
Below is the link to the electronic supplementary material.Supplementary file1 (DOCX 29 KB)
Data Availability
All data generated or analysed during this study are included in this published article (and its supplementary information files).
Declarations
Conflict of Interest
The authors declare no competing interests.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
1. Weiss ES Girard-Guyonvarc’h C Holzinger D De Jesus AA Tariq Z Picarsic J Interleukin-18 diagnostically distinguishes and pathogenically promotes human and murine macrophage activation syndrome Blood 2018 131 13 1442 55 10.1182/blood-2017-12-820852 29326099
2. Grom AA Horne A De Benedetti F Macrophage activation syndrome in the era of biologic therapy Nat Rev Rheumatol 2016 12 5 259 268 10.1038/nrrheum.2015.179 27009539
3. Bindoli S Giollo A Galozzi P Doria A Sfriso P Hyperinflammation after anti-SARS-CoV-2 mRNA/DNA vaccines successfully treated with anakinra: case series and literature review Exp Biol Med 2022 247 4 338 344 10.1177/15353702211070290
4. Gronich N Maman D Stein N Saliba W Culprit medications and risk factors associated with Stevens-Johnson syndrome and toxic epidermal necrolysis: population-based nested case–control study Am J Clin Dermatol 2022 23 2 257 266 10.1007/s40257-021-00661-0 35119606
5. Tsai TY Huang IH Chao YC Li H Hsieh TS Wang HH Treating toxic epidermal necrolysis with systemic immunomodulating therapies: a systematic review and network meta-analysis J Am Acad Dermatol 2021 84 2 390 397 10.1016/j.jaad.2020.08.122 32898587
| 36441289 | PMC9703394 | NO-CC CODE | 2022-11-29 23:21:42 | no | J Clin Immunol. 2022 Nov 28;:1-4 | utf-8 | J Clin Immunol | 2,022 | 10.1007/s10875-022-01408-0 | oa_other |
==== Front
Allergo J Int
Allergo J Int
Allergo Journal International
2197-0378
Springer Medizin Heidelberg
234
10.1007/s40629-022-00234-5
Case Report
Localized eczematous rash affecting left and right regions of breast and shoulder after Ad26.COV2.S vaccine against COVID-19 in a 30-year-old woman with comorbidities
Anasiewicz Natalie 12
Seeli Corsin 1
Brüggen Marie-Charlotte 123
Möhrenschlager Matthias [email protected]
4
1 grid.483388.c 0000 0000 8632 4866 Department of Dermatology, Hochgebirgsklinik, Davos, Switzerland
2 grid.507894.7 0000 0004 4700 6354 Christine Kühne Center for Allergy Research and Education, Davos, Switzerland
3 grid.7400.3 0000 0004 1937 0650 Department of Dermatology, University of Zurich, Zurich, Switzerland
4 grid.483388.c 0000 0000 8632 4866 Department of Dermatology, Hochgebirgsklinik, Herman-Burchard-Street 1, 7265 Davos Wolfgang, Switzerland
28 11 2022
12
14 9 2022
16 10 2022
© Springer Medizin Verlag GmbH, ein Teil von Springer Nature 2022
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
==== Body
pmcBackground
Cutaneous reactions after Coronavirus 2019 (COVID-19) vaccination with mRNA-1273 (Moderna, Cambridge, MA, USA), and BNT162b2 (Pfizer-BioNTech, New York, NY, USA) have been commonly reported [1–4]. In contrast, Ad26.COV2.S (Johnson & Johnson, New Brunswick, NJ, USA), which uses a nonreplicating viral vector, seems to have relatively fewer dermatological side effects [5].
Case presentation
We report the case of a 30-year-old woman who showed localized itching coin-like erythematous macules with scaling beginning at her left breast and left shoulder region 24 h after Ad26.COV2.S vaccination (Fig. 1) on her left upper arm. After another 24 h, her right breast and right shoulder showed similar cutaneous alterations (Fig. 2). Any involvement at the injection site was lacking. Unfortunately, a lesional biopsy was not approved by the patient. Systemic prednisolone and bilastine, and lesional topical application of mometason fuorate cream resulted in a full restitutio ad integrum. Nevertheless, 4 weeks after Ad26.COV2.S vaccination, the patient was infected by corona virus (as demonstrated by positive SARS-CoV‑2 PCR test). She encountered a moderate COVID-19 symptom severity score with fever and cough resolving at home without new cutaneous alterations.Fig. 1 Coin-like eczema-like erythematous lesion affecting left shoulder starting 24 h after first Ad26.COV2.S shot
Fig. 2 Coin-like eczema-like erythematous lesion affecting right shoulder starting 48 h after first Ad26.COV2.S shot
Patient’s medical history revealed a body mass index (BMI) > 25 kg/m2, allergic asthma, allergic rhinoconjunctivitis, drug allergies (paracetamol, ibuprofen), blood hypertension, endometriosis, and sleep apnoea syndrome. In regard to type IV allergic reactions, a positive sensitization to diethylenetriamine, a solvent for plastics and dyes and in chemical synthesis was confirmed by patch test. In regard to drugs, telmisartan 80 mg/amlodipine 5 mg tablet, dienogest 2 mg tablet, and formoterol 400 mcg/budesonide 12 mcg inhaler had been in continuous use for over 1 year. For sleep apnoea syndrome, the patient used an auto-adjusting positive airway pressure device during night hours. In regard to skin alterations in the past, the patient denied any cutaneous lesions except acne pustules during puberty.
In a large sample of cutaneous COVID-19 vaccine reactions to mRNA-1273, BNT162b2, and Ad26.COV2.S, McMahon et al. found robust papules with overlying crusts, pityriasis rosea-like eruptions, pink papules with fine scale (V-REPP), bullous pemphigoid-like lesions, dermal hypersensitivity, herpes zoster, lichen-planus-like lesions, urticarial, neutrophilic dermatosis, leukocytoclastic vasculitis, morbilliform, delayed large local reactions, erythromelalgia, and others [1].
Although macroscopic lesional inspection in our patient was consistent with the diagnosis of nummular eczema, dermal hypersensitivity, and id reaction, (defined as dermatitis distant to an initial site of inflammation or infection) must also be considered.
Conclusion
The cause of the encountered localized eczematous reactions remain unclear. It is possible that Ad26.COV2.S vaccine may act as an environmental trigger in a genetically susceptible individual. In our case, the patient suffered from allergic asthma and rhinoconjunctivitis, had a positive type IV sensitization to diethylenetriamine, and had a medical history of drug intolerances, perhaps making her more susceptible to an eczematous reaction. It is a matter of discussion whether this cutaneous reaction would have occurred in a similar way after administration of Moderna’s mRNA-1273 and Pfizer’s BNT162b2 vaccine. Further recording of cutaneous reactions following vaccination with Ad26.COV2.S seem mandatory to provide a complete picture of possible side effects.
Conflict of interest
N. Anasiewicz, C. Seeli, M.-C. Brüggen and M. Möhrenschlager declare that they have no competing interests.
==== Refs
References
1. McMahon DE Kovarik CL Damsky W Rosenbach M Lipoff JB Tyagi A Clinical and pathologic correlation of cutaneous COVID-19 vaccine reactions including V-REPP: a registry-based study J Am Acad Dermatol 2022 86 113 121 10.1016/j.jaad.2021.09.002 34517079
2. Oulee A Salem S Yahia R Yang K Garcia D Holmes A Cutaneous reactions due to Pfizer’s BNT162b2 mRNA and Moderna’s mRNA-1273 vaccine J Eur Acad Dermatol Venereol 2022 36 e332 4 10.1111/jdv.17925 35028998
3. Leasure AC Cowper SE McNiff J Cohen JM Generalized eczematous reactions to the Pfizer-Biontech COVID-19 vaccine J Eur Acad Dermatol Venereol 2021 35 e716 7 10.1111/jdv.17494 34236729
4. Schmidt V Blum R Möhrenschlager M Biphasic bullous pemphigoid starting after first dose and boosted by second dose of mRNA-1273 vaccine in a 84-year-old female with polymorbidity and polypharmacy J Eur Acad Dermatol Venereol 2022 36 e88 90 10.1111/jdv.17722 34606112
5. Sadoff J Gray A Vandebosch A Cárdenas V Shukarev G Grinsztejn B Safety and efficacy of single-dose Ad26.COV2.S vaccine against Covid-19 N Engl J Med 2021 384 1824 1835 10.1056/NEJMoa2034201 33440088
| 36466142 | PMC9703396 | NO-CC CODE | 2022-11-29 23:21:08 | no | Allergo J Int. 2022 Nov 28;:1-2 | utf-8 | Allergo J Int | 2,022 | 10.1007/s40629-022-00234-5 | oa_other |
==== Front
Int J Sociol Leis
International Journal of the Sociology of Leisure
2520-8683
2520-8691
Springer International Publishing Cham
124
10.1007/s41978-022-00124-8
Original Paper
The Unfortunate Inner Lives of Scholars of Color in Leisure and Tourism Studies
http://orcid.org/0000-0001-7265-9013
Bandyopadhyay Ranjan [email protected]
grid.444918.4 0000 0004 1794 7022 Institute for Social and Economic Research, Duy Tan University, Da Nang, Viet Nam
28 11 2022
118
17 2 2022
10 11 2022
© The Author(s), under exclusive licence to Springer Nature Switzerland AG 2022, Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
In this conceptual paper, I tried to articulate that in leisure and tourism studies “we still live in a wholly racialized world” (Morrison, 1992). Few leisure and tourism scholars cared to follow the clues to map the contours of the racial predicament of scholars of color as a way of their lives surviving in the academia. As a scholar of color, my everlasting quest has always been to feel at home without becoming “White”. The dilemmas and rejections in this journey created an omnipresent tension in my life which shaped the content of this paper. I understand that this study will certainly not set the Thames on fire but I seek to open new avenues of discussion to break this silence. While doing that, I tried to follow the philosophy of Hegel’s “master/slave dialectic: the search for self-consciousness” within the Bakhtinian (multiaccentuality of racial meaning) and Levinasian (his close equivalence between structuralist anthropology and genetics) context equipped with the wisdom of Stuart Hall, Frantz Fanon, W.E.B. Du Bois, Toni Morrison, Michel Foucault, Karl Marx, Jacques Derrida and Amartya Sen.
Supplementary Information
The online version contains supplementary material available at 10.1007/s41978-022-00124-8.
Keywords
Scholars of color
Race
Master/slave Dialectic
Leisure and Tourism Studies
Transformation
==== Body
pmcIntroduction
“We were lesser. Nicer, brighter, but still lesser … Guileless and without vanity, we were still in love with ourselves then. We felt comfortable in our skins, enjoyed the news that our senses released to us, admired our dirt, cultivated our scars, and could not comprehend this unworthiness”
(Toni Morrison, 1994, p. 27) – The Bluest Eye
“Where the mind is without fear and the head is held high
Where knowledge is free
Where the world has not been broken up into fragments by narrow domestic walls”
(Rabindranath Tagore, 1912, p. 11) - Gitanjali.
“The emergence of an idea of ‘the West’ was central to the Enlightenment (…).
The Enlightenment was a very European affair. European society, it assumed, was the most advanced type of society on earth, European man [sic] the pinnacle of human achievement. It treated the West as the result of forces largely internal to Europe’s history and formation”
(Stuart Hall, 1992, p. 37) - The West and the Rest: Discourse and Power.
This conceptual paper is about racism as social injustice, precisely it discusses the neocolonial oppression in leisure and tourism studies and the place of the scholars of color within it - to examine what else might be possible for us, collectively, in the realm of leisure and tourism studies. Tommie Shelby (2014, p. 70) in his provocative essay, Racism, Moralism, and Social Criticism emphasized an important point,
“I am not suggesting that racism is merely an ‘epiphenomenon’ that masks the ‘real’ injustice of economic exploitation or class domination. There are serious forms of injustice that are not essentially about money, property, or labor (e.g., being unfairly denied the right to vote or the right to due process) and racial ideology has played a significant role in buttressing such injustices”.
Encouraged by Shelby’s words, in this conceptual paper, I attempted to lay the groundwork necessary to render apparent this pervasive, corrosive, and dehumanizing form of domination that infects our society. Although discussion of several aspects of race now constitutes to a huge body of literature (e.g. Bonilla-Silva 1997; Bourdieu, 2003, 1988; Feagin, 2010, 1991; Cohen, 2004; Cudd, 2006; Hesselmann, 2018; Isaac, 2004; Lipsitz, 2011; McCall, 2005), regrettably, scholars have paid little importance to the covert form of institutional racism which is found in the ordinary practices of our lives (Desmond & Emirbayer, 2009). As Lillian Smith (1994, p. 96) observed in Killers of the Dream that our racist attitudes simply “slip from the conscious mind deep into the muscles”. Lives of scholars of color in academia is one such prominent contemporary issue. Recently, scholars have well documented the omnipresent racism in several disciplines (e.g. Budd & Magnuson, 2010; Buggs et al., 2020; Coleman, 2005; Harper 2012; Mueller, 2018; Rodríguez, 2018; Smith, 2012; Stanley, 2007; Thapar-Björkert & Farahani, 2019; Zuberi & Bonilla-Silva, 2008). While more has been written, this issue has yet to generate any sustained interdisciplinary critical inquiry in leisure and tourism studies as it appears that to most leisure and tourism scholars, expressions of interest in challenging “White supremacy” and “decolonizing academia” are still deeply unsettling (Bandyopadhyay, 2022). I attempted to attend to this significant lacuna in leisure and tourism studies and tried to go beyond Toni Morrison’s (1994, p. 27) stirring reminder in her magnum opus The Bluest Eye, “There is really nothing more to say - except why. But since why is difficult to handle, one must take refuge in how”.
Unfortunately, still today, talking about race in leisure and tourism studies is like breaking a taboo or as Derrida (1984) would say, “under erasure”. Hence, there remains only a handful of studies that discuss race in leisure studies (e.g. Anderson et al., 2021; Floyd 1998; Floyd & Stodolska, 2019; Fletcher et al., 2017; Higgins-Desbiolles, 2009; Mowatt, 2020, 2019; Mowatt et al., 2016; Outley et al., 2021; Pinckney et al., 2018; Ratna, 2018, 2011; Roberts, 2021; Spracklen, 2013; Torabian & Miller, 2016; Watson, 2022; Watson & Ratna, 2011; Watson & Scraton, 2001). Ditto is the case in tourism (e.g. Bandyopadhyay et al., 2022; Benjamin & Dillette, 2021; Chambers, 2020, 2018; Chambers & Buzinde, 2015; Chow-White, 2006; Higgins-Desbiolles, 2022; Hylton & Long, 2016; Jamerson, 2016; Singh, 2021). There are only a few laudable endeavours of scholars to critique tourism as an imperialistic practice (e.g. Bandyopadhyay, 2011; Brown & Hall 2008; Nash, 1977; Palmer, 1994; Sturma, 1999). Only recently (e.g. Bandyopadhyay, 2019; Bandyopadhyay & Patil, 2017), scholars have critically explored the racialized politics of tourism from the perspective of the “White savior complex” and argued that white supremacy continues to play an important role in tourism today. They claimed that colonial logics and discourses have shifted over time, from the erstwhile “civilizing mission” to the subsequent mandate for development to contemporary depoliticized social causes such as volunteer tourism to save and help the people in the global South – the main purpose of this third discourse is to resurrect imperial/colonial nostalgia. Hence, even in the twenty-first century, tourists visiting the global South still aspire to be living like a king or a queen for a day (Gottlieb, 1982) and make tourism a form of imperialism (Nash, 1977). The politics of representation play a paramount role in this as a mélange of traditionally stereotyped clichés are transferred from one generation to another so that global South-bound contemporary white tourists are primed to expect the worst of these peoples and their culture. As a result, contemporary white tourists, by integrating these expectations with their own experiences, continue to perpetuate this malicious cycle of white superiority. Indeed, “the conquest continues” (Chomsky, 1993).
Notwithstanding the praiseworthy efforts of leisure and tourism scholars to explore the issue of race, what is surprisingly missing is scholarship related to the unfortunate inner lives of scholars of color in leisure and tourism studies who arrive with grand plans and soon find themselves to be merely surviving. Very few leisure and tourism scholars have ever pondered how the “White man’s and woman’s burden” affects the scholars of color differently in the academic world of contact zones. The impacts are a matter of curiosity and gossip but unfortunately very few leisure and tourism scholars cared to follow the clues to map the contours of the racial predicament of scholars of color as a way of their lives surviving in the academia. Ironically, it is perhaps more troubling to witness the extraordinarily persistent “Whiteness” of leisure and tourism scholars who predominantly influence the field of study. And, for the Others, as Hall (1996, p. 7) reminded us, “to be racialized is to be denied entry into the mainstream of power and privilege”.
It is important to challenge the ever-existing Euro-Amero-centric benchmarks that judge the “Other” and call for a paradigm shift to emphasize that the “subaltern can speak!” (Spivak, 1988). But this is easier to “think” than to “do” in leisure and tourism studies. Perhaps, because as bell hooks (2004, p. 25) asked, “… how can we organize to challenge and change a system that cannot be named?” As a result, leisure and tourism scholars seem to shy away from more theoretical and normative discussions of what should be done to change the patterns of inequality, alienation, and anger as discussions of race is generally filled with intense and powerful emotions (Bell, 2003). In this study, I asked few pertinent questions for future scholars to ponder which will enable a critical dialogue in leisure and tourism studies. When will our attraction with the “white supremacy” come to an end, if at all? How do we want to redefine our identity in the post-Covid-19 world? This has devastating consequences as it created and continues to create leisure and tourism scholars with “Black Skin, White Masks” (Fanon, 1952). As Du Bois (1903) in his seminal The Souls of Black Folk lamented, “It is a peculiar sensation, this double-consciousness, this sense of always looking at one’s self through the eyes of others, of measuring one’s soul by the tape of a world that looks on in amused contempt and pity”. In a similar vein, Foucault (1988a) in Madness and Civilization: A History of Insanity in the Age of Reason opined, “People know what they do; frequently they know why they do what they do, but what they don’t know is what they do does”. But how can we change this? First, “we must understand how race works, developing tools to analyze this well-founded fiction responsible for so many cleavages and inequalities in our world today” (Desmond & Emirbayer, 2009, p. 335).
Hence, this conceptual study is about paranoia and ruthlessness of power and explored how to deal with the complex and shifting racial/racist landscape of contemporary leisure and tourism studies, and with how to engage across these complexities and transformations. Precisely, taking into consideration my experiences of nearly two decades in the US and UK as a scholar of color related to racism, in this conceptual study I tried to probe into the ideological factors behind the discrimination in leisure and tourism studies to pave the way for more serious consideration and empowerment of scholars of color. As such, I tried to use my experiences and feelings to track how I see the organic connections and dissonances between the three worlds – before (while growing up as a colonized man in India imitating the British), after (while pursuing doctoral studies in the US to become an academic) and eventually as an academic in the US and UK in leisure and tourism studies. In this reflexive journey, I was amazed to observe how my feelings constantly displaced, repeated and upset one another. I realized, looking back, that there was never a single moment in this trajectory which was not provoked by my racial positioning – though in the US and UK, the contours were perpetually unpredictable, but the intervening significance of the fact was persistent. My skin color was incontestably an issue. I understand that this study will certainly not set the Thames on fire but I seek to open new avenues of discussion to break this silence to improve understanding so that the policies, practices, and ideas that disseminate racial inequality can be acknowledged and dismantled. While doing that, I try to follow the philosophy of Hegel’s “master/slave dialectic: the search for self-consciousness” within the Bakhtinian (multiaccentuality of racial meaning) and Levinasian (his close equivalence between structuralist anthropology and genetics) context, and go beyond, equipped with the wisdom of Stuart Hall, Frantz Fanon, W.E.B. Du Bois, Toni Morrison, Michel Foucault, Karl Marx, Jacques Derrida and Amartya Sen. As Lawrie Balfour (2011, p. 415) movingly pointed out, “For the greatest legacy of the questions posed in Du Bois’s work and by the figure of Du Bois himself is to remind us just how intricate and complex ‘our tasks of emancipation’ remain”. It is important to note here, in this conceptual paper, I narrate my disjointed self but in no way seek to salve feelings of inferiority. Nor does my arguments try to act as propaganda against “White academics” but instead disparages the ideology of “White logic” (Zuberi & Bonilla-Silva, 2008) which “essentially renders Whiteness meritocratic and other colors deficient” (Fine et al., 1997, p. 64). My experiences narrated in this paper are fairly typical of the experiences of many scholars of color who prefer to remain silent. This acceptance of destiny by scholars of color is suggestive of form of “symbolic domination” (Bourdieu, 1977). However, I think, if we do not challenge these inequities then we are disseminating the ideology of White logic. As Szasz (1973, p. 20) reminded us, “In the animal kingdom, the rule is, eat or be eaten, in the human kingdom, define or be defined”.
Theoretical Foundation - Critical Race Theory, Whiteness and Social Justice
The theoretical background of this study is derived from the Harvard Economist and Nobel Laureate Amartya Sen’s fifty years of seminal work on “social justice”. One of the principal differences between Sen and the leading contemporary theorists of justice is that they have been concerned primarily with identifying what perfectly just social arrangements might be rather than clarifying how different realizations of justice might be compared and evaluated. The cornerstone of Sen’s argument is his insistence on the role of public reason in establishing what can make societies less unjust. Sen (2009, p. 127), in his The Idea of Justice, powerfully asks,
“The basic idea of human rights, which people are supposed to have simply because they are human, is seen by many critics as entirely without any kind of a reasoned foundation. The questions that are recurrently asked are: do these rights exist? Where do they come from?”.
“In the little world in which children have their existence”, says Pip in Charles Dickens’s Great Expectations, “there is nothing so finely perceived and finely felt, as injustice”. In a similar vein, Sen (2009, p. 9) argues that “strong perception of manifest injustice applies to adults as well and what moves us is not the realization that the world falls short of being completely just but that there are clearly remediable injustices around us which we want to eliminate”. It is in this context, this conceptual study argues that though ‘racism’ is a dirty word, significant inequalities persist in leisure and tourism studies today, which “we want to eliminate” - borrowing Sen’s words.
Commenting on the perpetual injustice to the topic of “race”, Mills (2014, p. 35) lamented that the prominent scholars in philosophy, while discussing John Rawls’ (1971, p. 27) ideal theory,
“have either no discussions at all of race, racism, and affirmative action, or at best a sentence or a paragraph or two. Nor do they indicate that this might be a problem, or comment anywhere on the absurdity of the most famous twentieth-century theorist of justice of a former White settler state having nothing useful to say about race—the central injustice on which that state rests. So, the simple fact that racial justice has not been central to the discussions of justice in American political philosophy over the last forty years is itself a clear-cut testimony to its ‘Whiteness.’ Nevertheless, it can obviously be replied that this lacuna, embarrassing as it may be (though it doesn’t seem to be), is still only a contingent one, unrelated to the apparatus”.
Intellectual History of Racism
White supremacy has been an ongoing racial project for almost 500 years (Mills, 2014, 1997). According to Wilson (1999, p. 14), racism is “an ideology of racial domination” in which the presumed biological or cultural superiority of the whites is emphasized to justify the inferior treatment of the Others. From a sociological perspective, Clair & Denis (2015, p. 857) clarified, “Through the process of racialization, perceived patterns of physical difference – such as skin color or eye shape – are used to differentiate groups of people, thereby constituting them as ‘races’; racialization becomes racism when it involves the hierarchical and socially consequential valuation of racial groups”. And from a psychological perspective, Salter et al (2017, p. 150) explained, “The term racism is often used synonymously with prejudice (biased feelings or affect), stereotyping (biased thoughts and beliefs, flawed generalizations), discrimination (differential treatment or the absence of equal treatment), and bigotry (intolerance or hatred). This practice implicitly conceptualizes racism as a set of basic social-psychological processes underlying the psychologies of individuals (i.e., stereotyping, prejudice, and discrimination) merely applied to the context of race”.
In the late nineteenth century, although sociology emerged as a social scientific discipline, few scholars studied racism. W.E.B. Du Bois was the only exception who analyzed the political economic roots of racism and its perverse impacts on western institutions and psyches. During this time, racism pervaded society. Beginning in the 1920s, the scientific validity of race came under closer scrutiny and sociologists at the Chicago School began to view racism as a distinct social problem worthy of study. The 1950 and 1960 s saw theories arose to explain why racism, racial discrimination and racial inequality persisted (Bobo, 2011). In the 1980 and 1990 s, various theories of new racisms emerged suggesting that racism itself has transformed into more covert forms.
Contemporary theorists suggest that new forms of racism that are expressed not in avowed racist attitudes but rather in contextually specific moral and symbolic principles that stereotype subordinated racial groups as underserving and thereby justify existing racial inequalities (Kinder & Sears, 1981). For example, surveys repeatedly show that many whites support racial equality in principle but resist policies to implement it (e.g. affirmative action and reparations) which caters to a “political agenda” (van den Berghe, 2001, p. 12,721). Another explanation for insistent racial inequality is “implicit bias” which is an unconsciously triggered belief in the inferiority of, or negative attitude toward, a group and can impact expectations and actions unconscious negative beliefs and feelings about racial groups may not appear on a survey but may be revealed in everyday interpersonal interactions at work, at school, or on the street (Clair & Denis, 2015). As Stuart Hall (2000b, p. 149) argued, “Black is not a question of pigmentation… [It] is a historical category, a political category, a cultural category”. As such, “black” needs to be understood in a particular time and space, as changing and contested, and as constructed and situational rather than taken for granted – as “without guarantees” (Hall, 1992). The story is similar with the case of scholars of color in leisure and tourism studies.
Racism as a Social Process
Discourse from anthropology, history and sociology characterizes the concept “race” as having a modern history. According to the Booker Prize winning author Arundhati Roy (2001, p. 81), “[r]ace was created mainly by Anglo-Europeans, especially English, societies in the 16th and 19th centuries”. In spite of several centuries of use as a concept representing a natural phenomenon, sociological studies on race critique the notion as lacking scientific clarity and specificity. Rather than emerging from a scientific perspective, the notion of race is informed by historical, social, cultural, and political values. Thus, we find that the concept of race is based on socially constructed, but socially, and certainly scientifically, outmoded beliefs about the inherent superiority of the whites (they are rational and brilliant, thus have a bright future) and inferiority of Others (they not only lack intelligence, rigor and relevance but are also considered inhuman or nonhuman so they cannot enjoy human rights, civil rights, labor rights, etc.) based on racial distinctions (Garbe, 2013; Grosfoguel, 2007; Hesselmann, 2018). British political theorist Lord Bhikhu Parekh (1994) explained this contradiction between superior and inferior people extensively in his article, Superior People: The Narrowness of Liberalism from Mill to Rawls. Bonilla-Silva’s (2010, 1997) conception of “racialized social systems” is significant in this context which highlights how political, economic, and social provisions are structured by white racial hierarchy (set of frames, styles, and scripts that are used to explain and justify the racial status quo without sounding racist) and supported by “colorblind racism”. This ideology can be traced back to sixteenth century slavery and imperialism, as Feagin (2010) argued, while studying the societal norms and “white racial frames”. Discussion of the social construction of “Whiteness” cannot be complete unless we acknowledge the social and political significance of race in America. Hacker (1993) in his groundbreaking study, Two Nations: Black and White, Separate, Hostile, Unequal explained that whiteness is a “condition like a virus of the psyche”, which makes whites react to blacks as superiors to inferiors. Moreover, as Hacker emphasized, whiteness perceives blackness as a “stain like a drop of ink in water”. He further claimed that whiteness is a delusional state of mind that was used to justify slavery and that endures to this day to validate white privilege. Hacker (1993, p. 4) rued,
“America is inherently a ‘white’ country: in character, in structure, in culture. Needless to say, black Americans create lives of their own. Yet as a people, they face boundaries and constrictions set by the white majority America’s version of apartheid, while lacking overt legal sanction, comes closest to the system even now… reformed in the land of its invention”.
Whatever its scientific validity, race is a social fact in which the social and political significance of “Whiteness” plays a critical role. Classical scholars have remarked about race as a social fact. Thus, according to Durkheim ([1895] 1938, p. 13), the concepts, race and whiteness, are social facts.
“A social fact is every way of acting, fixed or not, capable of exercising on the individual an external constraint; or again, every way of acting which is general throughout a given society, while at the same time existing in its own right independent of its individual manifestations”.
Statement of Positionality
“… one simply cannot and will never be able to fully recuperate one’s own processes of thought or creativity self-reflexively… I cannot become identical with myself”.
(Stuart Hall, 2017) – Through the Prism of an Intellectual Life.
I, the author of this study is a native of India. Being the only son in an upper middle-class Indian family, life was a rosy picture to me. I was sent to an elite English Convent School in India, where my teachers in their white robes were Anglo-Saxons. Being a Hindu by birth, I used to sing, “Jesus is my Lord” every morning in my school prayers. I was meticulously groomed in school to behave like an English boy in my mannerisms. My academic life has been a roller-coaster ride. After my schooling and undergraduate degree in India, I went to pursue my MBA degree in Europe. Then, I undertook my Doctoral studies in Leisure Studies and Socio-cultural Anthropology at a renowned university in the US. It was during this time, in the process of my metamorphosis, that I discovered myself anew. I began my academic life teaching at a Russell Group member university in the UK for a year followed by a decade long career at a public university in California. My wanderlust then took me to a public university in Thailand and finally am enjoying my sojourn at a private university in Viet Nam. During these 20 years of my life abroad, as a student and academic, I got the chance to travel globally (to 60 countries) which had impacted my research tremendously.
I serve as a reviewer for several leading Tourism journals and also important journals in Anthropology, Cultural Studies, Politics, Religious Studies, Ethnic & Racial Studies, History and Geography. I believe, as a mid-career scholar with 20 years of experience in leisure and tourism studies allows me to provide a deeper sense on the topic of this paper. Furthermore, my dual roles as both author and reviewer add to the etic and emic perspectives. Finally, it is important to recognize the dynamics of power relations - I am from the global South and educated in India, Europe and the US and taught in the UK, US, Thailand and Vietnam.
As mentioned above, growing up in India mimicking the British, I found myself in a conflicting situation after arriving in the west. The crisis of my identity occurred, which was manifested in the indecision over what to accept from western practices and what to abandon of Indian practices – it became a crusade! As Salman Rushdie (1983, p. 81) clarified:
“What does it mean to be an ‘Indian’ outside India? How can culture be preserved without becoming ossified? How should we discuss the need for change within ourselves and our community… what are the consequences, both spiritual and practical, of refusing to make concessions to Western ideas and practices? What are the consequences of embracing those ideas and the practices and turning away from ones that came with us?”.
I remember often reacting with disproportionate rage to my colleagues in the UK and US, who said they did not understand why I kept banging on about being colored since I was from an upper middle-class family, had been educated at an elite English Convent School and studied abroad in Europe and the US and taught in the UK and US. As if to say, what had I to complain about? Since these colleagues in the UK and US had unconsciously no doubt identified the central contradiction of my life, I felt, perhaps unfairly, that those who did not understand that was not likely to get much else about me right. It is this history which lay behind my decision to set about writing this conceptual paper.
In an effort to free myself from the baggage of a colonized man from the metropole (in India), I never had any aspiration to be British or American nor have I ever become such. I have always used my original surname and discarded the one given by the British (as they could not pronounce certain long Indian surnames, they renamed those). This reminds of Stuart Hall’s (2017, p. 14) poignant description, “Being English, it seemed to me, was not a repository of potential identification – rather it was just an unwelcome twist of historical fate. It had no traction on my actual life”. This is quite similar to what Edward Said described in his memoir about his struggles with his name – his fraught relationship with this unknowable, awkward persona - “Edward”, the other inside him who caused him much grief (Kennedy, 2000). In the last few years, I always wanted to write this conceptual paper and hoped that it might constitute an insight into the contradictory transition points in my old story – the long, mimicked and never-concluded route out of colonial subaltern hood. Politics has always been a passion, since those undergraduate college days in Calcutta in India (the city had the world’s longest ruling communist government – continuously for 34 years) when Marx, Lenin, Ho Chih Minh and Che Guevara were my idols – “rebellion”, “social justice” were cult words then (and still continues to be) - I dreamt of changing the world, for good. Now it is time to reminisce – how, like all students in our school, I also believed that one day, I must study and live in the UK or US. My feelings were same in college and thought that I could escape the lanes of my memories in India, by leaping onto the UK or US – another future, a new dream. All that running, when there was no way to escape. I feel trapped. “How are we to live in the world?” asks Salman Rushdie. Wish I knew!
Suffering in Life – the Common Fate of Scholars of Color
“Rise, people, rise up now, break the chains of caste
Throw off the corpse of slavery, smash the obstacles, Rise people –
We may be Maratha, Mahar, Brahman, Hindu, Muslim, Christian, Humanity is all one, all are brothers”.
(Joshi, 1986, p. 97).
The protest song mentioned above is of the oppressed group Dalit (untouchables) in India. This song reminds of Gayatri Spivak’s (1987) notion of “strategic essentialism” which can allow for appeals to humanism in the political interest of oppressed groups – in this context, I, the author of this paper, my relationship between power and powerlessness and the endless conflict I am engaged in, where I now turn to.
Nobel Laureate poet Rabindranath Tagore (2009, p. 87) opined, “Whenever the people in the west ask us for clarifications, they do that with a superior air. It is a sign of laziness and impotency to accept conditions imposed upon us by others who have other ideals than ours”. Similarly, while studying and teaching leisure and tourism studies at universities in the US and UK, whenever I introduced the topic of race with my colleagues, I was mocked. Below mentioned are some anecdotes during my tenure as a young lecturer at a British university.“Why do you care so much about race? It doesn’t exist. If it did, then we wouldn’t have hired you in the UK. You are teaching at a Russell Group British university being an Indian and you should be very proud of that fact. Rather than feeling that you are being discriminated, you should be happy that this experience will define your life always.”
“Your accent is so funny. You went to a British School in India, right? That’s why you are so British… your mannerisms are almost there… you just need to keep trying… Indians are smart and very good in mimicking.”
As a young faculty member, I started my career in the academia at a prestigious British university, where apart from some great experiences, I also learned my limitations for being a man of color. Every now and then, I was reminded of what was appropriate and what was not. I was overwhelmed by the emotional weight of my British mannerisms which I had to mimic all the time to please my colleagues. I was always under scrutiny. How true Stuart Hall’s (1992a, p. 17) acute observation was, “We’d all undertaken the journey to our many illusions. Embarrassingly, I found myself in tears, often”. Me too. I myself dreamt of this journey and there I were… embarked on a journey where “difference” meant everything, it could not be evaded. Standing at the crisis between an India where I did not know how to belong and an England which I realized I did not belong. From my colleagues’ disappointments with the smell of bananas and rotten tomatoes every day at my office to sophisticated Britons asking their dogs to poo in front of my house to being denied of several deserved privileges; not to mention the omnipresent mockery which became an integral part of my life. Hacker (1993, p. 29), while arguing the value of whiteness, reiterated that it “sets a floor on how far people of that complexion can fall. No matter how degraded their lives, white people are still allowed to believe that they possess the blood, the genes, the patrimony of superiority. No matter what happens, they can never become black”. To support his argument, Hacker explained the reason why there is no word in English that overwhelms for whites that “nigger” does for blacks. “White trash? No, that won’t do it. Honky? That won’t do it”, Hacker justified. This reminds of Mill’s (2014) poignant remark that the colonialism that is still with us is energized through signs, metaphors and nasty jokes. In a white dominant environment, I felt helpless, and obviously Fanon’s (1952, p. 111) reflection was the only solace:
“So, they were countering my irrationality with rationality, my rationality with the ‘true rationality.’ I couldn’t hope to win. I tested my heredity. I did a complete checkup of my sickness. I wanted to be typically black – that was out of the question. I wanted to be white – that was a joke. And when I tried to claim my negritude intellectually as a concept, they snatched it away from me”.
Fanon wrote this more than sixty years ago and it is still so relevant. This proves that Orientalism, even forty years after Said (1979) wrote his book, is still the same as it ever was.
I thought my life in the academia would be different in the US, hence I returned to the land of opportunities and this time to California – “the land of the free”. Soon I realized that my journey to an illusion was a fraught transition – I thought I was going forward however was confronted by the return of the suppressed. Quite similar to the UK, majority of my colleagues in the US were in denial of the fact that racism existed in leisure and tourism studies. Below are some excerpts.
“Come on, ‘racism’ is an obsolete word. This is California! We are the most progressive among all Americans… we love everyone! By the way, how do I pronounce your surname, why do you use this bloody surname and not what the British have given you – that sounds simple and sophisticated as well”.
“We teach in a beautiful discipline – ‘leisure and tourism’. People from so many countries study in our department. We are truly a melting pot, like America. We are like various flowers in one garden. No racism, no discrimination, united we stand”.
Every time, a colleague of mine said something like the above, were supported by others unanimously, leaving me, the only man with a brown, dark (or whatever) skin, alone - with no chance to debate. Indeed, “no dialectic was warranted or required” from me (Said, 1979). The camaraderie of my white British and American colleagues and their prophesy about leisure and tourism studies reminded me of Benedict Anderson’s (1991, p. 7) remarkable work Imagined Communities, “regardless of the actual inequality and exploitation that may prevail in each, the nation [here, leisure and tourism studies] is always conceived as a deep, horizontal comradeship” (emphasis mine). This is exactly what DiAngelo (2018, p. 27) articulated in her thought-provoking, White Fragility: Why It’s So Hard for White People to Talk About Racism,
“White supremacy is everywhere. Messages of pre-eminent white value and Black insignificance are raining down on us 24/7, and there are no umbrellas My psychosocial development was inculcated in this water and internalized white superiority is seeping out of my pores”.
The above romantic statements made little sense as in the American university (in California – “the land of the free”) where I was working then, three white students there were charged by authorities with hate crimes in the tormenting of a black roommate that allegedly went on for months before anyone intervened – the white students calling the black student “3/5” in a reference to the way the infamous Three-Fifths Compromise and when the black student objected, the white students called him “Fraction” (Jaschik, 2013). Ironically, the Olympic Black Power Statue of Tommie Smith and John Carlos with their famous gesture still stands mute in the university campus as a testimony to racial discrimination. Both in the UK and US, according to my colleagues, race was something that did not exist anymore, which supports Ladson-Billings (2012) opinion that the whites avoid the existence of race in daily lives. And this has been happening well over a century, as Yale historian Mercer (2017, p. 2) pointed out, the “troubling ambiguities in the concept of race that have vexed eminent thinkers”. In a similar vein, Stuart Hall (2017) in his lecture at Harvard, “Race – The Sliding Signifier” explained why the concept of race stubbornly persists despite every explanation showing its realities to be socio-historical and not biological. We are shattered and destroyed by the racializing White gaze, as so movingly portrayed by Fanon, “Look, a Negro!… Mama, see the Negro. I’m frightened”. I still remember an incident during my first visit to the UK for my job interview when I requested a British lady to take my photo with the university in the background. She obliged but her little boy screamed, “Don’t take the photo of the savage!” So, my experiences in the UK and US made me relate to Fanon’s (1952, p. 37) intense observation,
“… fixed, woven out of a thousand details, anecdotes, stories… sealed into that crushing objecthood… the movements, the attitudes, the glances of the other fixed me there, in the sense in which a chemical solution is fixed by a dye… now, the fragments have been put back together again by another self… a slave, not of the ‘idea’ that others have of me but of my own appearance”.
This fixing of me in the White mask as a result of my look is what Fanon (1952) articulately calls the process of epidermalization. More interestingly, whenever I attempted to introduce a radical thought, I was indirectly or sometimes directly, reminded of my origin that is originally from where I belong, emphasizing the brutal reality – how did I dare to eschew inherited truths? Hall (1992b) vociferously showed how old hierarchies of human identity in Western culture were forcefully broken apart when oppressed groups introduced new meanings to the representation of difference. Appiah (1986) in his radical, The Uncompleted Argument, commented on the scholars in the academy who have been too reluctant to share the “repudiation of race as a term of difference”.
Brunsma et al., (2012, p. 727) building on the ideas of Fanon (1966), claimed that, “theoretically and epistemologically, racism has not only colonized the lands, histories, cultures, bodies, and cognitions of non-whites but importantly, it has colonized the histories, cultures, bodies, and cognitions of whites as well”. This has had crushing significances, as DuBois (1903) indicated, “not only for non-Whites’ lived experience and self-determination but also for Whites’ ability to see beyond the veil of race”. As a result, we see the perpetuation of neo-colonial mentality (“tutelage”). Baldwin (1965, p. 7) decades ago discoursed:
“People who imagine that history flatters them (as it does, indeed, since they wrote it) are impaled on their history like a butterfly on a pin and become.
incapable of seeing or changing themselves or the world. This is the place in which it seems to me, most white Americans find themselves. Impaled. They are dimly, or vividly, aware that the history they have fed themselves is mainly a lie, but they do not know how to release themselves from it, and they suffer enormously from the resulting personal incoherence”.
Conclusion
“It is the racist who creates his inferior” (Fanon, 1952, p. 69).
In this conceptual paper, I have tried to articulate that in leisure and tourism studies “we still live in a wholly racialized world” (Morrison, 1992) where “privileged whites are supposed to [and in reality] protect and perpetuate not only white myths but also raciology itself” (Gilroy, 2002). Narrating my experiences as an academic in this “white world”, in this paper, I have expressed how my everlasting quest has always been to feel at home without becoming like the white culture. The dilemmas and rejections in this journey created an omnipresent tension in my life which shaped the content of this paper. My feelings can be best described through the poignant words of Toni Morrison (1994, p. 17), “A little black girl yearns for the blue eyes of a little white girl, and the horror at the heart of her yearning is exceeded only by the evil of fulfillment”.
William Shakespeare (1603, I, iv) in his play Measure for Measure cautions us, “Our doubts are traitors, and make us lose the good we oft might win, by fearing to attempt”. So, this conceptual paper is an attempt to consider reversing the gaze not for turning the tables rather give importance to scholars of color. It is important to note here, quoting Moisi (2009, p. 56), if “humiliation is impotence, hope is confidence” and I have benefited tremendously from all the humiliation in my academic life which gave fruition to this study. I suggest that leisure and tourism scholarship must continue to work within the creative space between a utopian idealism and the pragmatic requirements of politics at every level. As Spivak (1987) states:
“[I]f we engage ourselves not only in the end of exploitation of our own community, but for the distant and impossible but necessary horizon of the end of exploitation, then we will not be confined within fantasmic and divisive cultural boundaries”.
Precisely, to what extent should the postcolonial leisure and tourism studies agenda be aligned with the kind of anti-representationalism characteristic of much post-structuralist and postmodernist thought? As Nicholas Dirks (1992, p.12) asks us:“What does it mean that Edward Said, or Ranajit Guha and the Subaltern Studies collective of Indian historians, take the very same texts by Gramsci, Foucault, or Williams as fundamental that are recited elsewhere in the academy We ignore at our peril the manifestations of the postcolonial predicament in provincial universities in Asia and Africa where these theorists would all signify elitist forms of exclusion, new Western forms of domination”.
Let us believe in the transformation of leisure and tourism studies, where “the concept of race will come to be widely viewed as incoherent and empirically unsound” (Appiah, 1986, p. 57). From Marx, Sartre, Fanon, CLR James to Du Bois – all advocated “Freedom for All” based on mutual respect. Quite similar to these radical thinkers’ humanist and universalist perspectives, let’s imagine a world where, as philosopher Ludwig Wittgenstein (1963, p. 27) so eloquently put it, “All that one knows could be otherwise. All that one sees could be otherwise?” Indeed, let us imagine a new world of leisure and tourism studies, post-Covid-19, where all kinds of racism will disappear. Perhaps, “disappear” is too strong a word, so Stuart Hall’s (2017) perceptive vision can help us to understand today’s crisis of liberal democracy in leisure and tourism studies and provide us the much-needed hope:
“A different, postcolonial understanding of multiculturalism that both acknowledged and celebrated the hybrid and mongrelized nature of cultures that slavey and colonialism had both produced and displaced. Colonial history ensured that it was no longer possible to conceive of specific communities or traditions whose boundaries and identities were settled and fixed”.
Electronic Supplementary Material
The authors declare that they have no conflict of interest
Supplementary Material 1
Declaration
Conflict of interest statement
Not applicable.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
Anderson B Imagined Communities: Reflections on the Origins and Spread of Nationalism 1983 London Verso
Appiah, A. (1986). The Uncompleted Argument: Du Bois and the Illusion of Race. In Race, Writing and Difference. Edited by Henry Louis Gates, Jr., 21–37. Chicago: University of Chicago Press.
Bakhtin, M. M. (1990). Art and Answerability: Early Philosophical Essays by M. M. Bakhtin edited by Michael Holquist and Vadim Liapunov. Austin: University of Texas Press.
Bandyopadhyay, R. (2011). A Photo Ethnography of Tourism as Neo-colonialism. Annals of Tourism Research, 38(2): 714–718.
Bandyopadhyay, R. & Patil, V. (2017). “The White Woman’s Burden”: The Racialized, Gendered Politics of Volunteer Tourism. Tourism Geographies, 19(4): 644–57.
Bandyopadhyay, R. (2019). Volunteer Tourism and “The White Man’s Burden”: Globalization of Suffering, White Savior Complex, Religion and Modernity. Journal of Sustainable Tourism, 27(3): 327–343.
Bandyopadhyay, R. (2022). “A Wholly Racialized World”: Racial Inequalities and Peer Review in Leisure and Tourism Studies. Leisure Studies, 41(5): 605–619.
Bandyopadhyay, R., Ray, A. & Milde, M. (2022). A Phenomenological Analysis of Tourist Identity: Three Theses and Propositions. Leisure Sciences. 10.1080/01490400.2022.2138649
Bell, L. A. (2003). Telling tales: What stories can teach us about racism. Race, Ethnicity and Education, 6, 3–28.
Benjamin S Dillette AK Black Travel Movement: Systemic racism informing tourism Annals of Tourism Research 2021 88 103169 10.1016/j.annals.2021.103169
van den Berghe, P. L. (2001). Sociology of Racism. International Encyclopedia of Social and Behavioral Sciences. Elsevier.
Bernal DD Villalpando O An Apartheid of Knowledge in Academia: The Struggle over the “Legitimate” Knowledge of Faculty of Color Equity & Excellence in Education 2002 35 2 169 180 10.1080/713845282
Bhabha, H. (2004). Framing Fanon. Introduction to Frantz Fanon. The Wretched of the Earth vii-xli. New York: Grove Press.
Bhabha H The Location of Culture 1994 London Routledge
Bobo, L. (2011). Somewhere Between Jim Crow and Post-Racialism: Reflections on the Racial Divide in America Today. Daedalus, 140(2), 11–36.
Bonilla-Silva, E. (2010). Racism without Racists. Color-Blind Racism. Racial Inequality in Contemporary America. (3rd ed.). New York: Rowman & Littlefield.
Bonilla-Silva, E. (1997). Rethinking Racism: Toward a Structural Interpretation. American Sociological Review, 62, 465–480.
Bourdieu P Outline of a theory of practice 1977 Cambridge Cambridge University Press
Bourdieu P Homo Academicus 1988 Stanford Stanford University Press
Bourdieu, P. (2003). Language and Symbolic Power. Translated by Gino Raymond and Matthew Adamson. Cambridge, MA: Harvard University Press.
Brown, F., & Hall, D. (2008). Tourism and development in the global south: The issues. Third World Quarterly, 29(5),839–849.
Brunsma DL Brown ES Placier P Teaching Race at Historically White Colleges and Universities: Identifying and Dismantling the Walls of Whiteness Critical Sociology 2012 39 5 717 738 10.1177/0896920512446759
Budd JM Magnuson L Higher education literature revisited: Citation patterns examined Research in Higher Education 2010 51 3 294 304 10.1007/s11162-009-9155-6
Buggs, S. G., Sims, J. P., & Kramer, R. (2020). Rejecting white distraction:a critique of the white logic and white methods in academic publishing. Ethnic and Racial Studies, 1–9.
Chambers, D. (2020). Reflections on the relationship between gender and race in tourism In Tourism in Development: reflective essays, U. C. Peter, & Dieke (Eds.), Brian E.M. King Richard, A. J., & Sharpley, CAB International, Wallingford, pp.233–244.
Chambers, D. (2018). Tourism research: Beyond the imitation game. Tourism Management Perspectives, 25(1), 193–195.
Chambers, D., & Buzinde, C. (2015). Tourism and Decolonisation:Locating Research and Self. Annals of Tourism Research, 51,1–16.
Chow-White, P. A. (2006). Race, gender and sex on the net: semantic networks of selling and storytelling sex tourism. Media, Culture & Society, 28(6), 883–905.
Clair, M., & Denis, J. S. (2015). Sociology of Racism. In J.D. Wright (Ed.), The International Encyclopedia of the Social and Behavioral Sciences, 19, 857–863.
Cudd AE Analyzing Oppression 2006 New York Oxford University Press
Derrida, J. (1984). Difference. In Margins of Philosophy. Translated by Alan Bass. Chicago: University of Chicago Press.
Desmond M Emirbayer M What is Racial Domination? Du Bois Review 2009 6 2 335 355 10.1017/S1742058X09990166
Du Bois, W. E. B. (1903). The Souls of Black Folk. Dover Publications.
Durkheim, E. ([1895]1938). The Rules of the Sociological Method, eighth edition, translated by Sarah A. Solvay and John H. Mueller. Edited by George E.G. Catlin. The Free Press.
Fanon F The Wretched of the Earth 1966 London Penguin Books
Fanon F Black Skin, White Mask 1952 New York Grove Press
Feagin, J. R. (1991). The Continuing Significance of Race: Antiblack Discrimination in Public Places. American Sociological Review, 56,101–116.
Feagin, J. R. (2010). The White Racial Frame: Centuries of Racial Framing and Counter Framing. New York: Routledge.
Fine, M., Weis, L., Powell, L. C., & Wong, L. M. (1997). Off White: Readings on race, power and society. New York: Routledge.
Fletcher, T., Carnicelli, S., Lawrence, S., & Snape, R. (2017). Reclaiming the ‘L’ word: Leisure Studies and UK Higher Education in neoliberal times. Leisure Studies, 36(2), 293–304.
Floyd, M. F. (1998). Getting Beyond Marginality and Ethnicity: The Challenge for Race and Ethnic Studies in Leisure Research. Journal of Leisure Research, 30(1), 3–22.
Floyd, M., & Stodolska, M. (2019). Scholarship on Race and Ethnicity: Assessing Contributions to Leisure Theory and Practice. The Journal of Park and Recreation Administration, 37(1),80–94.
Foucault M Madness and Civilization: A History of Insanity in the Age of Reason 1988 New York Random House USA Inc
Foucault M Power/Knowledge. Selected Interviews & Other Writings 1972–1977 1988 New York Random House USA Inc
Garbe S Deskolonisierung des Wissens: Zur Kritik der epistemischen Gewalt in derKultur- und Sozialanthropologie Austrian Studies in Social Anthropology 2013 1 1 17
Gilroy P Against Race: Imagining Political Culture beyond the Color Line 2002 MA, USA Harvard University Press
Grosfoguel R The epistemic decolonial turn Cultural Studies 2007 21 2–3 211 223 10.1080/09502380601162514
Hacker, A. (1993). Two Nations: Black and White, Separate, Hostile, Unequal. New York: Charles Scribner’s Sons.
Hall S Familiar Stranger 2017 A Life Between Two Islands. Durham, NC Duke University
Hall, S. (1996). Ethnicities. In D. Morley & K. Chen (Eds.), Stuart Hall: Critical Dialogues in Cultural Studies. London: Routledge.
Hall S Race, Culture, and Communications: Looking Backward and Forward at Cultural Studies Rethinking Marxism 1992 5 1 10 18 10.1080/08935699208657998
Hall, S. (1992b). The West and the Rest: Discourse and Power. In Formations of Modernity, edited by Stuart Hall and Bram Gieben, 276–320. Cambridge: Polity.
Harper, S. R. (2012). Race without Racism: How Higher Education Researchers Minimize Racist Institutional Norms. The Review of Higher Education, 36(1),9–29..
Hesselmann, F. (2018). Science and its Others: examining the discourse about scientific misconduct through a postcolonial lens. Identities: Global Studies in Culture and Power, 26(4), 393–411.
hooks We real cool: Black men and masculinity 2004 New York and London Routledge
Hylton, K., & Long, J. (2016). Confronting ‘race’ and policy. ‘how can you research something you say does not exist?’ Journal of Policy Research in Tourism, Leisure and Events, 8(2),202–208.
Isaac B The Invention of Racism in Classical Antiquity 2004 Princeton, NJ Princeton
Jamerson, T. (2016). Race and critical tourism studies: An analytical literature review. Sociology Compass 10(11): 1038–1045.
Joshi, B. (1986). Untouchable! Voices of the Dalit Liberation Movement (Women in the Third World). London: Zed Books.
Kennedy V Edward Said: A Critical Introduction 2000 London Polity
Kinder, D. R., & Sears, D. O. (1981). Prejudice and politics:Symbolic racism versus racial threats to the good life. Journal of Personality and Social Psychology, 40, 414–431.
Ladson-Billings, G. (2012). Through a Glass Darkly. The Persistence of Race in Education Research & Scholarship. Educational Researcher, 41(4), 115–120..
Lévinas E Dieu, la Mort et le Temps 1993 Paris Bernard Grasset
Lipsitz G How Racism Takes Place 2011 Philadelphia, PA Temple University Press
Marx K Capital: A Critique of Political Economy 1967 New York International Publishers
McCall, L. (2005). The Complexity of Intersectionality.Signs: Journal of Women in Culture and Society, 30, 1771–1800.
Mills CW White Time Du Bois Review: Social Science Research on Race 2014 11 01 27 42 10.1017/S1742058X14000022
Mills CW The Racial Contract 1997 Ithaca, NY Cornell University Press
Moisi D The geopolitics of emotion: How cultures of fear, humiliation, and hope are reshaping the world 2009 New York Random House
Morrison T Playing in the Dark: Whiteness and the Literary Imagination 1992 USA Harvard University Press
Morrison T The Bluest Eye 1992 London Penguin
Mowatt, R. A. (2019). A People’s History of Leisure Studies: Where the White Nationalists Are.Leisure Studies, 1–18.
Mowatt RA Johnson CW Roberts NS Kivel BD “Embarrassingly White”: Faculty racial disparities in American recreation, park & tourism programs Schole: A Journal of Leisure Studies & Recreation Education 2016 31 1 37 55 10.18666/SCHOLE-2016-V31-I1-7268
Mueller JC Advancing a Sociology of Ignorance in the Study of Racism and Racial non-Knowing Sociology Compass 2018 12 8 1 22 10.1111/soc4.12600
Nash D Smith VL Tourism as a form of imperialism Hosts and guests. The anthropology of tourism 1977 Philadelphia, PA University of Pennsylvania Press 33 47
Outley C Bowen S Pinckney H Laughing While Black: Resistance, Coping and the Use of Humor as a Pandemic Pastime among Blacks Leisure Sciences 2021 43 1–2 305 314 10.1080/01490400.2020.1774449
Palmer C Tourism and colonialism Annals of Tourism Research 1994 21 4 792 811 10.1016/0160-7383(94)90084-1
Parekh B Superior People: the narrowness of liberalism from Mill to Rawls Times Literary Supplement 1994 February 25 11 13
Pinckney HP Mowatt RA Outley C Brown A Floyd MF Black KL Black Spaces/White Spaces: Black Lives, Leisure, and Life Politics Leisure Sciences 2018 40 4 267 287 10.1080/01490400.2018.1454361
Ratna A Not just merely different: Travelling Theories, post-feminism and the racialized politics of women of color Sociology of Sport Journal 2018 35 3 197 206 10.1123/ssj.2017-0192
Ratna A ‘Who wants to make aloo gobi when you can bend it like Beckham?’: British Asian females and their racialised experiences of gender and identity in women’s football Soccer & Society 2011 12 3 382 401 10.1080/14660970.2011.568105
Rawls J A Theory of Justice 1971 Cambridge Harvard University Press
Rawls J Political Liberalism 1993 New York Columbia University Press
Rawls J Justice as Fairness: A Restatement 2001 Cambridge Harvard University Press
Roberts NS “White Savior” as duly provocative or enforced civility with unintended consequences? A response to Anderson, Knee, & Mowatt Journal of Leisure Research 2021 52 5 551 556 10.1080/00222216.2020.1854547
Rodríguez CO Decolonizing Academia: Poverty, Oppression and Pain 2018 Winnipeg Fernwood Publishing
Roy A Ordinary Person’s Guide to Empire 2005 New Delhi Penguin Books India
Said E Orientalism 1978 New York Pantheon
Salter PS Adams G Perez MJ Racism in the Structure of Everyday Worlds: A Cultural–Psychological Perspective Current Directions in Psychological Science 2018 27 3 150 155 10.1177/0963721417724239
Sen A The Idea of Justice 2009 Cambridge Harvard University Press
Sen A Identity and Violence: The Illusion of Destiny 2006 New York Norton & Company
Sene-Harper, A., Mowatt, R. A., & Floyd, M. F. (2002).A People’s Future of Leisure Studies Political Cultural Black Outdoors Experiences. Journal of Park and Recreation Administration, 40(1).
Shelby T Racism, Moralism, and Social Criticism Du Bois Review 2014 11 1 57 74 10.1017/S1742058X14000010
Singh RB A Critical Yoga Studies Approach to Grappling with Race: Introducing “Racial Tourism,” “Racial Mobilities,” and “Justice Storytelling” in the Context of Racial Fraud in the Academy Genealogy 2021 5 2 44 10.3390/genealogy5020044
Smith L Killers of the Dream 1994 New York Norton
Smith, L. T. (2012). (). Decolonizing Methodologies: Research and Indigenous Peoples. London Zed, & Books.
Spivak GC Nelson C Grossberg L Can the subaltern speak? Marxism and the interpretation of culture 1988 Urbana University of Illinois Press 271 315
Spivak GC In Other Worlds: Essays in Cultural Politics 1987 London Taylor and Francis
Spracklen K Nazi punks folk off: Leisure, nationalism, cultural identity and the consumption of metal and folk music Leisure Studies 2013 32 4 415 428 10.1080/02614367.2012.674152
Stanley CA When Counter Narratives Meet Master Narratives in the Journal Editorial-Review Process Educational Researcher 2007 36 1 14 24 10.3102/0013189X06298008
Sturma M Packaging Polynesia’s image Annals of Tourism Research 1999 26 3 712 715 10.1016/S0160-7383(99)00022-5
Szasz, T. S. (1973). The second sin. Anchor Press.
Tagore R Gitanjali 1912 London MacMillan and Company Limited
Tagore R Nationalism 2009 New York Penguin
Thapar-Björkert S Farahani F Epistemic modalities of racialised knowledge production in the Swedish academy Ethnic and Racial Studies 2019 42 16 214 232 10.1080/01419870.2019.1649440
Watson B Leisure and the Legacy of bell hooks: A Personal Reflection from Beccy Watson Leisure Sciences 2022 44 3 401 402 10.1080/01490400.2022.2044985
Watson, B., & Ratna, A. (2011). Bollywood in the park: Thinking intersectionality about public leisure space. Leisure/ Loisir, 35(1),71–86.
Watson, B., & Scraton, S. (2001). Confronting Whiteness? Researching the leisure lives of South Asian mothers. Journal of Gender Studies, 10(3),265–277.
Williams RM The Sociology of Ethnic Conflicts: Comparative International Perspectives Annual Review of Sociology 1994 20 49 79 10.1146/annurev.so.20.080194.000405
Wilson WJ The Bridge over the Racial Divide: Rising Inequality and Coalition Politics 1999 Berkeley University of California Press
Wittgenstein L Philosophical Investigations 1963 Oxford Basil Blackwell
Zuberi, T., & Bonilla-Silva, E. (Eds.). (2008). White Logic, White Methods: Racism and Methodology. Lanham, M. A. :Rowman & Littlefield.
| 0 | PMC9703398 | NO-CC CODE | 2022-11-29 23:21:08 | no | Int J Sociol Leis. 2022 Nov 28;:1-18 | utf-8 | null | null | null | oa_other |
==== Front
Curr Breast Cancer Rep
Curr Breast Cancer Rep
Current Breast Cancer Reports
1943-4588
1943-4596
Springer US New York
468
10.1007/s12609-022-00468-w
Breast Cancer Disparities (LA Newman, Section Editor)
Breast Cancer Disparities and the Digital Divide
http://orcid.org/0000-0001-8765-2690
Bayard Solange [email protected]
Fasano Genevieve
Gillot Tamika
Bratton Brenden
Ibala Reine
Taylor Fortson Katherine
Newman Lisa [email protected]
grid.5386.8 000000041936877X Department of Surgery, Weill Cornell Medicine, 525 E 68Th Street, New York-PresbyterianNew York, NY 10065 USA
28 11 2022
18
31 10 2022
© The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2022, Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Abstract
Purpose of Review
Socioeconomically disadvantaged populations and minority groups suffer from high breast cancer mortality, a disparity caused by decreased access to specialty care, lower treatment adherence, co-morbidities, and genetic predisposition for biologically aggressive breast tumor subtypes. Telehealth has the potential to mitigate breast cancer disparities by increasing access to specialty care and health information. However, unequal access to high-speed/broadband internet service and telehealth itself magnifies breast cancer disparities in vulnerable populations. This review evaluates the impact of the digital divide on breast cancer outcomes, as well as strategies for leveraging telehealth to reduce breast cancer disparities.
Recent Findings
There is a paucity of research specific to employing telehealth to address breast cancer disparities. Previous studies provide examples of telehealth utilization for increasing screening mammography, in addition to improving access to breast cancer care, including breast cancer specialist, nurse navigators, and clinical trials. Telehealth can also be used as an approach to risk reduction, with strategies to support weight management and genetic testing.
Summary
Eliminating the digital divide holds enormous potential for mitigating breast cancer disparities through an intentional focus on improving access to telehealth. With increased accessibility, resource allocation, and improved digital infrastructure, telehealth can be used to address disparities in early detection, quality of breast cancer care, treatment adherence, and risk assessment. Further research is essential to elucidate best practices in breast cancer telehealth approaches in underserved communities.
Keywords
Disparities
Telehealth
Breast cancer
Digital divide
COVID-19
==== Body
pmcIntroduction
The COVID-19 pandemic was a catalyst for expanded utilization of telemedicine; however, it also uncovered additional racial disparities within the American healthcare system and among communities nationwide [1, 2]. Non-urgent aspects of breast cancer care were reallocated to remote services or postponed to minimize infectious exposure and maximize hospital capacity for emergencies [3–5]. This phenomenon highlighted potential advantages to incorporating telemedicine into long-term healthcare, while also underscoring the significance and impact of the digital divide on healthcare access [6].
One in 8 women will be diagnosed with breast cancer in their lifetime [7]. Compared to relatively affluent and White American patients, low-income populations and minorities suffer from higher breast cancer mortality, a disparity caused by decreased access to specialty care, lower treatment adherence, co-morbidities, and genetic predisposition for biologically aggressive breast tumor subtypes [7]. Telehealth has the potential to mitigate breast cancer disparities by increasing access to specialty care and health information. However, unequal access to high-speed/broadband internet service and telehealth itself magnifies breast cancer disparities in vulnerable populations. There is a paucity of research specific to telehealth disparities in breast cancer. This review evaluates the impact of the digital divide in breast cancer, as well as strategies to leverage telehealth to improve the breast health and address breast cancer disparities in vulnerable populations.
The Digital Divide
Access
Telehealth refers to clinical and non-clinical technology services, encompassing telephone calls, videos, video conferences, and digital communication, while telemedicine involves technology related to direct services to patients [8]. Over 120 million households have access to broadband internet in the United States; however, racial minorities, low-income populations, and rural communities use the internet and technology at a significantly lower rate [9].
During the COVID-19 pandemic, many aspects of breast cancer care were shifted to telehealth, including visits for initial evaluation, postoperative follow-up, oral chemotherapy/endocrine therapy treatment, radiation oncology, and survivorship [10–18]. Accounting for differential access to compatible devices, broadband internet, technological literacy, and provider reimbursement, at least one in every four Americans did not have access to the technology and digital literacy essential to engage in telehealth visits. Historically underserved populations are especially vulnerable to these disparities and therefore faced numerous barriers to telehealth utilization. It is imperative that the digital divide is addressed strategically as we seek broad implementation of telehealth services across the continuum of breast cancer care [9].
Strategies to Increase Accessibility
Despite inequities in technology, there is enormous potential for telehealth to bridge gaps in breast cancer disparities through increased access and connectivity. With appropriately allocated resources, studies have reported that telehealth was useful among minority populations to increase social support, monitor post-operative symptoms and treatment adherence, and support survivorship [19–23]. Racial minorities reported greater satisfaction with telehealth appointments for oncology versus in-person visits [24, 25•]. In addition, at least one study found that underserved women used telehealth more frequently than affluent women when provided with the relevant materials and technology, emphasizing the importance of resource allocation, policy, and community outreach to increase telehealth access in underserved populations [26].
Increased digital infrastructure can be supported by government-funded initiatives, private companies, and community outreach. This includes resource allocation through tablet loan programs and remote video telehealth clinics. In 2018, the Pittsburg Healthcare System of the United States Department of Veterans Affairs developed a Virtual Cancer Care Network, facilitating virtual oncology care at a tertiary center [27•]. Veterans’ affairs also partnered with large corporations like T-Mobile and Walmart to provide telehealth services in retail stores [28]. Previous studies also found community outreach focused on identifying free public Wi-Fi locations helped patients connect for virtual visits, while nurse navigator programs successfully assisted patients with technical issues [29]. These findings highlight the impact of developing sustainable technology infrastructure in communities and the need for increased community outreach and policy to support similar initiatives.
Digital literacy is also central to telehealth’s utility. Patients over 65 and those with low education have less digital literacy and more anxiety about using telehealth [30, 31]. The importance of improving digital literacy has been recognized nationally with recent efforts from the Department of Education to reinstate the Community Technology Centers (CTC) program that provides underserved communities with information technology and training to improve use. Previous studies have also shown that patients over 65 adopt telehealth well with instruction and practice models, identifying a critical opportunity for nurse navigators to aid in digital literacy programs to improve telehealth utilization. In fact, studies assessing patient comfort with telehealth revealed increases in patient satisfaction with increased use, emphasizing the importance of practice models and nurse navigators to facilitate digital learning, increased telehealth utility, and improved patient satisfaction [31, 32•].
The likelihood that providers will adopt telehealth services is impacted by liability, licensure reimbursement, and concerns about limitations related to the absence of physical exam, decreased ability to bond via in-person/face-to-face interaction, and decreased patient comprehension [33–35]. Strategies to increase telehealth effectiveness must therefore also include increased education at the provider level. During the COVID-19 pandemic, reimbursement for telehealth services was expanded, including Medicare and Medicaid coverage plans, and allowance of telehealth across state lines. Further investment in reimbursement must be prioritized to ensure adequate payment for telehealth services as well as equity in pay in comparison to in-person visits [36].
Disparities in Detection and Early Diagnosis
Breast Cancer Screening
Some studies have demonstrated that low-income women and racial/ethnic minorities including African Americans, Hispanics, and Native Americans/American Indians utilize screening mammography at a lower rate [37, 38]. These same groups also present with later-stage disease. African Americans have the highest breast cancer mortality rates compared to other population subsets in the United States [39, 40]. Screening mammography mitigates disparities in breast cancer mortality through early detection [41]. For these reasons, some societies have advocated for initiation of screening at younger ages for African American women. Patient and provider-level barriers to equitable breast cancer screening include lack of knowledge of guidelines, lack of awareness of personal health records, geographic barriers to accessing imaging centers, financial burdens, and employment conflicts [42].
Screening and Telehealth
Multiple telehealth approaches have been effective at increasing screening utilization. Offman et al. demonstrated effectiveness of automated telephone reminders to underserved urban populations [43]. Health maintenance information pooled from electronic health records and delivered as automated reminders for screening has been similarly effective. In addition, mobile apps have been used to notify patients of screening status and employ remote nurse navigators to aid in scheduling and appointment coordination, which was particularly useful for navigating cultural differences in minority communities [44, 45].
Despite scheduling reminders, geographic access to screening facilities remains as a barrier to breast cancer screening. Mobile mammography has been one of the most effective strategies to improve screening in underserved communities and among African Americans, Hispanic Americans, American Indians, low-income, underinsured, and rural communities [46]. Similar to telephone or app-based appointment reminders improving adherence, telehealth for mobile mammography visits could be used to further facilitate screening use by facilitating appointments, reminders, and location tracking [47].
In this way, telehealth can be leveraged to make screening programs more robust and increase early detection in underserved communities, which could prove particularly useful in accommodating increased need post-pandemic. Screening is particularly important in the context of recovery from reduced screening capacity during the peak of the COVID-19 pandemic—a phenomenon accounting for an approximate deficit of 3.9 million breast cancer screenings in the US [48, 49]. After imaging centers reopened, studies found that rates of screening were disproportionately low among Hispanic and Black populations [50], warranting intentional focus on screening in underserved communities post-pandemic.
Disparities in Access to Treatment
Access
It is well documented that individuals residing in rural settings, racial/ethnic minorities, and underinsured patients are more likely to face disparities in access to specialty breast cancer care and lower treatment adherence [51, 52]. During the COVID-19 pandemic surge, patients presenting for breast cancer treatment were triaged to different care strategies based upon risk stratified models [5]. Those determined to have relatively “low-risk” breast cancer were considered for telephone consultations, while patients considered to have higher risk disease (e.g., those requiring chemotherapy) were triaged to in-person care; the triage process itself was typically handled remotely [5]. Telehealth was in this way utilized for evaluation, follow-up, and symptom monitoring of patients undergoing active treatment with chemotherapy, biological agents, and endocrine therapy.
Increased telehealth in breast cancer treatment during the pandemic demonstrated the critical importance of internet access, connectivity, and communication among patients and providers. This experience supports the findings of several studies highlighting telehealth’s utility in breast cancer treatment. For example, telephone counseling and apps have been shown to decrease cancer-related stress, improve pain management, and increase social support among patients in medically underserved areas [21, 29, 53]. Additionally, telehealth was found to be effective peri-operatively and postoperatively as well as for rehabilitation services [54, 55]. Among breast cancer patients who used telehealth during the pandemic, telehealth services were perceived as increasing access, improving health, and saving time [56]. Few studies however have evaluated racial minorities in the context of socioeconomic, cultural, and language barriers. Although some studies focused on cultural sensitivity in telehealth and app development, further research is warranted to better understand the best methods to employ telehealth in breast cancer care long-term in underserved populations [23, 57].
Nurse Navigators
Nurse navigators have been critical in providing support, coordinating patient-provider communication, and mitigating overall disparities in minority breast cancer care, as racial minorities and underserved patients have less access to support systems during treatment and lower treatment adherence [52]. Remote nurse navigators are particularly important among minority patients who may otherwise be deterred from telehealth services as they face barriers related to language, technological literacy, and mistrust of the medical system [58]. Patient navigators have successfully helped manage pain, emotional support, appointment scheduling, and interpretation of results and are well established in the oncologic community. This model can be further strengthened in approaches to navigating telehealth services, including resource support and instruction to improve digital literacy.
Multidisciplinary Conference
Breast cancer treatment is multidisciplinary. Treatment involves consideration of genetic testing, surgical resection, reconstruction, chemotherapy, endocrine therapy, and radiation treatment. This involves the coordination between genetic counselors, breast surgeons, plastic surgeons, medical oncologist, and radiation oncologists. Minorities and underserved populations have less access to specialty care. Additionally, barriers related to work leave, childcare, transportation, and cost compound the impact of patient appointments. Coordination of individualized breast cancer treatment through multidisciplinary conferences (MDC) improves breast cancer outcomes. Several studies report using videoconferencing for MDC to bring specialized care to remote areas [1]. Increased use of telehealth for MDC among providers could improve communication between providers while increasing access to specialty care and improving efficiency of follow-ups among patients.
Clinical Trials
Clinical oncology trials are critical for the advancement of cancer treatment and patient care. The NIH Revitalization Act of 1993 mandated the inclusion of women and minorities in NIH clinical trials [59, 60]. Despite nationwide efforts to address minority clinical trial enrollment, the accrual of African American (AA) and Hispanic Americans (HA) remains significantly lower when compared to White American (WA) patients [61]. Most clinical trials are aimed at reflecting demographics of the general population. As racial minorities are overrepresented in advanced-stage cancers, clinical oncology trials should be aimed at proportionally enrolling higher rates of AA and HA patients, yet minorities remain disproportionately underrepresented [62].
Underrepresentation in clinical trials contributes to racial disparities, as scientific advancements have limited applicability in diverse populations. Barriers to minority accrual include availability and awareness of clinical trials, supportive infrastructure, strict eligibility criteria, and bias; however, electronic health records are an opportunity to identify cohorts, coordinate recruitment, and standardize enrollment efforts [63–65]. Increased awareness and knowledge of clinical trials through portal messaging has the potential to increase trial opportunities within minority communities [66–68]. Additionally, electronic patient-reported outcome systems allow clinicians to monitor symptoms in real time through online questions, downloadable questions, or telephone technologies [58]. Digital strategies to reduce complications and monitor symptoms can in this way be employed to reduce distress and trigger early medical intervention for concerning symptoms.
Survivorship and Risk Reduction
Breast Cancer and BMI
It is important to consider breast cancer primary and secondary prevention strategies in women, especially among those who are at high risk. A study by Pruthi et al. in 2013 found video telemedicine to be successful in risk-reducing strategies among Alaskan natives [20], Hispanic women [21, 22, 69], and African American women [53, 70].
Obesity is associated with increased risk of breast cancer, as well as later stage at diagnosis and poorer prognosis. Higher body mass index is particularly linked with breast cancer risk in African American women, making weight loss strategies a crucial focus among modifiable risk factors in breast cancer and in breast cancer disparities. Many mobile apps target weight loss and have been trialed in breast cancer patients, with success among minority populations. Oncology patients who engage in regular exercise were found to have better quality of life, reduced depression and anxiety, and decreased recurrence [6, 63–76]. These apps provide tools to track weight loss, monitor exercise and intake, and provide support; however, the majority of previous studies involve the application of these telehealth services among breast cancer survivors. Knowing the role of risk reduction and prevention of breast cancer, these apps could be further utilized to increase understanding of the correlation between breast cancer and obesity, as well as provide tools to reduce risk in the general population.
Genetic Testing
Another proposed method of increasing telehealth accessibility is prioritization of audio-only telephone services that do not require internet connection. This approach works well for certain specialties like telegenetics, the remote delivery of genetic counseling. Telegenetics increases access to genetic counseling and was explored even prior to the pandemic, as it offers similar diagnostic accuracy to in-person counseling [31, 77]. Discerning genetic mutations is essential to breast cancer treatment and patient outcomes due to its influences on screening recommendations, options for chemoprevention, prophylactic resections, and treatment with poly ADP ribose inhibitors. African Americans are 40% more likely to die from breast cancer, which is partially explained by genetic predisposition to triple-negative breast cancer (TNBC), an aggressive subtype linked to West African ancestry and associated with BRCA-1 germline mutations [78, 79].
With increased access through telegenetics, genetic testing has the potential for large-scale impact on precision medicine. Insurance policies to cover more extensive breast cancer panels should accordingly be expanded for high-risk patients. Recent guidelines released by the American Society of Breast Surgeons recommend all breast cancer patients are offered genetic testing [4]. Increased eligibility in comparison to commonly used NCCN genetic testing guidelines is based on evidence of high prevalence of genetic mutations among those who would be ineligible using NCCN guidelines—many of whom are disproportionately African American due to reduced completion of family history [80]. Identifying genetic mutations in patients with African ancestry is crucial to mitigating risk and improving outcomes with more treatment options.
Conclusion/Next Steps
Telehealth interventions have been utilized for breast cancer care for a wide range of applications and across diverse populations; however, there is a paucity of research specific to telehealth disparities in breast cancer. Telehealth holds enormous potential to mitigate breast cancer disparities through intentional focus on eliminating the digital divide. With increased accessibility, resource allocation, and improved digital infrastructure, telehealth can be used to address disparities in early detection, stage at presentation, treatment adherence, genetic testing, and recurrence. Further research is essential to elucidate best practices in breast cancer telehealth approaches in underserved communities.
Acknowledgements
We thank Michelle R Demetres of the Weill Cornell Medical Samuel J. Wood Library for her services in literature review.
Author Contribution
Solange Bayard: writing—original draft; Genevieve Fasano: writing—review and editing; Tamika Gillot: conceptualization, writing—review and editing; Brenden Bratton: writing—review and editing; Reine Ibala: writing—review and editing; Katherine Taylor Fortson: writing—review and editing; Lisa Newman: conceptualization, writing—review and editing.
Declarations
Ethics approval
This article does not contain any studies with human or animal subjects performed by any of the authors.
Competing Interests
The authors declare no competing interests.
This article is part of the Topical Collection on Breast Cancer Disparities
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
Papers of particular interest, published recently, have been highlighted as: • Of importance
1. Onesti CE Tagliamento M Curigliano G Expected medium- and long-term impact of the COVID-19 outbreak in oncology JCO Glob Oncol 2021 7 162 172 10.1200/GO.20.00589 33529077
2. Balogun OD Bea VJ Phillips E Disparities in cancer outcomes due to COVID-19—a tale of 2 cities JAMA Oncol 2020 6 1531 1532 10.1001/jamaoncol.2020.3327 32789508
3. Centers for disease control and prevention. Uses of telehealth during COVID-19 in low resource non-U.S. settings. July 21, 2021. https://www.cdc.gov/coronavirus/2019-ncov/global-covid-19/telehealth-covid19-nonUS.html#print.
4. Manahan ER, Kuerer HM, Sebastian M, Hughes KS, Boughey JC, Euhus DM, et al. Consensus Guidelines on Genetic` Testing for Hereditary Breast Cancer from the American Society of Breast Surgeons. Ann Surg Oncol. 2019;26(10):3025–3031. 10.1245/s10434-019-07549-8.
5. Luther A, Agrawal A. A practical approach to the management of breast cancer in the COVID-19 era and beyond. E cancer medical science. 2020: 14. 10.3332/ecancer.2020.1059.
6. Forum, National Quality. Creating a framework to support measure development for telehealth. 2017. https://www.healthlawinformer.com/wp-content/uploads/2017/06/Creating-a-Framework-to-Support-Measure-Development-for-Telehealth.pdf. Accessed 25 March 2022.
7. Institute, National Cancer. Cancer stat facts: breast cancer. https://seer.cancer.gov/statfacts/html/breast.html. Accessed 25 March 2022.
8. Ryu S. Telemedicine: Opportunities and Developments in Member States: Report on the Second Global Survey on eHealth 2009 (Global Observatory for eHealth Series, Volume 2). Healthc Inform Res. 2012;18(2):153–155. 10.4258/hir.2012.18.2.153.
9. Pew Research Center. . Internet user demographics. http://www.pewinternet.org/data-trend/internet-use/latest-stats/. Accessed 25 March 2022.
10. Coles CE Aristei C Bliss J International guidelines on radiation therapy for breast cancer during the COVID-19 pandemic Clin Oncol 2020 32 279 281 10.1016/j.clon.2020.03.006
11. Clinical guide for the management of cancer patients during the coronavirus pandemic. 2020. https://www.nice.org.uk/media/default/about/covid-19/specialty-guides/cancer-and-covid-19.pdf. Accessed March 2022.
12. NICE COVID-19 rapid guideline: delivery of systemic anticancer treatments. 2020. https://www.nice.org.uk/guidance/ng161. Accessed 25 March 2022.
13. Association of breast surgery statement re COVID-19. 2020. https://associationofbreastsurgery.org.uk/media/252009/abs-statement-150320-v2.pdf. Accessed 25 March 2022.
14. 14. COVID-19 guidelines for triage of breast cancer patients. 2020. https://www.facs.org/covid-19/clinical-guidance/elective-case/breast-cancer. Accessed 25 March 2022.
15. Breast cancer in the COVID-19 era. ESMO, 2020. https://www.esmo.org/guidelines/cancer-patient-management-during-the-covid-19-pandemic/breast-cancer-in-the-covid-19-era. Accessed 25 March 2022.
16. Dietz J, Katharine Yao Uh, Kurtzman S, et al. Recommendations for prioritization, treatment and triage of breast cancer patients during the COVID-19 pandemic: executive summary. 2020. https://www.accc-cancer.org/docs/document/cancer-program-fundamentals/oh-cco-. Accessed 25 March 2022.
17. COVID-19, Association of breast surgery statement re. Apr 27, 2020. https://associationofbreastsurgery.org.uk/media/252026/abs-statement-270420.pdf.
18. Miller RC, Simone BA, Lombardo JF, et all. A pilot trial using telemedicine in radiation oncology: the future of health care is virtual. Telemed Rep 2021: 2: 171–178. 10.1089/tmr.2021.0007.
19. Baik SH, Oswald LB, Buscemi J, et al .Patterns of use of smartphone-based interventions among Latina breast cancer survivors: secondary analysis of a pilot randomized controlled trial. JMIR Cancer. 2020: 6. 10.2196/17538.
20. Pruthi S Stange KJ Malagrino GD Jr Chawla KS LaRusso NF Kaur JS Successful implementation of a telemedicine-based counseling program for high-risk patients with breast cancer Mayo Clin Proc 2013 88 68 73 10.1016/j.mayocp.2012.10.015 23274020
21. Anderson KO Palos GR Mendoza TR Automated pain intervention for underserved minority women with breast cancer Cancer 2015 121 1882 1890 10.1002/cncr.29204 25711974
22. Buscemi J Oswald LB Baik SH My health smartphone intervention decreases daily fat sources among Latina breast cancer survivors J Behav Med 2020 5 732 742 10.1007/s10865-020-00136-3
23. . Donevant S, Heiney SP, Wineglass C, Schooley B, Singh A, Sheng J. Perceptions of endocrine therapy in African-American breast cancer survivors: mixed methods study. JMIR Form Res. 2021:5. 10.2196/23884.
24. O Parsonson A, Grimison P, Boyer M, et al. Patient satisfaction with telehealth consultations in medical oncology clinics: a cross-sectional study at a metropolitan centre during the COVID-19 pandemic. J Telemed Telecare. 2021. 10.1177/1357633X211045586.
25. Johnson BA Lindgren BR Blaes AH The new normal? Patient satisfaction and usability of telemedicine in breast cancer care Ann Surg Oncol 2021 28 5668 5676 10.1245/s10434-021-10448-6 34275045
26. Street RL Jr Treiman K Kranzler EC Oncology patients’ communication experiences during COVID-19: comparing telehealth consultations to in-person visits Support Care Cancer 2022 10.1007/s00520-022-06897-8 35141772
27. • Jiang C.Y., El-Kouri N.T., Elliot D., et al. Telehealth for cancer care in veterans: opportunities and challenges revealed by COVID. JCO Oncol Pract. 2021:22–29. 10.1200/OP.20.00520. Remote delivery of cancer care to geographically remote communities is possible through establishment of clinical video visit clinics.
28. V.L, Elliott. Department of Veterans Affairs (VA): a primer on telehealth. https://crsreports.congress.gov/product/pdf/R/R45834.
29. Dorfman CS Kelleher SA Winger JG Development and pilot testing of an mHealth behavioral cancer pain protocol for medically underserved communities J Psychosoc Oncol 2019 37 335 349 10.1080/07347332.2018.1479327 30585762
30. Loree JM, Dau H, Rebic N, et al. Virtual oncology appointments during the initial wave of the COVID-19 pandemic: an international survey of patient perspectives. Curr Oncol:28: 671–677. doi.org/10.3390/curroncol28010065.
31. Breen KE Tuman M Bertelsen CE Factors influencing patient preferences for telehealth cancer genetic counseling during the COVID-19 pandemic JCO Oncol Pract 2021 10.1200/OP.21.00301 34652959
32. Wu QL Street RL Jr Factors affecting cancer patients’ electronic communication with providers: implications for COVID-19 induced transitions to telehealth Patient Educ Couns. 2021 103 2583 7 10.1016/j.pec.2020.09.036
33. Gordon J, Tsay D, Coyne S, Barchi D, Fleischut PM, Deland EL. Four challenges of launching a telehealth program. NEJM Catal. 2017.
34. Kane CK Gillis K The use of telemedicine by physicians: still the exception rather than the rule Health Aff 2018 37 1923 1930 10.1377/hlthaff.2018.05077
35. Heyer A, Granberg RE, Rising KL, Binder AF, Gentsch AT, Handley NR. Medical oncology professionals’ perceptions of telehealth video visits. JAMA Netw Open. 2021: 4. 10.1001/jamanetworkopen.2020.33967.
36. McGrowder DA Miller FG Vaz K Anderson Cross M The utilization and benefits of telehealth services by health care professionals managing breast cancer patients during the COVID-19 pandemic Healthcare 2021 9 1401 10.3390/healthcare9101401 34683081
37. McGuinness JE Crew KD The challenges of screening mammography in racial/ethnic minority populations in the United States: a mini-review and observations from a predominantly Hispanic communit J Cancer Treatment Diagn 2018 2 16 20 10.29245/2578-2967/2018/2.1128 30112517
38. Ahmed AT Welch BT Brinjikji W Farah WH Henrichsen TL Murad MH Racial disparities in screening mammography in the United States: a systematic review and meta-analysis J Am Coll Radiol 2017 14 157 165 10.1016/j.jacr.2016.07.034 27993485
39. DeSantis CE Fedewa SA Goding Sauer A Breast cancer statistics, 2015: convergence of incidence rates between black and white women CA Cancer J Clin 2016 66 31 42 10.3322/caac.21320 26513636
40. DeSantis CE Ma J Goding Sauer A Newman LA Jemal A Breast cancer statistics, 2017, racial disparity in mortality by state CA Cancer J Clin 2017 67 439 448 10.3322/caac.21412 28972651
41. Bayard S Fasano G Chen Y Davis M Screening mammography mitigates breast cancer disparities through early detection of triple negative breast cancer Clin Imaging 2021 80 430 437 10.1016/j.clinimag.2021.08.013 34543867
42. Stoll CR Roberts S Cheng MR Crayton EV Jackson S Politi MC Barriers to mammography among inadequately screened women Health Educ Behav 2015 42 8 15 10.1177/1090198114529589 24722216
43. Offman J Myles J Ariyanayagam S A telephone reminder intervention to improve breast screening information and access Public Health 2014 128 1017 1022 10.1016/j.puhe.2014.09.007 25443131
44. Percefull J Butler J Improving mammography through effective screening, brief intervention, and referral to treatment at a rural health center J Am Assoc Nurse Pract 2021 33 324 330 10.1097/JXX.0000000000000356
45. Lee H, Ghebre R, Le C, Jang YJ, Sharratt M, Yee D. Mobile phone multilevel and multimedia messaging intervention for breast cancer screening: pilot randomized controlled trial. JMIR Mhealth Uhealth. 2017:5. 10.2196/mhealth.7091.
46. Vang S Margolies LR Jandorf L Mobile mammography participation among medically underserved women: a systematic review Prev Chronic Dis 2018 10.5888/pcd15.180291 30447104
47. Chen YR, Chang-Halpenny C, Kumarasamy NA, Venegas A, Braddock Iii CH. Perspectives of Mobile Versus Fixed Mammography in Santa Clara County, California: A Focus Group Study. Cureus. 2016;8(2):e494. 10.7759/cureus.494.
48. Chen RC, Haynes K, Du S, Barron J, Katz AJ. Association of cancer screening deficit in the United States with the COVID-19 pandemic. JAMA Oncol. 2021: 878–884. 10.1001/jamaoncol.2021.0884.
49. Preventive Cancer Screenings during COVID-19 Pandemic. Epic Health Research Network. (May 4, 2020). Retrieved from: https://epicresearchblob.blob.core.windows.net/cms-uploads/pdfs/Preventive-Cancer-Screenings-during-COVID-19-Pandemic.pdf. Accessed 25 March 2022.
50. Velazquez AI, Hayward JH, Gregory B, Dixit N. Pandemic trends in breast cancer screening in a safety-net hospital during the COVID-19. JAMA Netw Open. 2021: 4. 10.1001/jamanetworkopen.2021.19929.
51. Philipovskiy A, Campbell A, Heydarian R, et al. Adherence to adjuvant aromatase inhibitor therapy among postmenopausal Hispanic/Latino women with breast cancer. Anticancer Res. 2020. 40. 857–864. 10.21873/anticanres.14018.
52. Zavala VA Bracci PM Carethers JM Cancer health disparities in racial/ethnic minorities in the United States Br J Cancer 2021 124 315 332 10.1038/s41416-020-01038-6 32901135
53. Allicock M Kendzor D Sedory A A pilot and feasibility mobile health intervention to support healthy behaviors in African American breast cancer survivors J Racial Ethn Health Disparities 2021 8 157 165 10.1007/s40615-020-00767-x 32385847
54. Lai LL Player H Hite S Satyananda V Stacey J Sun V Jones V Hayter J Feasibility of remote occupational therapy services via telemedicine in a breast cancer recovery program Am J Occup Ther 2021 10.5014/ajot.2021.042119 33657345
55. Pelle F Cappelli S Graziano F Breast cancer surgery during the Covid-19 pandemic: a monocentre experience from the Regina Elena National Cancer Institute of Rome Exp Clin Cancer Res 2020 10.1186/s13046-020-01683-y
56. Zimmerman BS Seidman D Berger N Patient perception of telehealth services for breast and gynecologic oncology care during the COVID-19 pandemic: a single center survey-based study J Breast Cancer 2020 23 542 552 10.4048/jbc.2020.23.e56 33154829
57. Banas JR Victorson D Gutierrez S Cordero E Guitleman J Haas N Developing a peer-to-peer mHealth application to connect Hispanic cancer patients J Cancer Educ 2017 32 158 165 10.1007/s13187-016-1066-6 27364905
58. Basch E Stover AM Schrag D Clinical utility and user perceptions of a digital system for electronic patient-reported symptom monitoring during routine cancer care: findings from the PRO-TECT trial JCO Clin Cancer Inform 2020 4 947 957 10.1200/CCI.20.00081 33112661
59. The unequal burden of cancer: an assessment of NIH research and programs for ethnic minorities and the medically underserve. Medicine, Institute of. 1999, Washington, DC: National Academy Press.
60. Health, National Institute of. Monitoring adherence to the NIH policy on the inclusion of women and minorities as subjects in clinical research. http://www4.od.nih.gov/orwh/GoldReport_2000‐01.pdf .
61. Stewart JH Bertoni AG Staten JL Levine EA Gross CP Participation in surgical oncology clinical trials: gender-, race/ethnicity-, and age-based disparities Ann Surg Oncol 2007 14 3328 3334 10.1245/s10434-007-9500-y 17682824
62. Hamel LM Penner LA Albrecht TL Heath E Gwede CK Eggly S Barriers to clinical trial enrollment in racial and ethnic minority patients with cancer Cancer Control 2016 23 327 337 10.1177/107327481602300404 27842322
63. Ford JG Howerton MW Lai GY Barriers to recruiting underrepresented populations to cancer clinical trials: a systematic review Cancer 2008 112 228 242 10.1002/cncr.23157 18008363
64. Adams-Campbell LL Ahaghotu C Gaskins M Enrollment of African Americans onto clinical treatment trials: study design barriers J Clin Oncol 2004 22 730 734 10.1200/JCO.2004.03.160 14966098
65. Howerton MW Gibbons MC Baffi CR Provider roles in the recruitment of underrepresented populations to cancer clinical trials Cancer 2007 109 465 476 10.1002/cncr.22436 17200964
66. Skinner JS Fair AM Holman AS Boyer AP Wilkins CH The impact of an educational video on clinical trial enrollment and knowledge in ethnic minorities: a randomized control trial Front Public Health 2019 26 104 10.3389/fpubh.2019.00104
67. Ellis PM Butow PN Tattersall MH Dunn SM Houssami N Randomized clinical trials in oncology: understanding and attitudes predict willingness to participat J Clin Oncol 2001 19 3554 3561 10.1200/JCO.2001.19.15.3554 11481363
68. George S, Duran N, Norris K. A systematic review of barriers and facilitators to minority research participation among African Americans, Latinos, Asian Americans, and Pacific Islanders. Am J Public Health. 2014;104(2):e16–31. 10.2105/AJPH.2013.301706.
69. Meneses K Gisiger-Camata S Benz R Raju D Bail JR Benitez TJ Pekmezi D McNees P Telehealth intervention for Latina breast cancer survivors: a pilot Womens Health 2018 10.1177/1745506518778721
70. Ferrante JM Devine KA Bator A Feasibility and potential efficacy of commercial mHealth/eHealth tools for weight loss in African American breast cancer survivors: pilot randomized controlled trial Transl Behav Med 2020 10 938 948 10.1093/tbm/iby124 30535101
71. Cormie P Zopf E Zhang X Schmitz K The impact of exercise on cancer mortality, recurrence, and treatment-related adverse effects Epidemiol Rev 2017 39 71 92 10.1093/epirev/mxx007 28453622
72. Ariza-Garcia A Lozano-Lozano M Galiano-Castillo N Postigo-Martin P Arroyo-Morales M Cantarero-Villanueva I A web-based exercise system (e-CuidateChemo) to counter the side effects of chemotherapy in patients with breast cancer: randomized controlled trial J Med Internet Res 2019 10.2196/14418 31342907
73. Wagner LI Tooze JA Hall DL Targeted eHealth intervention to reduce breast cancer survivors’ fear of recurrence: results from the FoRtitude randomized trial J Natl Cancer Inst 2021 113 1495 1505 10.1093/jnci/djab100 34057469
74. Koç Z Kaplan E Tanrıverdi D The effectiveness of telehealth programs on the mental health of women with breast cancer: a systematic review J Telemed Telecare 2022 10.1177/1357633X211069663 35108131
75. van der Hout A Holtmaat K Jansen F Lissenberg-Witte BI The eHealth self-management application ‘Oncokompas’ that supports cancer survivors to improve health-related quality of life and reduce symptoms: which groups benefit most? Acta Oncol 2021 60 403 411 10.1080/0284186X.2020.1851764 33345659
76. Hauffman A Alfonsson S Bill-Axelson A Cocreated internet-based stepped care for individuals with cancer and concurrent symptoms of anxiety and depression: results from the U-CARE AdultCan randomized controlled trial Psychooncology 2020 29 2012 2018 10.1002/pon.5489 32691455
77. Stalker HJ, Wilson R, McCune H, et al. Telegenetic medicine: improved access to services in an underserved area. J Telemed Telecare. 2006: 182–185.
78. Newman LA Kaljee LM Health disparities and triple-negative breast cancer in African American women: a review JAMA Surg 2017 152 485 493 10.1001/jamasurg.2017.0005 28355428
79. Kohler BA, Sherman RL, Howlader N, et al. Annual report to the nation on the status of cancer, 1975–2011, featuring incidence of breast cancer subtypes by race/ethnicity, poverty, and state . J Natl Cancer Inst. 2015: 107.
80. Murff HJ Byrne D Haas JS Puopolo AL Brennan TA Race and family history assessment for breast cancer J Gen Intern Med 2005 20 75 80 10.1111/j.1525-1497.2004.40112.x 15693932
| 36467667 | PMC9703401 | NO-CC CODE | 2022-11-29 23:21:41 | no | Curr Breast Cancer Rep. 2022 Nov 28;:1-8 | utf-8 | Curr Breast Cancer Rep | 2,022 | 10.1007/s12609-022-00468-w | oa_other |
==== Front
AIDS Behav
AIDS Behav
AIDS and Behavior
1090-7165
1573-3254
Springer US New York
36441406
3925
10.1007/s10461-022-03925-w
Original Paper
The Relationship Between Intersectional Drug Use and HIV Stigma and HIV Care Engagement Among Women Living with HIV in Ukraine
http://orcid.org/0000-0003-2380-5533
Owczarzak Jill [email protected]
1
Fuller Shannon 2
Coyle Catelyn 3
Davey-Rothwell Melissa 2
Kiriazova Tetiana 4
Tobin Karin 2
1 grid.21107.35 0000 0001 2171 9311 Department of Health, Behavior and Society, Johns Hopkins Bloomberg School of Public Health, 624 N. Broadway Ave, Room 739, Baltimore, MD 21205 USA
2 grid.21107.35 0000 0001 2171 9311 Department of Health, Behavior and Society, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD 21205 USA
3 grid.21107.35 0000 0001 2171 9311 Department of Mental Health, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD 21205 USA
4 grid.478065.8 Ukrainian Institute on Public Health Policy, Kiev, 04050 Ukraine
28 11 2022
112
6 11 2022
© The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2022, Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
This study used an intersectional approach to explore the association between enacted and internalized drug use and HIV stigma on HIV care outcomes among HIV-positive women who inject drugs in Ukraine. Surveys were conducted in Kyiv in 2019–2020. Among the 306 respondents, 55% were engaged in HIV care. More than half (52%) of participants not engaged in care reported internalized stigma related to both drug use and HIV status (i.e., intersectional stigma), compared to only 35% of those who were engaged in HIV care. Among those engaged in care, 36% reported intersectional enacted stigma compared to 44% of those not engaged in care; however, this difference was not statistically significant in the univariable analysis (p = 0.06). In the univariable analysis, participants who reported intersectional internalized stigma had 62% lower odds of being engaged in HIV care (OR 0.38, 95% CI 0.22, 0.65, p < 0.001). In the adjusted model, reported intersectional internalized stigma (aOR 0.52, 95% CI 0.30, 0.92, p = 0.026), reported intersectional enacted stigma (aOR 0.47, 95% CI 0.23, 0.95, p = 0.036), and knowing their HIV status for more than 5-years (aOR 2.29, 95% CI 1.35, 3.87, p = 0.002) were significant predictors of HIV care engagement. These findings indicate that interventions to improve HIV care engagement must address women’s experiences of both HIV and drug use stigma and the different mechanisms through which stigma operates.
Supplementary Information
The online version contains supplementary material available at 10.1007/s10461-022-03925-w.
Keywords
Intersectional stigma
Drug use
HIV care engagement
Women living with HIV
Ukraine
National Institutes of Health Fogarty International CenterR21TW011060-01 Owczarzak Jill
==== Body
pmcIntroduction
In Ukraine, which has one of the largest HIV epidemics in Europe, about 250,000 people are living with HIV and 15,968 people acquired HIV in 2020 [1]. An estimated 82% of people living with HIV (PLWH) who knew there HIV status were linked to care and receiving antiretroviral therapy (ART) at the beginning of 2020, up from 42% in 2013 [2]. In Ukraine, HIV care is only delivered through specialized publicly funded AIDS centers and their affiliates, and HIV medication and treatment is provided for free to all patients registered with an AIDS center. Previous studies have identified factors such as low HIV knowledge and a fragmented, bureaucratically complex health care system as barriers to access HIV care [3]. Despite overall improvements in HIV care cascade indicators, significant gaps remain among women living with HIV (WLWH). New HIV infections among women (aged 25 and older) more than doubled in seven years, from 1814 in 2005 to 5057 in 2012 [4]. Only 58% of HIV-positive people who inject drugs (PWID) are aware of their HIV status and 70% of PWID aware of their HIV-positive status were on ART [5].
As of 2016, women constituted 42.5% of PLWH in Ukraine [6]. While only 18% of the total sample in one study of Ukrainian WLWH indicated intravenous drug use as the mode of transmission, these women represented nearly half of the women who presented with AIDS [7]. An analysis of the 2017 Integrated Bio-Behavioral Surveillance survey data collected among PWID across 30 Ukrainian cities found that the HIV prevalence among women who inject drugs was 31.4% [5]. Among the WLWH in the sample, 63.7% were aware of their HIV status and of those, 69.9% initiated ART and of those, 72.6% were virally suppressed. However, of the total number of WLWH in the sample, just 47% were on ART and 35% were virally suppressed [5].
Ukrainian WLWH with histories of drug use experience worse health outcomes and greater negative social interactions than other WLWH. Among Ukrainian WLWH who indicated injection drug use as their mode of infection, the median delay between diagnosis and registration in care (i.e., linking to care and enrolling as a new patient) was 87 days, compared to 34 days for women diagnosed through antenatal testing [7]. In a cohort study of 8884 HIV-positive Ukrainian women, negative birth outcomes (preterm delivery and small for gestational age) were associated with history of injection drug use [8]. The Stigma Index, which included results from 1500 PLWH, found that 24% of WLHW, compared to 7% of men living with HIV, were advised by a health care professional not to have children [9]. One study documented that women in Ukraine who received a positive HIV diagnosis as part of standard of care for pregnancy often encountered negative or stigmatizing experiences, such as being told not to get married and to have an abortion, and doctors talking about their test results openly or telling partners, employers, and family members about the test result without the woman’s consent or knowledge [10]. In response to internalized negative feelings and negative experiences with service providers and family [11], WLWH in Ukraine make self-discriminating decisions, including not applying for jobs or social assistance, avoiding medical visits, and isolating themselves from family and friends [12].
Women who use drugs must also navigate complicated bureaucratic systems that reinforce their marginalized status, create barriers to government support, and fail to meet their specific needs as mothers and people who use substances [13, 14, 15]. Research indicates that stigma, discrimination, and loss of confidentiality in the service delivery setting also impact the ability of women who use drugs to receive essential services [16, 17]. Women who use drugs are more likely to delay approaching health facilities than men who use drugs (World Health [18]. Women may try to hide their drug use to avoid negative reactions associated with breaching stereotyped gender roles, or may fear that their children will be removed from their custody if authorities learn of their drug use from entering or leaving a needle exchange site [19, 20]. Women who use drugs also face additional stigma in healthcare settings and from law enforcement due to gender disparities and the lack of tailored services towards women [21]. WLWH and histories of drug use must also navigate repressive drug policies, including detention without arrest, police harassment, registration as a drug user to receive services, and diminished employment opportunities [21]. Ciambrone [22] found that among WLWH in the United States, those with histories of drug use were less likely to have supportive familial networks than other WLWH, and women expressed concern about disclosing their HIV status due to its association with stigmatized behaviors such as drug use.
Stigma and HIV Outcomes
As the experiences of WLWH suggest, HIV and drug use stigma remain a major barrier to care, leading to poor health outcomes [23]. Stigma is a social phenomenon in which individuals or groups are devalued based on certain traits and the experiences of exclusion, rejection, judgment, and blame that result [24, 25]. As Link and Phelan [26, p. 367] note, “stigmatization is entirely contingent on access to social, economic, and political power” and people assigned to stigmatized categories are disadvantaged in terms of socioeconomic status, medical treatment, and health. Earnshaw & Chaudoir [24] use the concept of a stigma mechanism to explain how possessing a devalued attribute leads to a negative outcome, including how other people react to this knowledge and how the individual with the devalued attribute reacts. At the individual level, stigma affects health outcomes through three primary mechanisms: anticipated, internalized, and enacted. Anticipated stigma refers to degree to which an individual expects they will experience prejudice or discrimination. Enacted stigma is defined as experiences of stigma enacted on the individual from an external force, such as experiences of discrimination, stereotyping, and prejudice [24]. Individuals may experience poor treatment from health care providers, rejection from friends and family, social isolation, and lack of employment because they belong to a stigmatized group. Internalized stigma refers to the degree to which one endorses the negative attitudes and beliefs that one is devalued because of their possession of a devalued trait [24]. Internalized stigma is associated with negative health outcomes, particularly depression, psychological distress [27], feelings of helplessness, and lower acceptance of one’s health condition [28, 29]. As a result of these stigma mechanisms, people may not engage in health-seeking behaviors to avoid thinking about their condition or a lack of a sense of self-worth [30], delay seeking care, and experience low social support and its attendant consequences (e.g., economic insecurity, depression, isolation) [31, 32].
Contemporary stigma research recognizes that people may belong to multiple stigmatized groups and that focusing on a single dimension of disadvantage (such as socioeconomic status or gender) obscures the ways in which these multiple identities interact and are compounded to produce specific health outcomes or statuses [33, 34]. Recognizing that people may belong to multiple stigmatized groups and understanding how these multiple identities affect health and well-being requires an intersectional approach. An intersectional perspective is based on the tenet that social categories are interdependent and mutually constitutive; one aspect of identity cannot explain health outcomes or social disadvantage [23, 35]. Historically, PLWH have been highly stigmatized for both their HIV status and negative attitudes about modes of HIV transmission and identities of groups vulnerable to HIV, including injection drug use.
The scientific evidence is consistent across populations and settings that stigma is associated with poor outcomes. HIV stigma is associated with poor HIV medication adherence [36], worse health outcomes [32], lower health care utilization [37], and low self-reported health status, low HIV medication adherence, and diminished mental health [38, 39]. Drug use stigma is also associated with poorer access to health care [37], suboptimal engagement in HIV care [40], increased injection risk behaviors [41], and decreased use of harm reduction and medical services [42]. Other studies have shown enacted and internalized HIV stigma are associated with poorer HIV care outcomes [43, 44, 45].
Although there is an extensive body of literature on the impacts of various types of stigma on health outcomes, little is known about the intersection of stigma related to HIV and drug use among women and the mechanisms through which this stigma affects HIV care outcomes. The experiences of WLWH who also use drugs are embedded in and shaped by gender norms, attitudes, and expectations around ideal womanhood and negative stereotypes about people who use drugs [46, 47, 48, 49]. This paper explores the relationship between enacted and internalized drug use and HIV stigma and HIV care outcomes among WLWH who inject drugs in Ukraine. As Metsch et al. [21] note, data on women who use drugs is often combined with data from women involved sex work or partners of people who inject drugs and few countries disaggregate data by gender or drug use. Understanding the effects of intersectional stigma and the mechanisms through which stigma affects HIV care engagement is a key step in developing tailored interventions that can improve all points along the HIV care continuum. Identifying and modifying barriers to HIV care is particularly important in Ukraine, where significant resources have been invested to improve access to treatment and care but gaps in the care cascade remain.
Methods
Study Setting and Procedures
This study was conducted among WLWH who reported recent injection drug use in Kyiv City, Ukraine. Kyiv is Ukraine’s largest city and in 2019, 1490 new cases of HIV were detected in the city. Overall, in Kyiv City, only about 52% of newly diagnosed PLWH were linked to HIV care services (compared to 82.6% nationally) [2]. All participants provided oral informed consent prior to data collection. Study procedures were reviewed and approved by the Institutional Review Boards at Johns Hopkins Bloomberg School of Public Health and the Ukrainian Institute on Public Health Policy.
Participants
Between December 2019 and November 2020, 310 WLWH and histories of injection drug use were recruited. Inclusion criteria included: identifying as female, living in Kyiv City, being 18 years of age or older, reporting injection drug use in the previous 3 months, and being HIV-positive (confirmed with a rapid test at the time of study consent). Participants were recruited using a combination of direct outreach and participant referral methods. At the outreach routes (where PWID come to have rapid HIV test or to get syringes) and at the needle and syringe program (NSP) sites of two Kyiv-based nongovernmental organizations that work with PLWH, social workers approached those female clients who had been previously or recently tested HIV-positive and asked them if they were interested in participating in a research study. Women recruited from service agencies were asked to refer friends and associates to the study, particularly those who are not currently receiving services from that agency. Women were given information about the study to distribute to interested family and friends. Study staff told interested women about the study and offered them participation, contingent on eligibility. Individuals who agreed to take part in the study were given the contact information for study staff to schedule a time, date, and place to conduct an eligibility screening and complete the survey if eligible.
Data Collection
Study participants completed a single, two-part survey. The first part of the survey was interviewer-led in Russian or Ukrainian according to the participant’s preference. This section of the survey elicited demographic information and an egocentric social network. An egocentric social network, or personal network, asks a participant (i.e., ego) to identify social contacts (i.e., alters) and characterize their relationships [50]. In this study, the interviewer first prompted the participant to list all the individuals in their social network. To elicit a comprehensive list of social network members (alters), participants were asked a series of structured questions that included who they lived with, talked to about personal things, lent money to or borrowed money from, helped them with small tasks, socialized with, injected drugs with, had sex with, and worked with. Using this list of alters, the interviewer then followed a structured questionnaire to elicit demographic, relationship, and behavioral information about the alters. The second part of the survey (covering HIV care, drug use, social support, and mothering) was self-administered in Russian. The survey took one and half to two hours to complete. Participants received 400 hryvnia (about 15 USD) for completing the survey.
Measures
Enacted and Internalized Stigma
Enacted drug use stigma was assessed using eight questions, based on Earnshaw et al. [28] and modified based on preliminary research. For each alter in their social network, participants were asked, “How frequently does this person treat you this way because of your drug use.” The questions were (1) “avoid you,” (2) “look down on you,” (3) “treat you differently,” (4) “did not take you seriously,” (5) “avoid touching you,” (6) “treated you with less respect,” (7) “not listened to you,” (8) “been critical of your behavior.” The answer options for each were “never,” “rarely,” “50/50,” “sometimes,” “always.” The same questions were asked about how alters treated the participant based on their HIV status. Participants were categorized as experiencing enacted stigma related to drug use or HIV status if they replied “50/50,” “sometimes” or “always” to any of the eight questions for at least one alter.
Internalized drug use stigma was assessed using nine questions [28]. Participants were asked if: (1) Using drugs makes me feel like I’m a bad person; (2) I feel I’m not as good as others because I use drugs; (3) I feel ashamed of using drugs; (4) I think less of myself because I use drugs; (5) Using drugs makes me feel unclean; (6) Using drugs is disgusting to me; (7) I don’t pay attention to how others feel about me because I use drugs; (8) I accept using drugs as part of how I am right now; and (9) I take and make opportunities to educate people around me about drug use. The answer options were “strongly disagree,” “disagree,” “neither disagree nor agree,” “agree,” and “strongly agree.” To measure internalized HIV stigma, the participants were asked the same nine questions but as they related to their HIV status. The answer options were the same. Scores were calculated by summing responses (1 = strongly disagree, to 5 = strongly agree) and dividing by the 9 questions, resulting in a scale from 1 (no stigma) to 5 (high stigma). Items 7–9 were reverse coded. The median values of 3.56 and 2.78 were used as the cut-points to define internalized drug use and HIV stigma, respectively (cf. [37]).
Intersectional Drug Use & HIV Stigma
To assess the intersectionality of enacted HIV and drug use stigma, a categorical variable with four options was created: (1) the participants did not experience any enacted stigma by social network members; (2) the participant experienced enacted stigma from their social network related to their drug use only; (3) the participant experienced enacted stigma from their social network related to their HIV status only; and (4) the participant experienced enacted stigma from their social network related to their drug use and HIV status.
To assess the intersectionality of internalized HIV and drug use stigma, a categorical variable with four options was created: (1) the participants reported no internalized stigma; (2) the participant reported internalized stigma related to their drug use only; (3) the participant reported internalized stigma related to their HIV status only; and (4) the participant reported internalized stigma related to their drug use and HIV status.
Outcome Measure: HIV Care Engagement
The outcome of interest was HIV care engagement. Participants were considered “in care” if they reported two HIV visits more than 3-months apart within the last year or if they had one HIV visit within 3-months of their study visit, following the Health Resources and Services Administration (HRSA) definition of HIV care retention [51]. All other participants were classified as “not in care,” for example those who reported less than two HIV care visits in the previous 12-months or if they had not received HIV care within the last 12-months.
Other Covariates
Age was dichotomized using median age as the cut-point. Long-term relationship included participants who stated they were married, in a long-term relationship but not living together, or living as married but not officially married. If a participant was single, divorced, widowed, or other, then they were not considered being in a long-term relationship. Having any children was a dichotomous yes/no variable. Those who reported any children were dichotomized as those who had reported having any children less than 18 years old or 18 years of age or greater. Participants were asked to describe their current financial situation. Responses were categorized as either just able or unable to meet basic needs versus able to meet most or all needs. Employment status was categorized as regular/steady work, which was defined as regular full or part-time work; occasional work/unemployed, defined as occasional or seasonal work, or being unemployed; and not looking for work, defined as being on disability and therefore unable to work or a homemaker, full-time student, or retired. Education history was dichotomized as any post-secondary education (technical college, non-finished higher education, or completed higher education at a university) or less (10–12 years of secondary school, or 8–9 years or less of secondary school). The CESD-10 was used to assess depressive symptoms, with a score of ≥ 10 indicating signs of depressive symptoms [52]. Time since HIV diagnosis was dichotomized using the median number of years (5 years) as the cut-point. Finally, participants were asked to report which drug they used more frequently. The answer options were stimulant, street methadone (i.e., purchased on the street or not received through an OAT program), street buprenorphine, homemade opioids (e.g., poppy straw extract, shirka, chernaya), medical opioids (e.g., morphine, tincture opium, codeine, tramadol, omnopon, promedol, kodterpin), or other opioids from a pharmacy (e.g., tropicamid, rhinasolin). The variable was dichotomized as either stimulant or opioid.
Statistical Analysis
Participant characteristics by HIV care engagement were compared using Pearson’s chi-squared test or Fisher's exact test for categorical and binary variables. Four participants had missing data and were not included in the analysis. Univariable and multivariable logistic regression analyses were used to examine the association between experiencing enacted and/or internalized stigma and HIV care engagement. Variables in the final multivariable model were chosen based on a priori hypothesis because they were associated with HIV care in the literature or had a p-value < 0.01 in the univariable analysis. In addition to the exposures (internalized and enacted stigma), variables in the multivariable model included age, whether the participant has any children, the length of time that the participant has known their HIV-positive diagnosis, and injection drug use frequency. The model selection process explored the inclusion of an interaction term between the internalized and enacted stigma mechanisms; however, this model was limited by small cell sizes and did not improve model fit based on Akaike Information Criteria (AIC) values. Additionally, due to small cell sizes in the group that reported enacted HIV stigma only, a model was tested that excluded the three participants represented in that category. That model produced similar results as the final multivariable model outlined above. All analysis was performed using StataMP 17 (StataCorp, College Station, Texas).
Results
In this sample of 306 HIV-positive women who inject drugs, 55% were engaged in HIV care (Table 1). The median age was 34 years, 58% were in a long-term relationship, 9% had enough money to meet most financial needs, 36% had regular employment, and 64% had technical training or some level of higher education. A higher percentage of women engaged in HIV care had children (61% versus 41%, p = 0.04). In both groups, opioids were the most used drug and a higher percentage of participants who were not engaged in HIV care injected drugs daily (53% versus 44%, p = 0.1). Most participants did not report any experience in medication-assisted treatment (MAT) programs for drug use (37% had been enrolled in MAT among those engaged in HIV care and 36% among those not engaged in HIV care, p = 0.93). A more detailed breakdown of participant characteristics is available in Supplemental Table I.Table 1 Participant characteristics by HIV care engagement status for 306 HIV-positive women who inject drugs, Kyiv, Ukraine, November 2019–November 2020
Not in HIV care In HIV care Test statistic Degrees of Freedom p-value
N 138 168
Demographic characteristics
Age 34 years or older 66 (47.8%) 97 (57.7%) 2.99 1 0.08
Some higher education or more 87 (63.0%) 109 (64.9%) 0.11 1 0.74
In long-term relationship 77 (55.8%) 99 (59.3%) 0.74 1 0.54
Has enough money to make ends meet 15 (10.9%) 13 (7.7%) 0.89 1 0.34
Employment status
Regular/steady work 52 (38.8%) 55 (33.5%) 0.07
Occasional work/unemployed 70 (52.2%) 79 (48.2%) 5.38 2
Not looking for work 12 (9.0%) 30 (18.3%)
Has any children 41 (29.7%) 69 (41.1%) 4.25 1 0.04
Has children under the age of 18 years of age 36 (26.1%) 56 (33.3%) 1.89 1 0.17
HIV health-related
Known HIV positive status for at least 5 years 57 (41.3%) 103 (61.3%) 12.15 1 < 0.01
Mental health
Depressive symptoms (CES-D ≥ 10) 94 (68.1%) 109 (64.9%) 0.36 1 0.55
Substance use
Drug used most frequently
Opioids 99 (72.3%) 137 (81.5%) (81.5%) 3.72 1 0.05
Stimulants 38 (27.7%) 31 (18.5%)
Injects drugs almost daily or more 72 (52.9%) 73 (43.5%) 2.71 1 0.10
Ever involved in any drug treatment program 55 (40.1%) 62 (36.9%) 0.40 1 0.53
Ever enrolled in a MAT program 20 (36.0%) 23 (37%) 0.01 1 0.93
Currently enrolled in a MAT program 6 (11%) 9 (15%) 0.34 1 0.56
Stigma
Type of internalized stigma reported
Neither drug use nor HIV 34 (24.6%) 74 (44.1%) < 0.01
Drug use only 21 (15.2%) 20 (11.9%) 13.47 3
HIV only 12 (8.7%) 15 (8.9%)
Intersectional drug use and HIV 71 (51.5%) 59 (35.1%)
Type of enacted stigma reported
Neither drug use nor HIV 22 (15.9%) 43 (25.6%) 0.06 (Fisher’s exact)
Drug use only 55 (39.9%) 64 (38.1%) – 3
HIV only 0 (0.0%) 3 (1.8%)
Intersectional drug use and HIV 61 (44.2%) 61 (36.3%)
When comparing reported internalized stigma, there were significant differences (p < 0.01) between participants engaged in care versus not engaged in care. A larger percentage of women engaged in HIV care did not report any internalized stigma related to drug use or HIV status (44% versus 25%) compared to those not in care. More than half (52%) of the participants who were not engaged in care reported internalized stigma related to both drug use and HIV status, compared to only 35% of those who were engaged in HIV care. Although there were differences in the frequency of reported enacted stigma between the care engagement groups, these differences were not statistically significant (p = 0.10). Roughly 25% of those in HIV care and 16% of those not in HIV care did not report enacted stigma related to drug use or HIV. Among those not in care versus in care, 44% and 36% reported enacted stigma for both drug use and HIV status, respectively.
In the univariable analysis, participants who reported internalized stigma for drug use only had 56% lower odds of being engaged in HIV care compared those who reported no internalized stigma (OR 0.44, 95% CI 0.21, 0.91, p = 0.027). Those who reported internalized stigma related to HIV status only had 43% lower odds of being engaged in HIV care (OR 0.57, 95% CI 0.24, 1.36, p = 0.21), and those who reported intersectional stigma had 62% lower odds (OR 0.38, 95% CI 0.22, 0.65, p < 0.001) compared to those who reported no internalized stigma (Table 2). Participants who reported enacted intersectional stigma had 49% (OR 0.51, 95% CI 0.27, 0.96, p = 0.035) lower odds of being engaged in HIV care compared to those who reported no enacted stigma. Having children (OR 1.65, 95% CI 1.02, 2.66, p = 0.04) and being diagnosed with HIV for more than 5-years (OR 2.25, 95% CI 1.42, 3.57, p = 0.001) were both associated with higher odds of HIV care engagement. In the adjusted model, reported intersectional internalized stigma (aOR 0.52, 95% CI 0.30, 0.92, p = 0.026), intersectional enacted stigma (aOR 0.47, 95% CI 0.23, 0.95, p = 0.036), and knowing their HIV status for more than 5-years (aOR 2.29, 95% CI 1.35, 3.87, p = 0.002) remained significant predictors of HIV care engagement after adjusting for age, whether they had children, and injection drug use frequency.Table 2 Association between stigma and engagement in HIV care among 306 HIV-positive women who inject drugs participants, Kyiv, Ukraine, November 2019–November 2020
OR (95% CI) aOR (95% CI)
Internalized stigma
Neither stigma in both categories Reference Reference
Drug use stigma only 0.44 (0.21, 0.91)*** 0.40 (0.18, 0.89)*
HIV stigma only 0.57 (0.24, 1.36) 0.79 (0.32, 1.97)
Both drug use and HIV status 0.38 (0.22, 0.65)*** 0.54 (0.31, 0.95)*
Enacted Stigma
Neither stigma in both categories Reference Reference
Drug use stigma only 0.57 (0.30, 1.07) 0.49 (0.24, 0.98)*
HIV stigma only –a –a
Both drug use and HIV status 0.51 (0.27, 0.96)* 0.44 (0.21, 0.89)*
Age
Less than 34 years Reference Reference
34 years or older 1.49 (0.95, 2.34)* 1.13 (0.68, 1.88)
Has any children
No Reference Reference
Yes 1.65 (1.02, 2.66)* 1.58 (0.94, 2.66)
Time since HIV diagnosis
Less than 5 years Reference Reference
5 years or more 2.25 (1.42, 3.57)** 2.31 (1.36, 3.94)*
Injects drugs almost daily or more
No Reference Reference
Yes 0.68 (0.43, 1.08) 0.72 (0.44, 1.18)
*p < 0.05, **p < 0.01, ***p < 0.001
aOdds ratios not calculated for enacted HIV stigma group due to small cell sizes
Discussion
This study explored both internalized and enacted intersectional HIV and drug use stigma among Ukrainian WLWH and their impact on HIV care engagement. The analytic approach taken in this paper sought to understand the effect of experiencing a single type of stigma (HIV or drug use) versus multiple stigmas (HIV and drug use) and different stigma mechanisms (internalized and enacted) on HIV care engagement. This analytic strategy revealed that intersectional drug use and HIV stigma was associated with being not engaged in care through two distinct mechanisms (internalized and enacted stigma). When adjusting for other variables, enacted and internalized stigma for both HIV and drug use was negatively associated with HIV care engagement. The finding that intersectional enacted and internalized stigma remained significant predictors of HIV care engagement after adjusting for age contrasts with prior studies that found that age does not predict enacted stigma [53] while older age and longer time since HIV diagnosis may be associated with lower internalized stigma [54]. These findings mark an important advancement in understanding how intersectional stigma affects health outcomes by demonstrating that interpersonal acts of discrimination, rather than feelings of shame or worthlessness alone, are associated with lower care engagement.
Importantly, in the current study, enacted stigma for both HIV and drug use was assessed within participants’ social networks, rather than through generic measures of stigma or specifically in healthcare settings. Many prior studies identify stigmatizing experiences (enacted stigma) within healthcare settings (e.g., discriminating behaviors, loss of confidentiality, negative attitudes of healthcare workers) as possible ways in which HIV (or related) stigma discourages PLWH from seeking or remaining in care [55, 56]. The finding that enacted drug use and HIV stigma negatively affected HIV care engagement suggests that stigma reduction interventions must address attitudes and behaviors within participants’ social networks rather than among healthcare providers alone. This study’s findings also address that multiple stigma mechanisms (enacted and internalized) should be accounted for in interventions that aim to reduce stigma and increase HIV care engagement.
These findings indicate that interventions to address stigma and its role in HIV care engagement must address women’s experiences of both HIV and drug use stigma, which are often rooted in gender stereotypes and expectations around motherhood and gender identity [13, 57, 58]. In a qualitative study of WLHW in the United States, Rice et al. [59] found that participants experienced multiple forms of stigma, including racism, economic discrimination, and sexism, in addition to HIV-related stigma. Any single dimension of stigma could not be disentangled from others and negative connotations of specific aspects of their identity (e.g., being Black, female, and HIV-positive) were reinforced in their interactions with family, community members, and potential employers (e.g., [60]. While teaching coping skills and increasing resilience to address internalized stigma is important [61, 62], coping skills alone will likely be insufficient to address the stigma-related barriers to care that WLHW and histories of drug use encounter based on negative associations around drug use and HIV among women, particularly mothers. In this study, having children was associated with higher odds of HIV care engagement. Studies have document that among WLWH motherhood is an important part of identity, self-worth, personal joy, and motivation [63]. Attending to women’s roles as mothers HIV care engagement in interventions in antenatal care settings [64] and across the life course (outside the period of pregnancy and infancy) should be incorporated into stigma reduction and HIV care engagement interventions as well.
This study was conducted in 2020 and 2021, prior to the Russian attack on Ukraine in 2022. When Russia invaded Ukraine in 2014, resulting in the annexation of Crimea and a sustained military conflict in the Donbas region, essential services for PWID and PLWH were severely disrupted [65, 66] and people who remained in Russian-controlled territories were subjected to Russian’s harmful public health policies, including being cut off from medications for opioid use disorder [67, 68]. The outcome of the 2022 Russian invasion of Ukraine is unknown, but the consequences for marginalized populations such as the women who participated in this study could be severe. During periods of severe and active conflict, if HIV medication supplies are not maintained, people are unable to visit health care facilities, or HIV care programs are deprioritized in favor of other immediate concerns (shelter, security, food), then people risk falling out of care and skipping HIV medications [69, 70]. While short periods of treatment interruptions may not lead to resistance and HIV treatment can be successfully implemented in conflict settings [71], prolonged treatment gaps and disengagement from HIV care can lead to poor health outcomes [72, 73]. In contrast to the 2014 conflict, which affected a limited geographic region of Ukraine, the current war has had profound disruption throughout the country. The women in this study were particularly marginalized and may not have the economic or social resources to migrate out of the conflict zone and access HIV care in new location (within Ukraine or internationally). The instability and uncertainty caused by the war, compounded by the social and health impacts of the COVID-19 pandemic, are likely to have detrimental effects for HIV care engagement for WLWH in Ukraine [74, 75].
This study had several limitations. The analysis did not include other types of stigma, such as prior incarceration [76, 77] or sex work [78, 79], which other studies indicate may also affect HIV care engagement and other health outcomes. In addition, considerable attention has been paid to the ways in which intersectionality is measured both qualitatively [33, 80] and quantitatively [81, 82, 83]. Limitations in quantitative assessment of intersectionality and health outcomes include the narrow operationalizations of intersectional groups (e.g., limiting analysis to binary measures of social statuses) and inability to capture the cumulative effects of discrimination and disadvantage across the life course [82]. Similarly, the observational nature of the study precludes causal inference. In addition, while this paper focused on understanding the effect of intersectional and individual stigma mechanisms, the sample size in this study was not sufficiently powered to analyze an interaction effect between the two stigma mechanisms. Finally, although an effort was made to recruit participants from a range of community-based settings and not exclusively at sites where they receive HIV care, women who were in HIV care were overrepresented in the study population. The results may underestimate the true association between stigma and HIV care status given that women who felt the most stigmatized may also be the most hidden and disconnected from service providers.
Conclusion
The accumulation of scientific evidence indicates that stigma undermines HIV care engagement among diverse populations and across settings. Structural factors such as forced disclosure of HIV status, repressive drug policies, complex bureaucracies, and differential treatment of women who use drugs remain significant barriers to HIV care engagement. The findings of this paper indicate that it is also important to understand how interpersonal factors—specifically intersectional enacted and internalized stigma around HIV and drug use—can undermine engagement in care. Understanding how stigma operates in a highly marginalized population can be mobilized to create more effective, targeted, and multi-level interventions to reduce stigma and increase HIV care engagement. Service providers and programs directed at WLWH with histories of drug use should be tailored to better engage women who experience stigma across multiple settings, including within their social networks and in service provision contexts to keep them engaged in care. Low threshold and confidential services may help engage and retain women in care [84, 85] and serve as an entry point for marginalized women to be linked to other services [86]. At the same time, these services must not re-inscribe or ignore stigmatized identities and should be mobilized to help women to address the forms of stigma they experience in their everyday lives.
Supplementary Information
Below is the link to the electronic supplementary material.Supplementary file1 (DOCX 22 KB)
Acknowledgements
This research was supported by funding from the National Institutes of Health/Fogarty International Center (Grant No. 1R21TW011060-01). We also thank the staff at Club Eney and Convictus for their support of this project and their commitment to serving people living with HIV. We are also grateful to our study participants who generously shared their time and experiences to this project.
Author Contributions
JO: conceptualization, funding acquisition, methodology; writing—original draft preparation, review & editing; SF: writing, review, & editing; CC: formal analysis, writing, review & editing; MD-R: conceptualization, investigation, review & editing; TK: conceptualization, investigation, review & editing; KT: conceptualization, investigation, review & editing.
Funding
This research was supported by funding from the National Institutes of Health/Fogarty International Center (Grant No. 1R21TW011060-01).
Data Availability
Requests for data and instruments described in this study may be requested from the corresponding author.
Code Availability
Not applicable.
Declarations
Conflict of interest
The authors have no financial, consultative, or institutional interests that might lead to bias or conflict of interest.
Ethical Approval
This study was reviewed and approved by the Institutional Review Boards at the Johns Hopkins Bloomberg School of Public Health and the Ukrainian Institute for Public Health Policy.
Consent to Participate
All participants provided oral consent.
Consent for Publication
Not applicable.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
1. Vasylyev M Unified European support framework to sustain the HIV cascade of care for people living with HIV including in displaced populations of war-struck Ukraine Lancet 2022 9 6 E438 E448 10.1016/S2352-3018(22)00125-4
2. Kuzin I, Martzynovska V, Anontenko Z. HIV Infection in Ukraine, Newsletter No. 51. https://phc.org.ua/sites/default/files/users/user90/HIV_in_UA_51_2020_EN.pdf (2020)
3. Kiriazova T Postnov O Bingham T Myers J Flanigan T Vitek C Neduzhko O Patient and provider perspectives inform an intervention to improve linkage to care for HIV patients in Ukraine BMC Health Serv Res 2018 18 1 58 10.1186/s12913-018-2885-4 29378581
4. Vitek CR Čakalo J-I Kruglov YV Dumchev KV Salyuk TO Božičević I Baughman AL Spindler HH Martsynovska VA Kobyshcha YV Abdul-Quader AS Rutherford GW Slowing of the HIV epidemic in Ukraine: evidence from case reporting and key population surveys, 2005–2012 PLoS ONE 2014 9 9 e103657 10.1371/journal.pone.0103657 25251080
5. Sazonova Y Kulchynska R Sereda Y Azarskova M Novak Y Saliuk T Kornilova M Liulchuk M Vitek C Dumchev K HIV treatment cascade among people who inject drugs in Ukraine PLoS ONE 2020 15 12 e0244572 10.1371/journal.pone.0244572 33382768
6. HIV Infection in Ukraine. Newsletter No. 50. (2019). https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&uact=8&ved=2ahUKEwikoPP_z4jeAhUDT98KHfuGD3MQFjAAegQICRAC&url=https%3A%2F%2Fphc.org.ua%2Fuploads%2Fdocuments%2Fc21991%2F8996daa51af04118a9a24efbe09d1c49.pdf&usg=AOvVaw3oEzBv6dZpRaT0MFy
7. Dethier D Rybak N Hirway P Bachmaha M Carroll J Sorokolit A Flanigan T Sluzhynska M The changing face of women living with HIV in western Ukraine Int J STD AIDS 2018 29 4 318 323 10.1177/0956462417724708 28771077
8. Bagkeris E Malyuta R Volokha A Cortina-Borja M Bailey H Townsend CL Thorne C Pregnancy outcomes in HIV-positive women in Ukraine, 2000–12 (European Collaborative Study in EuroCoord): an observational cohort study Lancet HIV 2015 2 9 e385 e392 10.1016/S2352-3018(15)00079-X 26423552
9. Network 100 Percent Life. Людинa vs cтигмa: Як cьoгoднi живyть люди з вiл в yкpaїнi – нoвe дocлiджeння бo «мepeжa». https://network.org.ua/lyudyna-vs-stygma-yak-sogodni-zhyvut-lyudy-z-vil-v-ukrayini-nove-doslidzhennya-bo-merezha/ (2016)
10. Tripathi V King EJ Finnerty E Koshovska-Kostenko N Skipalska H Routine HIV counseling and testing during antenatal care in Ukraine: a qualitative study of the experiences and perspectives of pregnant women and antenatal care providers AIDS Care 2013 25 6 680 685 10.1080/09540121.2012.749332 23317036
11. Rice WS Turan B Fletcher FE Nápoles TM Walcott M Batchelder A Kempf M-C Konkle-Parker DJ Wilson TE Tien PC Wingood GM Neilands TB Johnson MO Weiser SD Turan JM A mixed methods study of anticipated and experienced stigma in health care settings among women living with HIV in the United States AIDS Patient Care STDS 2019 33 4 184 195 10.1089/apc.2018.0282 30932700
12. Demchenko I, Sokolova L, Buliga N, AC Consulting, & CO “100% Life.” The People Living with HIV Stigma Index 2.0: Ukraine. https://www.stigmaindex.org/wp-content/uploads/2021/11/Ukraine-SI-2.0-Report-2020_English.pdf (2020)
13. Carroll J A woman among addicts: the production and management of identities in a Ukrainian harm reduction program Anthropol East Eur Rev 2011 29 1 23 34
14. Owczarzak J Kazi AK Mazhnaya A Alpatova P Zub T Filippova O Phillips SD “You’re nobody without a piece of paper:” visibility, the state, and access to services among women who use drugs in Ukraine Soc Sci Med 2021 269 113563 10.1016/j.socscimed.2020.113563 33309442
15. Owczarzak J Phillips SD Cho W ‘Pure’ drug users, commercial sex workers and ‘ordinary girls’: gendered narratives of HIV risk and prevention in post-Soviet Ukraine Cult Health Sex 2018 10.1080/13691058.2017.1421708 30311861
16. Booth RE Dvoryak S Sung-Joon M Brewster JT Wendt WW Corsi KF Semerik OY Strathdee SA Law Enforcement practices associated with HIV infection among injection drug users in Odessa, Ukraine AIDS Behav 2013 17 8 2604 2614 10.1007/s10461-013-0500-6 23754613
17. Shokoohi M Bauer GR Kaida A Logie CH Lacombe-Duncan A Milloy M-J Lloyd-Smith E Carter A Loutfy M Patterns of social determinants of health associated with drug use among women living with HIV in Canada: a latent class analysis Addiction 2019 114 7 1214 1224 10.1111/add.14566 30698902
18. World Health Organization. Policy guidelines for collaborative TB and HIV services for injecting and other drug users: an integrated approach. (2008)
19. Barnard M Needle sharing in context: patterns of sharing among men and women injectors and HIV risks Addiction 1993 88 805 812 10.1111/j.1360-0443.1993.tb02094.x 8329971
20. Grella C Etheridge RM Joshi V Anglin MD Delivery of HIV risk-reducation services in drug treatment programs J Subst Abuse Treat 2000 19 229 237 10.1016/S0740-5472(00)00102-1 11027892
21. Metsch L Philbin MM Parish C Shiu K Frimpong JA Giang LM HIV testing, care, and treatment among women who use drugs from a global perspective: progress and challenges JAIDS J Acquir Immune Defic Syndr 2015 69 162 168 10.1097/QAI.0000000000000660
22. Ciambrone D Informal networks among women with HIV/AIDS: present support and future prospects Qual Health Res 2002 12 7 876 896 10.1177/104973202129120331 12214676
23. Turan B Hatcher AM Weiser SD Johnson MO Rice WS Turan JM Framing mechanisms linking HIV-related stigma, adherence to treatment, and health outcomes Am J Public Health 2017 107 6 863 869 10.2105/AJPH.2017.303744 28426316
24. Earnshaw VA Chaudoir SR From conceptualizing to measuring HIV stigma: a review of HIV stigma mechanism measures AIDS Behav 2009 13 6 1160 10.1007/s10461-009-9593-3 19636699
25. Hatzenbuehler ML Phelan JC Link BG Stigma as a fundamental cause of population health inequalities Am J Public Health 2013 103 5 813 821 10.2105/AJPH.2012.301069 23488505
26. Link BG Phelan JC Conceptualizing stigma Ann Rev Sociol 2001 27 1 363 385 10.1146/annurev.soc.27.1.363
27. Logie CH Wang Y Lacombe-Duncan A Wagner AC Kaida A Conway T Webster K de Pokomandy A Loutfy MR HIV-related stigma, racial discrimination, and gender discrimination: pathways to physical and mental health-related quality of life among a national cohort of women living with HIV Prev Med 2018 107 36 44 10.1016/j.ypmed.2017.12.018 29277410
28. Earnshaw VA Smith LR Chaudoir SR Amico KR Copenhaver MM HIV stigma mechanisms and well-being among PLWH: a test of the HIV stigma framework AIDS Behav 2013 17 5 1785 1795 10.1007/s10461-013-0437-9 23456594
29. Mak WWS Poon CYM Pun LYK Cheung SF Meta-analysis of stigma and mental health Soc Sci Med 2007 65 2 245 261 10.1016/j.socscimed.2007.03.015 17462800
30. Mahajan AP Sayles JN Patel VA Remien RH Sawires SR Ortiz DJ Szekeres G Coates TJ Stigma in the HIV/AIDS epidemic: a review of the literature and recommendations for the way forward AIDS 2008 22 Suppl 2 S67 S79 10.1097/01.aids.0000327438.13291.62 18641472
31. Gesesew HA Tesfay Gebremedhin A Demissie TD Kerie MW Sudhakar M Mwanri L Significant association between perceived HIV related stigma and late presentation for HIV/AIDS care in low and middle-income countries: a systematic review and meta-analysis PLoS ONE 2017 12 3 e0173928 10.1371/journal.pone.0173928 28358828
32. Rueda S Mitra S Chen S Gogolishvili D Globerman J Chambers L Wilson M Logie CH Shi Q Morassaei S Rourke SB Examining the associations between HIV-related stigma and health outcomes in people living with HIV/AIDS: a series of meta-analyses BMJ Open 2016 6 7 e011453 10.1136/bmjopen-2016-011453 27412106
33. Bowleg L When Black + Lesbian + Woman ≠ Black Lesbian woman: the methodological challenges of qualitative and quantitative intersectionality research Sex Roles 2008 59 5–6 312 325 10.1007/s11199-008-9400-z
34. Crenshaw K Demarginalizing the intersection of race and sex: a black feminist critque of antidiscrimination doctrine, feminist theory and antiracist politics Univ Chic Leg Forum 1989 1989 1 139 167
35. Bowleg L the problem with the phrase women and minorities: intersectionality—an important theoretical framework for public health Am J Public Health 2012 102 7 1267 1273 10.2105/AJPH.2012.300750 22594719
36. Aung S, Hardy N, Chrysanthopoulou SA, Kyaw A, San Tun M, Aung KW, Rana A, Kantor R. Stigma determines antiretroviral adherence in adults with HIV in Myanmar. JAIDS. 2022;89(1):19–26.
37. Vetrova MV Cheng DM Bendiks S Gnatienko N Lloyd-Travaglini C Jiang W Luoma J Blokhina E Krupitsky E Lioznov D Ekstrand ML Raj A Samet JH Lunze K HIV and substance use stigma, intersectional stigma and healthcare among HIV-positive PWID in Russia AIDS Behav 2021 25 9 2815 2826 10.1007/s10461-021-03172-5 33506305
38. Orza L Bewley S Logie CH Crone ET Moroz S Strachan S Vazquez M Welbourn A How does living with HIV impact on women’s mental health? Voices from a global survey J Int AIDS Soc 2015 18 20289 10.7448/IAS.18.6.20289 26643460
39. Sumari-de Boer IM Sprangers MAG Prins JM Nieuwkerk PT HIV stigma and depressive symptoms are related to adherence and virological response to antiretroviral treatment among immigrant and indigenous HIV infected patients AIDS Behav 2012 16 6 1681 1689 10.1007/s10461-011-0112-y 22198315
40. Batchelder AW Foley JD Wirtz MR Mayer K O’Cleirigh C Substance use stigma, avoidance coping, and missed HIV appointments among MSM who use substances AIDS Behav 2021 25 5 1454 1463 10.1007/s10461-020-02982-3 32737816
41. Latkin C Srikrishnan AK Yang C Johnson S Solomon SS Kumar S Celentano DD Solomon S The relationship between drug use stigma and HIV injection risk behaviors among injection drug users in Chennai, India Drug Alcohol Depend 2010 110 3 221 227 10.1016/j.drugalcdep.2010.03.004 20462707
42. Stringer KL Marotta P Baker E Turan B Kempf M-C Drentea P Stepanikova I Turan JM Substance use stigma and antiretroviral therapy adherence among a drug-using population living with HIV AIDS Patient Care STDS 2019 33 6 282 293 10.1089/apc.2018.0311 31166784
43. Brewer R Hood KB Moore M Spieldenner A Daunis C Mukherjee S Smith-Davis M Brown G Bowen B Schneider JA An exploratory study of resilience, HIV-related stigma, and HIV care outcomes among men who have sex with men (MSM) living with HIV in Louisiana AIDS Behav 2020 24 7 2119 2129 10.1007/s10461-020-02778-5 31916097
44. Steward WT Bharat S Ramakrishna J Heylen E Ekstrand ML Stigma is associated with delays in seeking care among HIV-infected people in India J Int Assoc Provid AIDS Care (JIAPAC) 2013 12 2 103 109 10.1177/1545109711432315 22282878
45. Valenzuela C Ugarte-Gil C Paz J Echevarria J Gotuzzo E Vermund SH Kipp AM HIV stigma as a barrier to retention in HIV Care at a General Hospital in Lima, Peru: a case–control study AIDS Behav 2015 19 2 235 245 10.1007/s10461-014-0908-7 25269871
46. Kearney MH Murphy S Rosenbaum M Mothering on crack cocaine: a grounded theory analysis Soc Sci Med 1994 38 2 351 361 10.1016/0277-9536(94)90405-7 8140462
47. Meksin A, Meksin L. Balka: Women, HIV and Drug Use in Ukraine. Open Society Foundations. https://www.youtube.com/watch?v=Z79QGaEqsDM (2011)
48. Pinkham S Malinowska-Sempruch K Women, harm reduction and HIV Reprod Health Matters 2008 16 31 168 181 10.1016/S0968-8080(08)31345-7 18513618
49. Pinkham S Shapoval A Making harm reduction work for women: the Ukrainian experience 2010 New York Open Society Institute, Public Health Program
50. Burgette JM Rankine J Culyba AJ Chu K-H Carley KM Best practices for modeling egocentric social network data and health outcomes HERD 2021 14 4 18 34 10.1177/19375867211013772 33973482
51. Mugavero MJ Amico KR Horn T Thompson MA The state of engagement in HIV care in the United States: from cascade to continuum to control Clin Infect Dis 2013 57 8 1164 1171 10.1093/cid/cit420 23797289
52. Radloff L The CES-D scale: a self-report depression scale for research in the general population Appl Psychol Meas 1977 1 3 385 401 10.1177/014662167700100306
53. Emlet CA Brennan DJ Brennenstuhl S Rueda S Hart TA Rourke SB The impact of HIV-related stigma on older and younger adults living with HIV disease: does age matter? AIDS Care 2015 27 4 520 528 10.1080/09540121.2014.978734 25397643
54. Emlet CA Brennan DJ Brennenstuhl S Rueda S Hart TA Rourke SB Protective and risk factors associated with stigma in a population of older adults living with HIV in Ontario, Canada AIDS Care 2013 25 10 1330 1339 10.1080/09540121.2013.774317 23452022
55. Nyblade L Mingkwan P Stockton MA Stigma reduction: an essential ingredient to ending AIDS by 2030 Lancet HIV 2021 8 2 e106 e113 10.1016/S2352-3018(20)30309-X 33539757
56. Nyblade L Stangl A Weiss E Ashburn K Combating HIV stigma in health care settings: what works? J Int AIDS Soc 2009 12 1 15 15 10.1186/1758-2652-12-15 19660113
57. Owczarzak J Phillips SD Filippova O Alpatova P Mazhnaya A Zub T Aleksanyan R A “common factors” approach to developing culturally tailored HIV prevention interventions Health Educ Behav 2016 43 3 347 357 10.1177/1090198115602665 27178497
58. Paudel V Baral KP Women living with HIV/AIDS (WLHA), battling stigma, discrimination and denial and the role of support groups as a coping strategy: a review of literature Reprod Health 2015 12 1 53 10.1186/s12978-015-0032-9 26032304
59. Rice WS Logie CH Napoles TM Walcott M Batchelder AW Kempf M-C Wingood GM Konkle-Parker DJ Turan B Wilson TE Johnson MO Weiser SD Turan JM Perceptions of intersectional stigma among diverse women living with HIV in the United States Soc Sci Med 2018 208 9 17 10.1016/j.socscimed.2018.05.001 29753137
60. Pretorius JB Greeff M Freeks FE Kruger A A HIV stigma reduction intervention for people living with HIV and their families Health SA Gesondheid 2016 21 187 195 10.1016/j.hsag.2015.11.005
61. Ma PHX Chan ZCY Loke AY Self-stigma reduction interventions for people living with HIV/AIDS and their families: a systematic review AIDS Behav 2019 23 3 707 741 10.1007/s10461-018-2304-1 30298241
62. Rao D Desmond M Andrasik M Rasberry T Lambert N Cohn SE Simoni J Feasibility, acceptability, and preliminary efficacy of the unity workshop: an internalized stigma reduction intervention for African American women living with HIV AIDS Patient Care STDS 2012 26 10 614 620 10.1089/apc.2012.0106 22984780
63. Kennedy VL Serghides L Raboud JM Su D Blitz S Hart TA Walmsley SL Angel JB Smaill FM Ralph ED Tharao WE Loutfy MR The importance of motherhood in HIV-positive women of reproductive age in Ontario, Canada AIDS Care 2014 26 6 777 784 10.1080/09540121.2013.855295 24206065
64. Watt MH Minja L Knettel BA Mwamba RN Osaki H Ngocho JS Kisigo GA Renju J Vissoci JRN Sao SS Mmbaga BT Pilot outcomes of Maisha: an HIV stigma reduction intervention developed for antenatal care in Tanzania AIDS Behav 2021 25 4 1171 1184 10.1007/s10461-020-03093-9 33180253
65. Carroll J. The Quietest Casualties: Russian Public Health Policies Cause Patient Deaths in Crimea. CritCom. http://councilforeuropeanstudies.org/critcom/the-quietest-casualties-russian-public-health-policies-cause-patient-deaths-in-crimea/ (2014)
66. Owczarzak J Karelin M Phillips SD A View from the frontlines in Slavyansk, Ukraine HIV prevention, drug treatment, and help for people who use drugs in a conflict zone Int J Drug Policy 2014 10.1016/j.drugpo.2014.11.009 25512117
67. Holt E Fears over future of opioid substitution therapy in Crimea Lancet 2014 383 9923 1113 10.1016/S0140-6736(14)60234-8 24693545
68. Kazatchkine M Russia’s ban on methadone for drug users in Crimea will worsen the HIV/AIDS epidemic and risk public health BMJ 2014 348 g3118 g3118 10.1136/bmj.g3118 24812004
69. Holt E Russia’s invasion of Ukraine threatens HIV response Lancet HIV 2022 10.1016/S2352-3018(22)00064-9 36088914
70. Sheather J As Russian troops cross into Ukraine, we need to remind ourselves of the impact of war on health BMJ 2022 10.1136/bmj.o499 35217578
71. O’Brien DP Venis S Greig J Shanks L Ellman T Sabapathy K Frigati L Mills C Provision of antiretroviral treatment in conflict settings: the experience of Médecins Sans Frontières Confl Heal 2010 4 1 12 10.1186/1752-1505-4-12
72. Culbert H Tu D O’Brien DP Ellman T Mills C Ford N Amisi T Chan K Venis S HIV treatment in a conflict setting: outcomes and experiences from Bukavu, Democratic Republic of the Congo PLoS Med 2007 4 5 e129 10.1371/journal.pmed.0040129 17535100
73. Ellman T Culbert H Torres-Feced V Treatment of AIDS in conflict-affected settings: a failure of imagination Lancet 2005 365 9456 278 280 10.1016/S0140-6736(05)17802-7 15664207
74. Chamla DD Olu O Wanyana J Natseri N Mukooyo E Okware S Alisalad A George M Geographical information system and access to HIV testing, treatment and prevention of mother-to-child transmission in conflict affected Northern Uganda Confl Health 2007 1 1 12 10.1186/1752-1505-1-12 18053189
75. Shannon K Kaida A Rachlis B Lloyd-Smith E Gray G Strathdee SA Reconsidering the impact of conflict on HIV infection among women in the era of antiretroviral treatment scale-up in sub-Saharan Africa: a gender lens AIDS 2008 22 14 1705 1707 10.1097/QAD.0b013e328308de0e 18753938
76. Chimoyi L Hoffmann CJ Hausler H Ndini P Rabothata I Daniels-Felix D Olivier AJ Fielding K Charalambous S Chetty-Makkan CM HIV-related stigma and uptake of antiretroviral treatment among incarcerated individuals living with HIV/AIDS in South African correctional settings: a mixed methods analysis PLoS ONE 2021 16 7 e0254975 10.1371/journal.pone.0254975 34329311
77. Haley DF Golin CE Farel CE Wohl DA Scheyett AM Garrett JJ Rosen DL Parker SD Multilevel challenges to engagement in HIV care after prison release: a theory-informed qualitative study comparing prisoners’ perspectives before and after community reentry BMC Public Health 2014 14 1 1253 10.1186/1471-2458-14-1253 25491946
78. Hargreaves JR Busza J Mushati P Fearon E Cowan FM Overlapping HIV and sex-work stigma among female sex workers recruited to 14 respondent-driven sampling surveys across Zimbabwe, 2013 AIDS Care 2017 29 6 675 685 10.1080/09540121.2016.1268673 27998178
79. Kerrigan D Karver TS Barrington C Davis W Donastorg Y Perez M Gomez H Mbwambo J Likindikoki S Shembilu C Mantsios A Beckham SW Galai N Chan KS Development of the experiences of sex work stigma scale using item response theory: implications for research on the social determinants of HIV AIDS Behav 2021 25 S2 175 188 10.1007/s10461-021-03211-1 33730252
80. Abrams JA Tabaac A Jung S Else-Quest NM Considerations for employing intersectionality in qualitative health research Soc Sci Med 2020 258 113138 10.1016/j.socscimed.2020.113138 32574889
81. Bauer GR Scheim AI Methods for analytic intercategorical intersectionality in quantitative research: discrimination as a mediator of health inequalities Soc Sci Med 2019 226 236 245 10.1016/j.socscimed.2018.12.015 30674435
82. Harari L Lee C Intersectionality in quantitative health disparities research: a systematic review of challenges and limitations in empirical studies Soc Sci Med 2021 277 113876 10.1016/j.socscimed.2021.113876 33866085
83. Kalichman SC Shkembi B Eaton LA Finding the right angle: a geometric approach to measuring intersectional HIV stigma AIDS Behav 2022 26 S1 27 38 10.1007/s10461-021-03437-z 34424389
84. Edland-Gryt M Skatvedt AH Thresholds in a low-threshold setting: an empirical study of barriers in a centre for people with drug problems and mental health disorders Int J Drug Policy 2013 24 3 257 264 10.1016/j.drugpo.2012.08.002 23036653
85. Mofizul Islam M Topp L Conigrave KM Day CA Defining a service for people who use drugs as ‘low-threshold’: what should be the criteria? Int J Drug Policy 2013 24 3 220 222 10.1016/j.drugpo.2013.03.005 23567101
86. Kim SR Goldenberg SM Duff P Nguyen P Gibson K Shannon K Uptake of a women-only, sex-work-specific drop-in center and links with sexual and reproductive health care for sex workers Int J Gynecol Obstet 2015 128 3 201 205 10.1016/j.ijgo.2014.09.026
| 36441406 | PMC9703403 | NO-CC CODE | 2022-11-29 23:21:42 | no | AIDS Behav. 2022 Nov 28;:1-12 | utf-8 | AIDS Behav | 2,022 | 10.1007/s10461-022-03925-w | oa_other |
==== Front
Comp Clin Path
Comp Clin Path
Comparative Clinical Pathology
1618-5641
1618-565X
Springer London London
3417
10.1007/s00580-022-03417-2
Original Article
The prognostic value of procalcitonin in critically ill cases of systematic inflammatory response syndrome in dogs
Chadorneshin Javad Rahnama 1
http://orcid.org/0000-0001-7363-0300
Khaksar Ehsan [email protected]
2
Sharif Maysam Tehrani 2
Jahandideh Alireza 1
1 grid.411463.5 0000 0001 0706 2472 Department of Clinical Science, Science and Research Branch, Islamic Azad University, Tehran, Iran
2 grid.449232.a 0000 0004 0494 0390 Department of Clinical Science, Garmsar Branch, Islamic Azad University, Garmsar, Iran
28 11 2022
17
4 7 2022
11 11 2022
© The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2022, Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Using markers for early diagnosis can help to reduce mortality and morbidity in systemic inflammatory response syndrome (SIRS). This study investigates the role of procalcitonin (PCT) as a prognostic value in dogs with SIRS in the intensive care unit. Fifty-five dogs were selected and studied. Blood samples were collected and investigated for PCT, white and red blood cells, iron, creatinine, platelet, glucose, albumin, urea, interleukin-6 (IL-6), tumor necrosis factor-α (TNF-α), bandcell, body temperature, and hospitalized days and costs. The animals were grouped into survivors and deceased groups, and their results were compared. The results showed negative significant relations between PCT with hematocrit (r2 = 0.294, P < 0.05) and the serum concentration of iron (r2 = 0.280, P < 0.05) and also positive relation with IL-6 (r2 = 0.456, P < 0.01) and TNF-α (r2 = 0.391, P < 0.01). Significant relations were not seen between PCT with other parameters (P > 0.05). The results also showed a significant relation between glucose and albumin with body temperature (P < 0.05). The results showed that the serum concentrations of PCT, IL-6, and TNF-α were significantly higher in deceased dogs compared with survivors, while white blood cells, glucose, albumin, urea, lymphocyte, neutrophil, and body temperature were higher in survivors compared with others. PCT can be utilized as a prognostic value and helps early diagnosis in dogs with SIRS.
Keywords
Early diagnosis
Procalcitonin
Prognostic values
SIRS
==== Body
pmcIntroduction
Sepsis is known as one of the most common causes of morbidity and mortality in intensive care and is presented in different severities of systemic inflammatory response syndrome (SIRS) (Goggs and Letendre 2019). SIRS is an exaggerated defense response of the body against stressor factors, such as infection, trauma, and surgery for removing the endogenous and exogenous sources of the insult (Chakraborty and Burns 2019). The disease is known with signs such as abnormal body temperature, heart and respiratory rate, and leukocyte count (Pierini et al. 2019). The hematopoietic system and hematological disorders are also seen in patients with SIRS (Napolitano 2017). The disease causes the increase in cytokines in the blood plasma by 2–6 times (Yelins’Ka et al. 2019). In dogs with SIRS, criteria are including body temperature < 37.8 °C or > 39.4 °C, heart rate > 140 bpm, respiratory rate > 30 breaths/min or PCO2 < 32 mmHg, white blood cells < 6000 or > 16,000 cells/μL, or > 3% band neutrophils (Alves et al. 2020).
Clinicians have sought new tools to diagnose sepsis promptly. Early diagnosis and treatment may reduce mortality and morbidity (Rowe et al. 2018). A marker can help early detection and also distinguish between infectious and non-infectious causes of SIRS (Binnie et al. 2020). Traditional markers cannot be specific, such as heart and respiratory rates and white blood cell (WBC) count (Abedini et al. 2012). In addition, physicians are still faced with challenges in the correct use of antibiotics. The incorrect application of antibiotics not only increases mortalities but also leads to antibiotic-resistant, toxic side effects and healthcare costs (Gregoriano et al. 2020). Among biomarkers, procalcitonin is known to be a precise indicator of bacterial infection and/or severity of infection, and also good control of the success of a therapeutic procedure (Battaglia et al. 2021). It has been reported 79.00%, 84.00%, 69.00%, and 77.00% for sensitivity, specificity, sensitivity, and specificity (Yu et al. 2010).
Procalcitonin is an inactive propeptide of calcitonin released by C cells of the thyroid gland, hepatocytes, and peripheral monocytes. It has been reported as a useful biochemical marker to distinguish sepsis from other non-infectious causes of systemic inflammation (Rhodes et al. 2017). Indeed, pro-inflammatory cytokines stimulate the expression of genes responsible for the production of procalcitonin (Downes et al. 2020). Procalcitonin value is related to the severity of illness, and its changes are linked with severity of infection in patients (Tsui et al. 2021). Production of procalcitonin is increased 24 h after induction of infection. An appropriate treatment decreases its production, while incorrect treatment leads to future increases (Paudel et al. 2020). The serum concentration of procalcitonin is very low in healthy people while its concentration is higher in patients with infection (> 100 ng/mL) (Dever and Sheikh 2015).
Despite the role of procalcitonin as a value in human studies, it has not been still investigated as a prognostic value in dogs with SIRS. This new study works on the serum concentration of procalcitonin as a prognostic value in dogs with SIRS in the intensive care unit.
Materials and methods
Animals and protocols
This study was conducted in infectious and intensive care of a veterinary hospital for 7 months during 2020–2021 years. All the dogs with SIRS were studied. Animals with tachycardia, tachypnea, leukopenia, fewer, and/or hypothermia were studied. Clinical examinations of dogs were conducted as reported by previous studies (Giunti et al. 2017). In the current study, all the efforts were conducted to minimize stress. All the experimental procedures were in agreement with protocols advised by the Ethical Committee of Islamic Azad University, Sciences and Research Branch (IR.IAU.SRB.REC.1400.048). All dogs with trauma, surgery, burning, and other diseases interfering with this study were excluded. Finally, 55 dogs were selected and studied. Vital signs were daily recorded.
Blood sampling
Blood samples were collected from all the dogs and transferred in two tubes with and without anticoagulant (Marschner et al. 2012). A complete blood count (CBC) was daily performed to monitor dogs. To evaluate serum samples, the blood samples were centrifuged at 3000 rpm for 10 min. Sera were stored at−80 °C.
The measurement of biochemical parameters
To assess procalcitonin, specific kits (Procalcitonin Kryptor Sensitive B.R.A.H.M.S) were used as recommended by producer companies. To measure IL-6 and tumor necrosis factor-α (TNF-α), Bender Medsystem Kits were used and the results were reported as pg/mL. The serum concentration of glucose was assessed by glucose-hexokinase method using spectrophotometer (Biowave, S2100, England). The serum concentration of urease was assessed by Pars Azmoon Kit (UREA Berthelot, 18,940) as recommended by producer company. Creatinine was evaluated using Pars Azmoon Kit (UREA Berthelot, 18,940) based on recommendations of producer company. The serum concentration of iron was also assessed by Pars Azmoon Kit (BT-2000, 18,923) as recommended by producer company.
The measurement of CBC
The blood samples were analyzed for white and red blood cells, hematocrit, lymphocyte, monocytes, neutrophils, lymphocytes, and basophiles, as suggested by previous studies (Nazerian et al. 2013). At the end of the study, the animals were grouped and compared into survivors and deceased groups. Hospitalized days and treatment costs were also evaluated.
Data analysis
The blood samples were collected from 55 dogs, and sera were obtained from all the collected samples. The data were investigated for normality by Kolmogorov–Smirnov test in SPSS software (version 23). All the data were normalized and analyzed by t test. Pearson correlation was used to investigate the relationship between variables.
Results
The relation between parameters
Table 1 depicts the results for the relation between parameters. The results showed negative significant relations between procalcitonin with hematocrit (r2 = 0.294, P < 0.05) and the serum concentration of iron (r2 = 0.280, P < 0.05) and also positive relation with IL-6 (r2 = 0.456, P < 0.01) and TNF-α (r2 = 0.391, P < 0.01). Significant relations were not seen between procalcitonin with other parameters (P > 0.05). There were positive relations between hematocrit with red blood cells (r2 = 0.846, P < 0.001), hemoglobin (r2 = 0.959, P < 0.001), iron (r2 = 0.791, P < 0.001), and creatinine (r2 = 0.281, P < 0.05). The results also showed significant relations between white blood cells with glucose (r2 = 0.466, P < 0.01), albumin (r2 = 0.589, P < 0.01), urea (r2 = 0.322, P < 0.05), neutrophil (r2 = 0.904, P < 0.001), eosinophil (r2 = 0.363, P < 0.05), monocyte (r2 = 0.983, P < 0.001), IL-6 (r2 = 0.486, P < 0.001), and TNF-α (r2 = 0.409, P < 0.001). There were significant relations between stress indexes of glucose and albumin with body temperature. Other significant relations are shown in Table 1. Table 1 Correlation between parameters
Pro-C Hct WBC RBC Hb Fe Crt Plt Glu Alb
Pro-C
Hct −0.294*
WBC −0.228 0.260
RBC −0.159 0.846*** 0.130
Hb 0.239 0.959*** 0.165 0.863**
Fe −0.280* 0.791*** 0.173 0.769*** 0.768***
Crt 0.051 0.281* 0.109 0.201 0.0301 0.193
Plt 0.180 −0.083 0.215 −0.026 −0.067 −0.087 0.009
Glu 0.058 0.070 0.466** 0.047 0.025 0.087 0.210 0.017
Alb 0.196 0.114 0.589** 0.014 0.088 −0.061 0.209 -0.034 0.642**
Urea Lymph Neut Eos Mon IL-6 TNF-α Bandcell Tem Day Cost
Pro-C 0.102 0.108 −0.146 0.102 −0.081 0.456** 0.391** 0.065 −0.005 0.083 0.055
Hct 0.013 0.099 0.241 −0.251 0.167 −0.011 0.025 0.178 0.051 0.095 0.058
WBC 0.322* 0.038 0.904** 0.363* 0.683** 0.486** 0.409** 0.304* 0.556** 0.310* 0.245
RBC 0.026 0.224 0.216 −0.260 0.126 0.038 0.052 −0.051 0.054 0.090 −0.051
Hb 0.010 0.110 0.158 −0.349* −0.030 0.052 0.041 0.126 0.025 0.069 0.010
Fe −0.116 −0.189 0.185 −0.192 −0.027 0.062 0.055 0.184 0.028 0.000 0.009
Crt 0.165 0.019 0.017 0.270 −0.275 0.032 0.047 0.118 0.324** 0.103 0.115
Plt 0.019 0.138 0.307* 0.412* 0.005 0.076 0.055 −0.256 0.060 0.180 0.175
Glu 0.375** 0.474** 0.501** 0.421* 0.357* −0.303* −0.296 0.348 0.837** 0.082 0.145
Alb 0.455** 0.551** 0.573** 0.430* 0.532* −0.527* 0.402* 0.465* 0.733** 0.112 0.128
Urea − 0.412** 0.327* 0.236 0.140 −0.376* −0.334* 0.043 0.410** 0.007 0.027
Lymph −0.173 0.355* −0.350* −0.264 −0.112 −0.098 0.607** −0.191 −0.227
Neut 0.264 0.786** −0.518** −0.415** 0.291 0.521** 0.474** 0.425**
Eos −0.455** −0.322* 0.061 −0.361* 0.469* 0.018 0.018
Mon −0.589* −0.512* 0.353* 0.429* 0.475 0.362
IL-6 0.512** −0.371* −0.447** −0.321* −0.203
TNF-α −0.323* −0.441** −0.358* −0.195
Bandcell 0.554** 0.019 0.061
Tem −0.154 −0.219
Day 0.961
Cost
Pro-C Procalcitonin, Hct Hematocrit, WBC White bBlood Cells, RBC Red Blood Cells, Hb Hemoglobin, Fe Iron, Crt Creatinine, Plt Platelet, Glu Glucose, Alb Albumin, Lymph Lymphocyte, Neut Neutrophil, Eos Eosinophil, Mon Monocyte, IL-6 Interleukin-6, TNF-α Tumor necrosis factor-α, Tem Temperature
Superscripts *, **, and *** show significant relationships at P < 0.05; P < 0.001 and P < 0.0001 respectively
The concentrations of parameters in survivors and deceased groups
Table 2 shows the results for the concentrations of blood parameters and other parameters in survivors and deceased groups. The results showed significant differences between survivors and deceased groups for procalcitonin (P = 0.006), white blood cells (P = 0.017), glucose (P = 0.008), albumin (P = 0.0001), urea (P = 0.007), lymphocyte (P = 0.001), neutrophil (P = 0.020), IL-6 (P = 0.001), TNF-α (P = 0.001), and body temperature (P = 0.004). Based on findings, the serum concentrations of procalcitonin, IL-6, and TNF-α were significantly higher in deceased dogs compared with survivors, while white blood cells, glucose, albumin, urea, lymphocyte, neutrophil, and body temperature were higher in survivors compared with others.Table 2 The concentrations of parameters and other characteristics in survivors and deceased groups
Deceased Survivors T-statistic P value
Pro-C 120.57 ± 34.71 89.83 ± 39.30 2.89 0.006
Hct (%) 25.05 ± 7.28 26.26 ± 8.53 −0.530 0.598
WBC (103/µL) 2427.78 ± 1295.15 3331.82 ± 995.90 −2.49 0.017
RBC (106/µL) 4.19 ± 1.02 4.20 ± 1.17 −0.007 0.995
Hb (%) 10.36 ± 2.99 10.49 ± 3.52 −0.140 0.890
Fe 84.35 ± 12.91 82.91 ± 13.34 0.387 0.700
Crt 0.97 ± 0.36 1.01 ± 0.35 −0.416 0.679
Plt (103/µL) 231.90 ± 148.80 265.88 ± 120.91 −0.915 0.364
Glu (mg/dL) 59.53 ± 27.88 81.06 ± 28.02 −2.76 0.008
Alb (mg/dL) 2.33 ± 0.55 3.37 ± 0.96 −5.058 0.0001
Urea 12.28 ± 5.38 19.72 ± 9.15 −3.30 0.007
Lymph 528.75 ± 245.15 1587.79 ± 1008.08 −4.96 0.001
Neut 1477.40 ± 1249.72 2376.25 ± 1086.77 −2.44 0.020
Eos 203.42 ± 129.76 330.87 ± 250.46 −1.70 0.102
Mon 34.70 ± 23.14 54.28 ± 26.80 −1.61 0.128
IL-6 (pg/mL) 896.78 ± 437.14 235.07 ± 122.19 5.52 0.001
TNF-α (ng/mL) 495.23 ± 63.52 245.52 ± 89.23 7.81 0.001
Bandcell 68.86 ± 29.63 61.17 ± 37.74 0.464 0.649
Tem (°C) 32.99 ± 8.18 37.50 ± 2.56 −2.98 0.004
Day 3.70 ± 1.84 4.38 ± 1.76 −1.35 0.181
Cost 1,432,500 ± 879,101 1,632,764 ± 856,510 −0.822 0.415
The data are presented as mean ± SD
Pro-C Procalcitonin, Hct Hematocrit, WBC White Blood Cells, RBC Red Blood Cells, Hb Hemoglobin, Fe Iron, Crt Creatinine, Plt Platelet, Glu Glucose, Alb Albumin, Lymph Lymphocyte, Neut Neutrophil, Eos Eosinophil, Mon Monocyte, IL-6 Interleukin-6, TNF-α Tumor Necrosis factor-α, Tem Temperature
Discussion
The pathogenesis of sepsis and SIRS causes to involve a complex interplay of factors and dysregulation of immunity. Early diagnosis, differentiation, and prognosis of SIRS help human and veterinary medicine with the use of appropriate treatments (Thames et al. 2019). In the current study, procalcitonin was evaluated as a prognosis factor in dogs in ICU. Several studies have reported the increase in procalcitonin in patients with sepsis and SIRS (Mustafić et al. 2018; Patil and Patil 2020; Sharma et al. 2020). The increased bacterial infection leads to increased procalcitonin (Moustafa et al. 2021). However, the serum concentration of procalcitonin has not been evaluated in dogs with SIRS in the intensive care unit.
The results showed an inverse relation between procalcitonin with hematocrit and iron. The results are in agreement with previous studies for the relation between procalcitonin with hematocrit in dogs with pyometra (Ahn et al. 2021). Our findings concur with other studies that reported a negative relationship between procalcitonin with hematocrit and iron in patients with coronary artery abnormalities (Liu et al. 2021). It has been also reported that an increase in serum concentration of procalcitonin concurs with a decrease in hematocrit and iron in patients with COVID-19 (Mertoglu et al. 2021). The increased serum concentration of procalcitonin could be attributed to greater inflammation and bacterial infection in patients in ICU. Iron and hematocrit deficiencies occur during infections (Riaty and Nursyam 2022). The decreased hematocrit and iron could be attributed to repeated hemorrhages and the administration of agents affecting iron and hematocrit.
Based on the findings, procalcitonin can be considered as a prognostic value in dogs with SIRS. There was also a positive relationship between the serum concentrations of IL-6 and TNF-α with procalcitonin. Several studies have reported a significant relations between procalcitonin with IL-6 and TNF-α in rats with mild and severe pancreatitis (Soyalp et al. 2017) and white swine (Chalkias et al. 2021). The increased serum concentrations of pro-inflammatory cytokines are directly associated with the severity and mortality of human sepsis. In fact, the cytokines lead to an increase in the variety of pathologic reactions and hypotension and shock (Chalkias et al. 2021). In fact, IL-6 is the main cytokine responsible to induce the systemic changes and promotes postoperative neurological dysfunction (Chalkias et al. 2021). TNF-α is a pro-inflammatory cytokine that participates in several physiological and pathophysiological processes and may encourage monocyte/macrophage differentiation and tumor cell necrosis/apoptosis (Sedger and McDermott 2014). SIRS is largely mediated by pro-inflammatory cytokines, and a SIRS-like clinical picture can be induced by the administration of pro-inflammatory cytokines (Natanson et al. 1989). IL-6 is known as a good diagnostic and prognostic marker in people with SIRS (Pettilä et al. 2002; Reinhart et al. 2000), and this has also been confirmed in dogs with SIRS (Rau et al. 2007). The results indicate significant relations between procalcitonin with TNF-α and IL-6. In addition, the results showed increase of 1.3 times serum concentrations of procalcitonin in deceased dogs compared with survivors. In fact, the concentration of procalcitonin is significantly higher in dogs with a chronic condition of disease. The increased procalcitonin in deceased dogs means to be appropriate as a prognostic value in dogs with SIRS in ICU.
The results also showed increased serum concentrations of IL-6 and TNF-α in deceased dogs. The findings are in agreement with those reported by Ghazizadeh et al. who showed to be high procalcitonin concentration in the acute phase of sepsis (Ghazizadeh et al. 2021). Other studies have also reported an increase in inflammatory factors in dogs with systematic infection (Gommeren et al. 2018; Kuzi et al. 2020). It has been reported that procalcitonin concentration is a marker for distinguishing infectious episodes from non-infectious episodes (Haeusler et al. 2013). The results are also parallel with those reported by Neumann and others who reported higher serum concentration of procalcitonin in dogs with sepsis in the acute phase compared with control healthy dogs (Neumann 2022). In addition, white blood cells, glucose, albumin, urea, lymphocyte, neutrophil, and body temperature were higher in survivors compared with others. The increase in parameters implicates the presence of infection and fighting the body against disease. Fighting against disease requires an energy source, and the increased glucose is a response to supplying energy for fighting against disease (Suleiman et al. 2022). The increased white blood cells and pro-inflammatory cytokines in survivor dogs highlight greater immunity in them compared with deceased dogs. All the findings show to be appropriate and specific procalcitonin as a prognostic value in dogs with SIRS.
The results did not show significant differences between survivors and non-survivors for hospitalized costs and days, although costs and days were numerically lower in survivors compared with other groups. The decrease in hospitalized days directly reduces the costs for dog owners.
Conclusion
In sum, the serum concentration of procalcitonin was higher in dogs with SIRS and especially in deceased dogs compared with survivors. It shows that pro-calcitonin is a good value for the prognosis of SIRS. In addition, procalcitonin showed a strong and positive correlation with pro-inflammatory cytokines of IL-6 and TNF-α. Based on findings, procalcitonin is a good marker for early diagnosis of SIRS in dogs admitted to ICU. Using procalcitonin helps early diagnosis of SIRS and reduces hospitalized days and costs.
Acknowledgements
Authors would like to appreciate Islamic Azad University, Science and Research Branch for their support.
Compliance with ethical standards
Funding
This study was not supported by any funding.
Conflict of interest
The authors declare that they have no conflict of interest.
Ethical approval
All the experimental procedures were in agreement with protocols advised by Ethical Committee of Islamic Azad University, Sciences and Research Branch (IR.IAU.SRB.REC.1400.048).
Informed consent
For this type of study informed consent is not required.
Consent for publication
For this type of study consent for publication is not required.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
Abedini M Delpisheh A Nikkhu B Vahabi A Afkhamzadeh A Procalcitonin and white blood cell count (WBC), erythrocyte sedimentation rate (ESR) and serum C-reactive protein (CRP) alterations in children with systemic inflammatory response syndrome before and after treatment Afr J Biotechnol 2012 11 10989 10993
Ahn S Bae H Kim J Kim S Park J Kim SK Jung DI Yu D Comparison of clinical and inflammatory parameters in dogs with pyometra before and after ovariohysterectomy Canad J Vet Res 2021 85 271 278 34602731
Alves F Prata S Nunes T Gomes J Aguiar S Aires da Silva F Tavares L Almeida V Gil S Canine parvovirus: a predicting canine model for sepsis BMC Vet Res 2020 16 1 11 10.1186/s12917-020-02417-0 31900161
Battaglia F Baldoneschi V Meucci V Intorre L Minunni M Scarano S Detection of canine and equine procalcitonin for sepsis diagnosis in veterinary clinic by the development of novel MIP-based SPR biosensors Talanta 2021 230 122347 10.1016/j.talanta.2021.122347 33934796
Binnie A Lage J Dos Santos CC How can biomarkers be used to differentiate between infection and non-infectious causes of inflammation? Evidence-Based Prac Crit Care 2020 14 319 335 10.1016/B978-0-323-64068-8.00055-9
Chakraborty RK Burns B Systemic inflammatory response syndrome Talanta 2019 275 12547
Chalkias A Spyropoulos V Georgiou G Laou E Koutsovasilis A Pantazopoulos I Kolonia K Vrakas S Papalois A Demeridou S Baseline values and kinetics of IL-6, procalcitonin, and TNF-α in landrace-large white swine anesthetized with propofol-based total intravenous anesthesia BioMed Res Int 2021 14 39 45
Dever J Sheikh M spontaneous bacterial peritonitis–bacteriology, diagnosis, treatment, risk factors and prevention Aliment Pharmacol Therap 2015 41 1116 1131 10.1111/apt.13172 25819304
Downes KJ Fitzgerald JC Weiss SL Utility of procalcitonin as a biomarker for sepsis in children J Clin Microbiol 2020 58 e01851 e1919 10.1128/JCM.01851-19 32350044
Ghazizadeh F Noroozi M Shekari M Comparison of the relationship between procalcitonin and crp with prognosis of fever and neutropenia in children with acute lymphoblastic leukemia Studies Med Sci 2021 32 335 341
Giunti M Troia R Battilani M Giardino L Dondi F Andreani G Fracassi F Retrospective evaluation of circulating thyroid hormones in critically ill dogs with systemic inflammatory response syndrome J Vet Sci 2017 18 471 477 10.4142/jvs.2017.18.4.471 28057899
Goggs R Letendre J Evaluation of the host cytokine response in dogs with sepsis and noninfectious systemic inflammatory response syndrome J Vet Emerg Crit Care 2019 29 593 603 10.1111/vec.12903
Gommeren K Desmas I Garcia A Bauer N Moritz A Roth J Peeters D Inflammatory cytokine and C-reactive protein concentrations in dogs with systemic inflammatory response syndrome JJ Vet Emerg Crit Care 2018 28 9 19 10.1111/vec.12685
Gregoriano C Heilmann E Molitor A Schuetz P Role of procalcitonin use in the management of sepsis J Thoracic Dis 2020 12 S5 10.21037/jtd.2019.11.63
Haeusler GM Carlesse F Phillips RS An updated systematic review and meta-analysis of the predictive value of serum biomarkers in the assessment of fever during neutropenia in children with cancer Pediatr Infect Dis J 2013 32 e390 e396 10.1097/INF.0b013e31829ae38d 23673421
Kuzi S Mazaki-Tovi M Suchodolski JS Rimer D Lidbury JA Steiner JM Buono A Nivy R Segev G Aroch I Protease inhibitors, inflammatory markers, and their association with outcome in dogs with naturally occurring acute pancreatitis J Vet Inter Med 2020 34 1801 1812 10.1111/jvim.15895
Liu HH Chen WX Niu MM Jiang Q Qiu Z Fan GZ Li RX Mammadov G Wu YF Luo HH A new scoring system for coronary artery abnormalities in Kawasaki disease Pediatr Res 2021 14 1 9
Marschner CB Kristensen AT Spodsberg EH Wiinberg B Evaluation of platelet aggregometry in dogs using the M ultiplate platelet analyzer: impact of anticoagulant choice and assay duration J Vet Emerg Crit Care 2012 22 107 115 10.1111/j.1476-4431.2011.00709.x
Mertoglu C Huyut MT Arslan Y Ceylan Y Coban TA How do routine laboratory tests change in coronavirus disease 2019? Scand J Clin Lab Invest 2021 81 24 33 10.1080/00365513.2020.1855470 33342313
Moustafa R Albouni T Aziz G The role of procalcitonin and presepsin in the septic febrile neutropenia in acute leukemia patients PLoS ONE 2021 16 e0253842 10.1371/journal.pone.0253842 34324506
Mustafić S Brkić S Prnjavorac B Sinanović A Porobić-Jahić H Salkić S Diagnostic and prognostic value of procalcitonin in patients with sepsis Med Glas 2018 15 10 20
Napolitano LM Anemia and red blood cell transfusion: advances in critical care Crit Care Clin 2017 33 345 364 10.1016/j.ccc.2016.12.011 28284299
Natanson C Eichenholz PW Danner RL Eichacker P Hoffman WD Kuo GC Banks S MacVittie T Parrillo J Endotoxin and tumor necrosis factor challenges in dogs simulate the cardiovascular profile of human septic shock J Exper Med 1989 169 823 832 10.1084/jem.169.3.823 2647895
Nazerian S Gholipour KH Jafarian MH Soltani M Esmaeili MA Effect of Dietary Garlic Powder on Some Hematological Indices of Beluga (huso Huso) 2013 10 45 68
Neumann A Rapid release of sepsis markers heparin-binding protein and calprotectin triggered by anaerobic cocci poses an underestimated threat Anaerobe 2022 75 102584 10.1016/j.anaerobe.2022.102584 35545182
Patil HV Patil VC Comparative study of procalcitonin and C-reactive protein in patients with sepsis J Natural Sci Biol Med 2020 11 85 93
Paudel R Dogra P Montgomery-Yates AA Yataco AC Procalcitonin: a promising tool or just another overhyped test? Int J Med Sci 2020 17 332 352 10.7150/ijms.39367 32132868
Pettilä V Hynninen M Takkunen O Kuusela P Valtonen M Predictive value of procalcitonin and interleukin 6 in critically ill patients with suspected sepsis Intens Care Med 2002 28 1220 1225 10.1007/s00134-002-1416-1
Pierini A Gori E Lippi I Ceccherini G Lubas G Marchetti V Neutrophil-to-lymphocyte ratio, nucleated red blood cells and erythrocyte abnormalities in canine systemic inflammatory response syndrome Res Vet Sci 2019 126 150 154 10.1016/j.rvsc.2019.08.028 31493682
Rau S Kohn B Richter C Fenske N Küchenhoff H Hartmann K Härtle S Kaspers B Hirschberger J Plasma interleukin-6 response is predictive for severity and mortality in canine systemic inflammatory response syndrome and sepsis Vet Clin Pathol 2007 36 253 260 10.1111/j.1939-165X.2007.tb00220.x 17806073
Reinhart K Karzai W Meisner M Procalcitonin as a marker of the systemic inflammatory response to infection Intens Care Med 2000 26 1193 10.1007/s001340000624
Rhodes A Evans LE Alhazzani W Levy MM Antonelli M Ferrer R Kumar A Sevransky JE Sprung CL Nunnally ME Surviving sepsis campaign: international guidelines for management of sepsis and septic shock: 2016 Intensive Care Med 2017 43 304 377 10.1007/s00134-017-4683-6 28101605
Riaty Z, Nursyam DE (2022) Analysis of hemoglobin levels and hematocrite value in stunting children in bukit sileh, solok regency. In "1st International Conference on Health Sciences and Biotechnology (ICHB 2021)", pp. 156–159. Atlantis Press
Rowe CR Newberry DM Jnah AJ Congenital syphilis: a discussion of epidemiology, diagnosis, management, and nurses' role in early identification and treatment Adv Neonatal Care 2018 18 438 445 10.1097/ANC.0000000000000534 30020089
Sedger LM McDermott M TNF and TNF-receptors: from mediators of cell death and inflammation to therapeutic giants–past, present and future Cytokine Growth Factor Rev 2014 25 453 472 10.1016/j.cytogfr.2014.07.016 25169849
Sharma A Ray S Mamidipalli R Kakar A Chugh P Jain R Ghalaut MS Choudhury S A comparative study of the diagnostic and prognostic utility of soluble urokinase-type plasminogen activator receptor and procalcitonin in patients with sepsis and systemic inflammation response syndrome Ind J Crit Care Med 2020 24 245 10.5005/jp-journals-10071-23385
Soyalp M Yalcin M Oter V Ozgonul A Investigation of procalcitonin, IL-6, oxidative stress index (OSI) plasma and tissue levels in experimental mild and severe pancreatitis in rats Bratisl Lek Listy 2017 118 137 141 28319407
Suleiman MA Umaru T Dauda K John SR Usman MA Hyponatraemia and hypokalaemia relationship with alterations of glucose, cholesterol and total protein levels during human infection with Plasmodium falciparum Compar Clin Pathol 2022 31 557 563 10.1007/s00580-022-03354-0
Thames BE Barr JW Suchodolski JS Steiner JM Heilmann RM Prospective evaluation of S100A12 and S100A8/A9 (calprotectin) in dogs with sepsis or the systemic inflammatory response syndrome J Vet Diagn Invest 2019 31 645 651 10.1177/1040638719856655 31170888
Tsui T-L Huang Y-T Kan W-C Huang M-S Lai M-Y Ueng K-C Shiao C-C A novel procalcitonin-based score for detecting sepsis among critically ill patients PLoS ONE 2021 16 e0245748 10.1371/journal.pone.0245748 33481913
Yelins’K, A, Akimov OY, Kostenko V, Role of AP-1 transcriptional factor in development of oxidative and nitrosative stress in periodontal tissues during systemic inflammatory response Ukrain Biochem J 2019 10 80 85
Yu Z Liu J Sun Q Qiu Y Han S Guo X The accuracy of the procalcitonin test for the diagnosis of neonatal sepsis: a meta-analysis Scand J Infect Dis 2010 42 723 733 10.3109/00365548.2010.489906 20840003
| 36466191 | PMC9703405 | NO-CC CODE | 2022-11-29 23:21:08 | no | Comp Clin Path. 2022 Nov 28;:1-7 | utf-8 | Comp Clin Path | 2,022 | 10.1007/s00580-022-03417-2 | oa_other |
==== Front
Wien Klin Wochenschr
Wien Klin Wochenschr
Wiener Klinische Wochenschrift
0043-5325
1613-7671
Springer Vienna Vienna
36441338
2111
10.1007/s00508-022-02111-1
Original Article
Impact of noninvasive ventilation at a municipal emergency department on ICU admissions
http://orcid.org/0000-0002-2497-1544
Abulesz Yannic-Tomas [email protected]
http://orcid.org/0000-0002-2953-831X
Haugk Moritz MBA [email protected]
Department of Emergency Medicine, Klinik Hietzing, Wolkersbergenstraße 1, 1130 Vienna, Austria
28 11 2022
17
17 10 2021
17 10 2022
© The Author(s), under exclusive licence to Springer-Verlag GmbH Austria, part of Springer Nature 2022
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Background
In 2015, the emergency department of a municipal hospital in Vienna began to perform noninvasive ventilation (NIV) on patients admitted for acute respiratory failure, given no intubation criteria were met. The intention of this study was to show to which type of hospital unit patients were transferred after undergoing NIV in the emergency department. Additionally, the impact of the underlying disease, a patient’s sex and age and the year of intervention were analyzed.
Methods
A single-center retrospective exploratory study was performed on 371 patients. All patients with acute respiratory failure who were noninvasively ventilated at the study center emergency department from 2015 to 2018 were eligible. Relevant data were extracted from the patient’s medical records.
Results
A total of 43.7% (95% confidence interval, CI 38.8–48.5%) of patients were successfully stabilized in the emergency department through NIV and subsequently transferred to a normal care unit or discharged. This nonintensive care admission rate was significantly associated with certain underlying medical conditions, age and year of intervention. A further 19.7% (95% CI 15.6–23.7%) of patients were transferred to an intermediate care unit instead of an intensive care unit.
Conclusion
These findings emphasize the importance of noninvasive ventilation at the emergency department in reducing load on intensive care units and ensuring an efficient hospital workflow. Nonintensive care admission rate appears to be the highest in patients with pulmonary edema, especially in the higher age range and is also associated with the level of staff training. Prospective trials are needed to accurately confirm these correlations.
Keywords
Emergency medicine
Acute respiratory failure
Critical care
COPD
==== Body
pmcIntroduction
Noninvasive ventilation (NIV) is nowadays frequently implemented in the routine of emergency departments, with a 2009 study at 132 academic hospital emergency departments across the USA reporting 64% of physicians to be very familiar with it, with 41% of sites being able to initiate NIV in less than 10 min [1]. Numerous studies have shown the efficacy of noninvasive ventilation in relieving acute respiratory distress in different conditions, such as chronical obstructive pulmonary disease (COPD) [2] or acute cardiogenic pulmonary edema [3], such as a reduction in mortality and a decreased rate of intubation; however, research on NIV specifically in the context of short-term treatment at an emergency department remains scarce.
In a 2009 trial including 198 patients, the initial application of NIV at an emergency department in patients with acute respiratory failure showed a significant reduction of subsequent in-hospital mortality and length of Intensive-Care-Unit or Intermediate-Care-Unit (ICU/IMC) stay compared to a control group without initial NIV treatment [4]. It was also demonstrated in another study that the average time it takes from the admission of a patient in respiratory failure to the application of NIV is significantly lower (56 min) when implemented at an emergency department than at an intensive care unit (84 min) [5].
To the authors’ knowledge, no study has yet addressed the impact of NIV at an emergency department on admissions to the intensive care unit. In a multitude of countries, intensive care units regularly operate on the limit capacity-wise [6, 7]. Sometimes, ICU bed availability can even influence decisions on ICU admissions of critical care patients [8]. The pending COVID-19 pandemic aggravated this problem further, with guidelines emerging on how to deal with acute shortages of intensive care beds [9]. Additionally, cost per bed per day is substantially higher in intensive care units than in normal care units [10].
The emergency department at the study center, a municipal hospital in Vienna, implemented NIV in the beginning of 2015. Formerly, all arriving patients in acute respiratory failure were admitted directly to the intensive care unit. Now, they are treated with NIV and corresponding medication in the emergency department for several hours. Once a thorough work-up on the patient is done, a decision is made where the patient should be transferred to. This can either be an intensive care unit, an intermediate care unit or a normal care unit in the same hospital. If a patient’s condition improves significantly, there is also the option of an overnight stay directly in the emergency department and subsequent discharge from there. The decision where to transfer a patient to is based on the cardiorespiratory situation of the patient as assessed by interpretation of vital parameters and blood gas analysis by the emergency physician on duty.
The primary outcome of this study was to show to which type of hospital unit the patients were transferred after receiving NIV in the emergency department. As secondary outcomes, we stratified this distribution by medical condition, age, sex and year of intervention. This enables discussion on whether a certain condition, a certain age range or a certain sex responds better to NIV in the emergency department than another. We also deemed the year of intervention interesting as in the later years of the study’s timeframe, the staff became more experienced with the procedure, supposedly raising the quality of NIV application.
Patients, material and methods
Patient recruitment
A single-center retrospective exploratory observational study was performed on 371 patients.
Eligible for this study were all patients admitted to the emergency department from 1 January 2015, through 30 November 2018 by ambulance or self-admission (around 120,000 patients).
An inclusion criterion was the application of NIV in the emergency department. NIV was applied to all patients in acute respiratory failure (defined by either PaO2 < 60 mm Hg or PaCO2 > 50 mm Hg or both in arterial blood gas analysis at the time of admission), given no intubation criteria were met. According to guidelines [11], these are: absence of spontaneous breathing, airway obstruction, gastrointestinal hemorrhage or ileus and non-hypercapnic coma. The second inclusion criterion was the presence of one of four specific admission diagnoses: COPD, pulmonary edema (cardiogenic, hypertensive or combined), pneumonia and asthma. This criterion enables meaningful statistical analysis, but still accounts for the vast majority of NIV indications.
Patients without sufficient documentation were excluded (Fig. 1).Fig. 1 Flowchart demonstrating the process of patient selection
Measurements
Data on the following parameters were collected retrospectively by the study’s first author:Primary outcome (categorical)
The primary outcome indicates the course of the patients after having received NIV at the study center’s emergency department and is comprised of the four categories intensive care, intermediate care, normal care and discharged.
Primary composite outcome (categorical)
For inferential statistical evaluation, it was decided to facilitate the primary outcome parameter from the four categories intensive care, intermediate care, normal care and discharged to two, intensive care admission and nonintensive care admission. Patients who were transferred to an intensive or intermediate care unit were assigned to the intensive care admission subgroup, patients admitted to a normal care unit or discharged to the nonintensive care admission subgroup.
Underlying medical condition (categorical)
The parameter underlying medical condition is defined as the condition that caused the acute respiratory distress which ultimately made noninvasive ventilation necessary. The four categories were acutely exacerbated COPD, pulmonary edema, pneumonia and asthma.
Year of intervention (metrical)
The year of intervention parameter identified the year in which the patient was admitted to the emergency department and received noninvasive ventilation. The possible values were 2015, 2016, 2017 and 2018.
Age and sex
The age parameter referred to the patient’s age at which noninvasive ventilation occurred, scaled metrically. Sex referred to the patient’s sex assigned at birth, scaled categorically.
Each patient who received NIV at the study center’s emergency department was manually recorded in a documentation booklet. After receiving approval and a waiver from informed consent from the study center’s responsible ethics committee (“Ethikkommission der Stadt Wien”—Ethics Committee of the City of Vienna, reference number EK 18-285-VK), the relevant data were non-anonymously extracted from the hospital’s documentation system and filled into anonymous, numbered case report forms. The data were then entered into a newly created database of the statistics program SPSS 23.0 by IBM (IBM, Armonk, NY, USA), where statistical calculations were executed.
Analysis
No proper sample size calculation was performed due to the exploratory character of this study; however, we deemed a patient number of 371 as high enough to allow for statistically significant results in our study setting.
For the baseline values in our study (age, sex, underlying medical condition and year of intervention) the total number for each subgroup as well as their percentual relation were calculated. For age, we additionally calculated the mean age and its standard deviation. The primary outcome and the primary composite outcome were also descriptively analyzed and 95% confidence intervals were calculated via bootstrapping (replications: 2000). Since the cases that were missing data were eliminated prior to statistical analysis, a complete case analysis was performed and no missing data imputations were used.
A bar chart was plotted visualizing the primary outcome grouped by underlying medical condition. (Fig. 2).Fig. 2 Bar chart representing the primary outcome, clustered by underlying medical condition, with 95% confidence intervals. NCU normal care unit, IMCU intermediate care unit, ICU intensive care unit
The secondary outcomes compare different populations, allowing for inferential statistics. Since this study was of exploratory character, no Bonferroni correction was calculated for the significance values, α remains 0.05.
Our aim was to estimate whether certain circumstances like a patient’s age or the underlying medical condition had an influence on subsequent intensive care admission. We therefore employed a logistic regression model with the dependent variable primary composite outcome and the independent variables underlying medical condition, age, sex and year of intervention.
All statistical assumptions for binary logistic regression were met in our sample. First, null and alternative hypotheses were formulated for each correlation (for example: H0 = age cannot estimate intensive care admission; H1 = age can estimate intensive care admission). For independent parameters that were categorical, dummy variables were coded for each category. For underlying medical condition, asthma was chosen as the reference category, for sex male and for year of intervention the last year of the study’s timeframe, 2018. Finally, the calculation was undertaken in the statistics software. To prove the model’s validity, an omnibus test of model coefficients and subsequently the Hosmer-Lemeshow goodness-of-fit test were performed and 95% confidence intervals were calculated for the odds ratios.
Results
Characteristics of the study subjects
All descriptive data can be found accumulated in (Table 1), the values for the logistic regression are given in (Table 2).Table 1 Baseline values and descriptive statistics
Primary outcome
All patients ICU IMCU NCU Discharged
Cases (%) 371 136 (36.7%) 73 (19.7%) 158 (42.6%) 4 (1.1%)
[95% CI] – [31.8–41.2] [15.6–23.7] [37.7–47.7] [0.3–2.2]
Mean age in years [SD] 69.7 [11.8] 66.8 [12.8] 71.0 [9.1] 71.2 [11.6] 66.8 [15.7]
Female (%) 181 (48.8%) 72 (39.8%) 35 (19.3%) 73 (40.3%) 1 (0.6%)
Male (%) 190 (51.2%) 64 (33.7%) 38 (20%) 85 (44.7%) 3 (1.6%)
Medical condition (%)
COPD 214 78 (36.4%) 42 (19.6%) 93 (43.5%) 1 (0.5%)
Pulmonary edema 82 22 (26.8%) 14 (17.1%) 45 (54.9%) 1 (1.2%)
Pneumonia 66 20 (30.3%) 16 (24.2%) 29(43.9%) 1 (1.5%)
Asthma 9 7 (77.8%) 1 (11.1%) 0 (0.0%) 1 (11.1%)
Year of intervention (%)
2015 11 5 (45.5%) 2 (18.2%) 3 (27.3%) 1 (9.1%)
2016 53 28 (52.8%) 10 (18.9%) 15 (28.3%) 0 (0.0%)
2017 148 48 (32.4%) 33 (22.3%) 67 (45.3%) 0 (0.0%)
2018 159 55 (34.6%) 28 (17.6%) 73 (45.9%) 3 (1.9%)
Primary composite outcome
Intensive care admission Nonintensive care admission
Cases (%) 209 (56.3) 162 (43.7)
[95% CI] [51.5–61.2] [38.8–48.5]
Female (%) 181 107 (59.1) 74 (40.9)
Male (%) 190 102 (53.7) 88 (46.3)
Medical condition (%)
COPD 214 120 (56.1%) 94 (43.9%)
Pulmonary edema 82 36 (43.9%) 46 (56.1%)
Pneumonia 66 45 (68.2%) 21 (31.8%)
Asthma 9 8 (88.9%) 1 (11.1%)
Year of intervention (%)
2015 11 7 (63.6%) 4 (36.4%)
2016 53 38 (71.7%) 15 (28.3%)
2017 148 81 (54.7%) 67 (45.3%)
2018 159 83 (52.2%) 76 (47.8%)
Percentage values always to be read in line
ICU intensive care unit, IMCU intermediate care unit, NCU normal care unit, SD standard deviation, CI confidence interval
Table 2 Nonintensive care admission estimation by binary logistic regression
Logistic regression model
χ2-test Degrees of freedom p‑value
Omnibus test – 27.61 8 < 0.01
Hosmer-Lemeshow test – 5.74 8 0.68
% correctly estimated 61.20% – – –
Independent variables Odds ratio 95% CI for odds ratio p‑value
Medical condition
COPD – 5.37 0.65–44.69 0.12
Pulmonary edema – 9.41 1.10–80.53 0.04
Pneumonia – 3.07 0.35–26.80 0.31
Asthma (ref.) – – – 0.01
Year of intervention
2015 – 0.85 0.54–1.34 0.48
2016 – 0.40 0.20–0.80 0.01
2017 – 0.57 0.15–2.12 0.40
2018 (ref.) – – – 0.07
Sex (ref. = male)
Female – 0.76 0.50–1.17 0.21
Age (years) – 1.03 1.00–1.05 0.03
The dependent variable of this logistic regression model was the primary composite outcome. The odds ratio gives the ratio of the odds for the primary composite outcome being 1 (=non-intensive care admission) in a certain category compared to the reference category in the variable while “controlling” for all other variables (while assuming all other variables stay the same). Values are shown to two decimal points
CI confidence interval
A total of 371 patient cases were analyzed. The median age in our patient population was 70 years with a standard deviation of 11.8 years and a range of 21–98 years. Of the patients 51.2% were male, 48.8% female. 3% of patients were treated in the year 2015, 14.3% in 2016, 39.9% in 2017 and 42.8% in 2018.
The underlying medical condition that led to admission the most often was acutely exacerbated COPD with 57.1% of patients, followed by pulmonary edema with 22.1%. Pneumonia accounted for 17.8% of admissions and acute asthma attacks for 2.4%.
Primary outcome
Our data show that 43.7% of patients in the study population were not admitted to intensive care, meaning they improved at the emergency department in such a way that they could be transferred to a normal care unit or even discharged. Normal care admissions accounted for 42.6% and discharges for 1.1%.
Of our patients 56.3% had to be admitted to an intensive care (ICU) or intermediate care unit (IMCU), with intensive care admissions of 36.7% and intermediate care admissions of 19.7%.
Secondary descriptive outcomes
For underlying medical condition, the nonintensive care admission rate was the highest in the pulmonary edema subgroup (56.1%), followed by the COPD (43.9%), pneumonia (31.8%) and finally the asthma subgroup (11.1%). The year of intervention showed higher nonintensive care admission rates in the later years (2017: 45.3; 2018: 47.8%) compared to when NIV was first implemented (2015: 36.4%; 2016: 28.3%). Finally, male patients had a slightly higher nonintensive care admission rate (46.3%) than females (40.9%).
Secondary inferential outcomes
The employed logistic regression model was able to estimate 61.2% of the cases correctly. The omnibus test for model coefficients and the Hosmer-Lemeshow goodness-of-fit test further indicated that the logistic regression model was valid.
Regarding underlying medical condition, the data show that patients with pulmonary edema had a significantly higher estimated nonintensive care admission rate than patients with asthma with an odds ratio of 9.41 (95% CI: 1.1–80.53). Asthma itself was also strongly associated with a low nonintensive care admission rate. Patients treated in the year 2018 had a significantly higher chance of not requiring intensive care than patients treated in 2016 (odds ratio: 2.53, 95% CI 1.26–5.05), with the year 2018 itself associated with NIV success.
A patient’s age was also able to estimate the nonintensive care admission rate with an odds ratio of 1025 (95% CI 1003–1047), suggesting that with each additional year, the odds of NIV success increased by 2.5%. A patient’s sex had no significant influence on the estimated nonintensive care admission rate.
Discussion
The results of this study show that an emergency department with NIV capability on site plays an important role in stabilizing patients in acute respiratory failure and therefore reducing the load on the hospital’s intensive care unit.
In the treatment of these patients, simple oxygen masks are often insufficient, however invasive ventilation always goes along with sedation and subsequent admission to the ICU. Noninvasive ventilation seems to be the golden mean on this spectrum: it offers sufficient ventilation assistance but is also very easy to apply with a simple strap-on mask and no obligatory need for sedation. Every patient can undergo a trial of noninvasive ventilation given no contraindications are met, and if invasive ventilation becomes necessary, the patient can be intubated in the emergency department and transferred to the intensive care unit. If noninvasive ventilation combined with adequate medication is sufficient the patient can avoid intensive care admission and be further treated in normal care.
This event is oftentimes referred to as “NIV success” in other studies. While we do think that most of the patients in our nonintensive care admission subgroup experienced NIV success due to their clinical course, we cannot scientifically attribute it to NIV alone due to the lack of a control group that did not receive NIV.
In our study population, the nonintensive care admission rate in either hypertensive or cardiogenic pulmonary edema was the highest. Numerous studies prove the efficacy of noninvasive ventilation in this condition [3] and guidelines strongly suggest using it as the main treatment [11, 12]. It is therefore a prime example of NIV treatment at an emergency department. While acute respiratory distress is alleviated by ventilation support through NIV, the appropriate medication is given. The medication diminishes the pulmonary edema over the course of minutes to hours and NIV can then be terminated. The underlying chronic condition is subsequently treated in a normal care unit. The treatment of acutely exacerbated COPD follows a similar pattern.
Other frequent causes of acute respiratory failure, for example pneumonia, had a lower nonintensive care admission rate, in accordance with studies showing lower overall NIV success rates in this condition [13, 14]. One reason might be that fast alleviation through medical intervention is not possible. Therefore, NIV treatment for pneumonia in the emergency department seems to be less efficient. Acute asthma attacks can happen very fast and are sometimes resistant to medical treatment, resulting in life-threatening situations. To avoid such incidents, the indications for invasive ventilation in an acute asthma attack are set widely [11, 12], which might result in the observed low nonintensive care admission rates in the emergency department.
With age, prior studies show a lower NIV success rate in the higher age range [15] or no significant correlation [16]. In our study, however, older patients had a higher rate of nonintensive care admission than younger ones. We assume that diseases with higher NIV success rates, such as acutely exacerbated COPD or pulmonary edema, are the ones that normally become acute in older people. NIV is very effective in correcting these acute exacerbations back into a steady state. Acute asthma attacks and pneumonia on the other hand, conditions that also occur in the younger age range more often, have lower NIV success. This correlation shows that NIV in the emergency department might be especially effective in older patients. Since our result contradicts earlier studies, further clarification in this area is needed.
Another significant correlation concerns the year of intervention in the emergency department. Studies suggest that the level of staff training is associated with overall NIV success rate [17, 18]. When the emergency department at the study center was inaugurated in 2015, the application of NIV was not daily routine, but in the following months and years, experience in applying this treatment effectively and efficiently grew. Our data confirm that not only was it applied more frequently every year, but also nonintensive care admission rates ascended, showing that the observation made in prior studies is also true for an emergency department. We also think that rising familiarity with NIV at the department made the operators more confident in its effectiveness, leading to less transfers to intensive care units due to insecurity.
The main limitation of this study was its retrospective nature, which restricted the study design and results by available data. Information on ventilation parameters, patient’s physiology and comorbidity scores or treatment goals were not available but would have allowed grouping of patients with similar characteristics. Due to missing blood gas values and vital parameters, a patient’s course of disease could only be defined by the type of hospital care unit they were subsequently transferred to. Since there is no control group that did not receive NIV, it cannot be said with certainty that a reduction of ICU admissions was entirely due to NIV.
As usual in pilot studies, we opted for an exploratory study design. This gave us the opportunity to find interesting correlations more easily but limited the validity of the correlation significance due to missing Bonferroni correction. The sample size (9 cases) in the asthma subgroup was not too small to invalidate statistical analysis, however, we would suggest further research with higher sample sizes in this subgroup.
In conclusion, our study was able to give interesting insights into the impact of NIV application at a municipal emergency department. NIV was most effective in patients suffering from pulmonary edema, as well as patients in the higher age range and patients who were treated later in the study’s timeframe due to more experienced staff.
In order to solidify these results, more research is needed especially in confirming the correlations between nonintensive care admission rates and underlying medical condition and age. We would suggest prospective studies with additional variables such as physiology and comorbidity indices and possibly ventilation parameters.
These results emphasize the importance of an emergency department for an efficient hospital-wide workflow in treating acute care patients and preserving viable intensive care resources. We hope this study can serve as an incentive for care points who do not have NIV in use yet to implement it into their workflow.
Conflict of interest
Y.-T. Abulesz and M. Haugk declare that they have no competing interests.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
1. Hess DR Pang JM Camargo CA Jr. A survey of the use of noninvasive ventilation in academic emergency departments in the United States Respir Care 2009 54 10 1306 1312 19796409
2. Osadnik CR Tee VS Carson-Chahhoud KV Picot J Wedzicha JA Smith BJ Non-invasive ventilation for the management of acute hypercapnic respiratory failure due to exacerbation of chronic obstructive pulmonary disease Cochrane Database Syst Rev 2017 10.1002/14651858.CD004104.pub4 28702957
3. Weng CL Zhao YT Liu QH Fu CJ Sun F Ma YL meta-analysis: noninvasive ventilation in acute cardiogenic pulmonary edema Ann Intern Med 2010 152 9 590 600 10.7326/0003-4819-152-9-201005040-00009 20439577
4. Tomii K Seo R Tachikawa R Harada Y Murase K Kaji R Impact of noninvasive ventilation (NIV) trial for various types of acute respiratory failure in the emergency department; decreased mortality and use of the ICU Respir Med 2009 103 1 67 73 10.1016/j.rmed.2008.08.001 18804357
5. Horvath CM Brutsche MH Schoch OD Schillig B Baty F vonOw D NIV by an interdisciplinary respiratory care team in severe respiratory failure in the emergency department limited to day time hours Intern Emerg Med 2017 12 8 1215 1223 10.1007/s11739-016-1546-z 27722910
6. OECD. Hospital beds (indicator). 2020. https://data.oecd.org/healtheqt/hospital-beds.htm. Accessed 11 June 2020.
7. Ma X Vervoort D Critical care capacity during the COVID-19 pandemic: global availability of intensive care beds J Crit Care 2020 58 96 97 10.1016/j.jcrc.2020.04.012 32408107
8. Robert R Coudroy R Ragot S Lesieur O Runge I Souday V Influence of ICU-bed availability on ICU admission decisions Ann Intensive Care 2015 5 1 55 10.1186/s13613-015-0099-z 26714805
9. Maves RC Downar J Dichter JR Hick JL Devereaux A Geiling JA Triage of scarce critical care resources in COVID-19 an implementation guide for regional allocation: an expert panel report of the task force for mass critical care and the American college of chest physicians Chest 2020 158 1 212 225 10.1016/j.chest.2020.03.063 32289312
10. Norris C Jacobs P Rapoport J Hamilton S ICU and non-ICU cost per day Can J Anaesth 1995 42 3 192 196 10.1007/BF03010674 7743567
11. Westhoff M Schonhofer B Neumann P Bickenbach J Barchfeld T Becker H Noninvasive mechanical ventilation in acute respiratory failure Pneumologie 2015 69 12 719 756 26649598
12. Rochwerg B Brochard L Elliott MW Hess D Hill NS Nava S Official ERS/ATS clinical practice guidelines: noninvasive ventilation for acute respiratory failure Eur Respir J 2017 50 2 1602426 10.1183/13993003.02426-2016 28860265
13. Confalonieri M Potena A Carbone G Porta RD Tolley EA Meduri UG Acute respiratory failure in patients with severe community-acquired pneumonia. A prospective randomized evaluation of noninvasive ventilation Am J Respir Crit Care Med 1999 160 5 Pt 1 1585 1591 10.1164/ajrccm.160.5.9903015 10556125
14. Jolliet P Abajo B Pasquina P Chevrolet JC Non-invasive pressure support ventilation in severe community-acquired pneumonia Intensive Care Med 2001 27 5 812 821 10.1007/s001340100869 11430536
15. Antonelli M Conti G Moro ML Esquinas A Gonzalez-Diaz G Confalonieri M Predictors of failure of noninvasive positive pressure ventilation in patients with acute hypoxemic respiratory failure: a multi-center study Intensive Care Med 2001 27 11 1718 1728 10.1007/s00134-001-1114-4 11810114
16. Poponick JM Renston JP Bennett RP Emerman CL Use of a ventilatory support system (BiPAP) for acute respiratory failure in the emergency department Chest 1999 116 1 166 171 10.1378/chest.116.1.166 10424521
17. Lopez-Campos JL Garcia PC Jimenez LA Arnedillo A Gonzalez-Moya E Fenandez BJJ Staff training influence on non-invasive ventilation outcome for acute hypercapnic respiratory failure Monaldi Arch Chest Dis 2006 65 3 145 151 17220104
18. Yalcinsoy M Salturk C Oztas S Gungor S Ozmen I Kabadayi F Can patients with moderate to severe acute respiratory failure from COPD be treated safely with noninvasive mechanical ventilation on the ward? Int J Chron Obstruct Pulmon Dis 2016 11 1151 1160 10.2147/COPD.S104801 27330283
| 36441338 | PMC9703406 | NO-CC CODE | 2022-11-29 23:21:41 | no | Wien Klin Wochenschr. 2022 Nov 28;:1-7 | utf-8 | Wien Klin Wochenschr | 2,022 | 10.1007/s00508-022-02111-1 | oa_other |
==== Front
J Bank Regul
Journal of Banking Regulation
1745-6452
1750-2071
Palgrave Macmillan UK London
207
10.1057/s41261-022-00207-2
Original Article
Environmental sustainability and financial stability: can macroprudential stress testing measure and mitigate climate-related systemic financial risk?
http://orcid.org/0000-0001-8309-8098
DeMenno Mercy Berman [email protected]
Mercy Berman DeMenno
is Senior Fellow of the Duke Global Financial Markets Center and Nonresident Affiliate of the Duke Center on Risk as well as Principal of Bosque Advisors, a boutique consultancy specializing in risk, regulation, and resilience. A political economist by training, her research focuses on financial markets and regulation, energy and environmental policy, regulatory governance, and corporate sustainability. Dr. DeMenno has advised executives and policymakers at the local, national, and international levels. She holds a PhD and an MA from Duke University and an MBA and a BA from the University of New Mexico.
grid.26009.3d 0000 0004 1936 7961 Duke University, Durham, NC USA
28 11 2022
129
20 9 2022
© The Author(s), under exclusive licence to Springer Nature Limited 2022, Springer Nature or its licensor holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Despite widespread recognition among financial regulators and central banks that climate change may threaten financial stability, the causes and consequences of climate-related systemic financial risk remain underexplored. Stress testing has emerged as one of the most prevalent regulatory tools for addressing climate-related financial risks, and this article analyzes the role of stress testing in mitigating the effects of climate change on financial stability. Specifically, it synthesizes the multi-disciplinary literature on climate-related financial risk, financial stability, and stress testing to develop a framework for evaluating the capacity and effectiveness of stress tests for measuring and managing climate-related systemic financial risk. It then takes stock of climate stress testing proposals and early practices globally and applies the evaluative framework in comparative case studies of the Bank of England and US Federal Reserve. It concludes that stress testing can support the measurement and management of both microprudential and macroprudential climate-related financial risks, but the benefits of stress testing vis-à-vis climate change and financial stability are largely unrealized. Addressing the disconnect between climate stress testing policy motivation and implementation as well as the divergence between leading and lagging jurisdictions will require both interagency and international regulatory cooperation.
Keywords
Climate-related financial risk
Systemic risk
Financial stability
Macroprudential regulation
Stress testing
==== Body
pmcIntroduction
In 2015, Mark Carney referred to climate change as the “tragedy of the horizon,” remarking that because of the incongruous temporal scopes of financial stability policy and climate change consequences, “once climate change becomes a defining issue for financial stability, it may already be too late” [1 p.3]. There has since been a remarkable shift: a 2019 survey of financial regulators and central banks (n = 33) found that some 70% view climate change as a “major threat” to financial stability [2 p.7]. In some ways, this shift is unsurprising given the increasingly widespread appreciation of the dire consequences of unmitigated climate change and interconnectedness of these consequences across sectors and geographies [3]. Yet, addressing the financial risks resulting from both the physical manifestations of, and societal responses to, climate change introduces a host of novel analytical and governance challenges for financial regulators and central banks [4]. Over the last several years, financial regulators around the globe have begun identifying, assessing, and, to a lesser extent, addressing climate-related financial risks. For example, a 2020 survey of Financial Stability Board members (n = 33)—composing international and domestic banking, insurance, and investment regulators—found that nearly three-quarters (73%) consider, or plan to consider, climate-related financial risks in financial stability monitoring [5]. Another 2020 survey (n = 27) highlights the various activities that financial regulators and central banks are undertaking to better understand climate-related risks, including conducting research (89%), convening stakeholder conversations (92%), surveying firm practices (75%) and disclosures (87%), and issuing supervisory guidance (41%) [6].
Stress testing is among the most common regulatory tools for assessing and addressing climate-related financial risks. Myriad governmental, intergovernmental, and nongovernmental organizations have issued proposals calling for the institutionalization of climate stress testing [7–17]. Although a 2019 survey of financial regulators and central banks (n = 33) found that only 15% of respondents include climate considerations in routine stress tests, some 79% reportedly planned to incorporate climate-related financial risks into future stress tests [2]. A 2021 survey of supervisors and regulators focused more broadly on the incorporation of environmental, social, and governance risks into stress tests (n = 14) found similar results, with some 14% of respondents already integrating these risks into supervisory stress tests, 79% planning to do so in the next three years, and 7% engaging in related activities, such as voluntary exercises [18]. Despite this widespread interest among policymakers, climate stress testing practices are nascent [5, 6, 18–21]. Moreover, although financial regulators and central banks have framed climate change as a financial stability issue, there has been relatively little attention paid to the causes and consequences of climate-related systemic financial risk in extant policy discussions. The relationship between climate change, financial stability, and potential regulatory responses is similarly underexplored in the literature [22–24]. Recognizing that financial market policymakers and participants cannot manage what cannot be measured, and cannot measure what cannot be defined, this article provides a novel analysis of climate-related systemic financial risk and the role of stress testing in mitigating the effects of climate change on financial stability.
The article proceeds as follows: The “Climate change and financial instability” section defines and contextualizes the motivating problem—the potential effects of climate change on financial stability—through a discussion of the causes and consequences of climate-related systemic financial risk, integrating the multi-disciplinary literature on climate-related financial risk and financial stability. The “Stress testing in theory and practice” section reviews the literature on stress testing, with a focus on lessons learned in the post-global financial crisis implementation of microprudential and macroprudential stress tests. The “Evaluative framework for climate stress testing” section draws on the preceding sections to develop a framework for evaluating the capacity and effectiveness of stress tests for measuring and managing climate-related systemic financial risk. The “Climate stress testing practices” section takes stock of climate stress testing proposals and practices globally and applies the evaluative framework in comparative case studies of the Bank of England (BOE) and Board of Governors of the US Federal Reserve System (FRB). The “Conclusion” section summarizes the article and describes areas for future research. To preview the key findings: stress testing can support the measurement and management of both microprudential and macroprudential climate-related financial risks, but the benefits of stress testing vis-à-vis climate change and financial stability are unrealized. Addressing the disconnect between climate stress testing policy motivation and implementation as well as the divergence between leading and lagging jurisdictions will require both interagency and international regulatory cooperation.
Climate change and financial instability
Assessing the effects of climate change on financial stability requires attentiveness to both the causes and consequences of climate-related systemic financial risk. While there has recently been considerable focus on climate-related financial risks and a recognition that such risks may be systemic, the underlying transmission channels and propagation pathways are underexplored in both the academic literature and extant policy discussions [5, 6, 20, 25–28]. Indeed, although there is a long standing literature on the macroeconomic impacts of climate change and an emerging literature on the relationship between certain climate-related risks and financial assets, the literature on both the transmission channels and consequences of climate-related risks for financial institutions and, particularly, systems is much less developed [20, 23, 24]. Synthesizing the climate-related financial risk and financial stability literatures, this section provides a high-level discussion of the environmental, economic, and financial risks resulting from climate change (light, medium, and dark blue boxes, respectively, in Fig. 1), their transmission channels, and the pathways by which climate-related financial risks may become systemic, thereby inhibiting financial stability (gray arrows in Fig. 1).Fig. 1 Transmission of climate risk to financial markets and propagation of climate-related systemic financial risk
Climate-related physical and transition risks
Climate change creates and exacerbates natural hazards, which can be categorized as chronic or acute. Chronic natural hazards are persistent stressors (i.e., slow onset events) resulting from a warming climate, such as sea-level rise [12, 20, 29–31]. Acute natural hazards are episodic shocks (i.e., sudden onset events) driven by climate change, such as hurricanes [12, 20, 29–31]. As the climate warms more rapidly, the probability distributions of climate-driven natural hazards are shifting, resulting in an increased likelihood of catastrophic environmental consequences (i.e., fatter tails) [3].
The economic risks associated with climate change are categorized in the literature as physical and transition risks. Physical risk encompasses the economic consequences of an acute or chronic climate-driven natural disaster, such as the disruption, repair, and recovery costs a firm and its stakeholders incur as the result of inundation from a hurricane or extreme cold from a winter storm [12, 25].1 Transition risk encompasses the economic consequences of policies (e.g., carbon tax), technological advancements (e.g., long-duration energy storage), or societal actions (e.g., shifting consumer or investor preferences) to mitigate or adapt to climate change [12, 25]. Extant discussions of transition risk often focus on the economic consequences of an abrupt transition to a lower-carbon economy, such as via rapid devaluation of carbon-intensive assets, but the transition to a sustainable economy should be defined more broadly. Specifically, as the Intergovernmental Panel on Climate Change’s (IPCC) most recent report highlights, transitioning to a sustainable economy requires mitigation of all greenhouse gas emissions (i.e., carbon dioxide, methane, nitrous oxide, and fluorinated gases)—which involves not only a reduction in the production of greenhouse gas emissions but also the capture, use, and storage of emitted greenhouse gases to reduce atmospheric concentrations—as well as adaptation to the existing irreversible effects of climate change [31–34]. Transition risk is distinct from physical risk in that it involves interaction among physical, political, social, and market dynamics, such that the adverse consequences may arise from both substantive changes as well as from uncertainty surrounding potential changes [25]. As such, the risks associated with the transition to a sustainable economy are often described as varying based on the extent to which the transition is “orderly” (i.e., immediate and measured) versus “disorderly” (i.e., sudden and unanticipated) [14].
Climate-related financial risks
Climate-related physical and transition risks affect both sectors in the real economy and the financial services sector. Climate-related financial risk can be quantified at the asset, institution, or system level. Climate change can affect financial assets by directly reducing the value or productivity of underlying real assets. For example, a natural disaster may destroy or rapidly depreciate the value of capital assets (i.e., property, plant, and equipment) while a carbon tax may increase operating costs for carbon-intensive activities, thereby reducing productivity of a given asset [35]. At the institution level, the effects of climate change may manifest in credit, market, liquidity, operational, or reputational risk. The materiality of these different risks will vary by institutional characteristics. An acute shock like a hurricane, for example, may increase a bank’s credit risk (e.g., impairment of collateral assets), market risk (e.g., volatility in bonds and stocks for climate-exposed corporates), liquidity risk (e.g., fire sales for impacted assets), and operational risk (e.g., business interruption costs for locations or supply chains). Table 1 provides examples of climate-related financial risks by transmission channel (i.e., physical, transition), financial risk type (i.e., credit, market, liquidity, operational, reputational), and subsector (i.e., banking, insurance, investment); because large financial institutions often provide diversified financial services, climate-related financial risks for a given firm may span multiple subsectors.Table 1 Examples of climate-related financial risk by transmission channel, risk type, and subsector
Credit Market Liquidity Operational Reputational
Physical Banking Loan defaults resulting from disruption costs for climate-exposed borrowers and impairment of collateral assets Valuation uncertainty for climate-exposed assets Interruption of electronic payment systems and associated demand for cash reserves Business interruption or liability costs for climate-exposed locations or supply chains Customer dissatisfaction with climate hazard mitigations
Investment Bond defaults resulting from disruption costs for climate-exposed issuers Volatility in stocks, bonds, and derivatives for climate-exposed industries, regions, and underlying assets Fire sales for climate-exposed assets Business interruption or liability costs for climate-exposed locations or supply chains Customer dissatisfaction with climate hazard mitigations
Insurance Large-scale correlated claims for climate-exposed property and casualty policyholders due to increasingly frequent and severe acute and chronic climate hazards Risk transfers to reinsurers and government insurers of last resort for climate-exposed underwriting Customer dissatisfaction with climate hazard mitigations
Transition Banking Loan defaults resulting from technology- or policy-driven transition costs for carbon-intensive borrowers Valuation uncertainty for carbon-intensive (or transition-enabling) assets Rapid draw down of deposits or credit lines for counterparties affected by climate change mitigation or adaptation policies Legal and compliance costs associated with carbon-intensive lending portfolios Decreased shareholder value due to deficiencies in climate risk management
Investment Bond defaults resulting from technology- or policy-driven transition costs for carbon-intensive issuers Abrupt price corrections in stocks and bonds for carbon-intensive industries and economies Fire sales for assets affected by climate change mitigation or adaptation policies Legal and compliance costs associated with carbon-intensive investment portfolios Decreased shareholder value due to deficiencies in climate risk management
Insurance Premia mispricing for carbon-intensive (or transition-enabling) policyholders given uncertainty around climate change mitigation and adaptation policies and technologies Risk transfers to reinsurers and government insurers of last resort for carbon-intensive underwriting Decreased shareholder value due to deficiencies in climate risk management
Climate-related systemic financial risk
Climate change may threaten financial stability if climate-related financial risks are, or become, systemic. Despite widespread interest in the concept of systemic risk in the post-global financial crisis era, there is no widely accepted definition. Drawing on the existing literature, this article defines systemic risk as risk that propagates at the financial system level (i.e., financial shock amplification), impeding the functioning of the financial sector (i.e., financial system impairment), and creating spillovers into the real economy (i.e., financial externalities) [38]. Systemic risk should be distinguished from specific risk (i.e., idiosyncratic risk), which manifests at the asset and institution levels and can be managed at the portfolio level through diversification. Although often conflated, systemic risk should also be distinguished from systematic risk (i.e., market risk), which manifests at the sector or subsector level but does not have the potential to impede the operating of the entire financial system or create spillovers into the real economy [39]. While much of the discussion about systemic risk focuses on country-level macrofinancial linkages, a “system” can be defined at various scales ranging from local to global. Recognizing the regional dynamics of climate-related financial exposures—for example, regions susceptible to hurricanes and primarily served by small and community banks—the concept of “sub-systemic” risk has been articulated in recent regulatory discussions [10]. One conceptually useful way to understand systemic risk is vis-à-vis the potential of a risk to inhibit financial stability (i.e., the ability of the financial system to manage risks and absorb shocks in order to continually support economic processes) [40, 41].
Climate change may trigger both exogenous and endogenous sources of systemic risk. Unmitigated global warming may cause sufficiently widespread and consequential exogenous shocks to be destabilizing to the financial system and the real economy. However, systemic risk may also be endogenous, resulting from the amplification of an initial localized shock and ensuing system-wide propagation of risk [42]. Various system dynamics—including interconnectedness (e.g., tight coupling within and across banking, insurance, and investment services), complexity (e.g., multi-sector and multi-jurisdiction scope), brittleness (e.g., mispricing due to uncertainty), and procyclicality (e.g., excessive leverage)—can make a system more vulnerable to systemic risk amplification, including as a result of climate change [5]. Because anticipating and preventing exogenous shocks is exceedingly difficult, systemic risk management often focuses on identifying propagation pathways and addressing firm- and system-level vulnerabilities that enable risk magnification [42]. However, climate change is a known risk, and as such both the exogenous triggers and endogenous propagation of climate-related systemic financial risk can be analyzed.
As depicted by the feedback loops in Fig. 1, the propagation pathways for climate-related systemic risk are complex and interdependent. Indeed, the Bank for International Settlements (BIS) has described climate change as a “green swan,” because it is a source of systemic risk characterized by “interacting, nonlinear, fundamentally unpredictable, environmental, social, economic and geopolitical dynamics” [36 p. 1, 43]. Building greater understanding of these dynamics is an active area of research [5, 6, 25, 27], and the following hypothetical examples illustrate two potential climate-related systemic financial risk propagation pathways (depicted by the “P” and “T” arrows in Fig. 1).
The first example focuses on physical risk: a severe hurricane that negatively affects individuals and firms in the real economy (P1). The resulting inundation of residential and commercial property leads to asset devaluation, such as loan collateral for mortgages (P2). This specific risk becomes a systematic risk as the resulting devaluation leads to widespread losses among financial firms insuring, investing in, or lending against those assets via property insurance, mortgage-backed securities, or collateralized real estate loans (P3). This systematic risk becomes a systemic risk as the mispricing of climate-related risk and interconnectedness amplify initial losses via fire sales and counterparty contagion, which in turn impedes the functioning of the financial system (P4) and constrains lending to the real economy (P5) [44, 45]. This credit supply shock contributes to macroeconomic contraction, which, when coupled with other direct effects of the precipitating risk (e.g., reduced gross regional product due to business disruptions) creates a negative feedback loop to financial system instability (P6). This simplified example is most plausible when the system is defined relatively locally. However, these dynamics could also be observed at the regional or national level as climate-driven natural hazards become more frequent and severe, thereby increasing the probability of coinciding and compounding climate hazards (e.g., heatwaves and wildfires) within geographies and sectors [3].
The second example focuses on transition risk: a climate mitigation policy that internalizes emissions externalities, for which affected industries have not sufficiently prepared and thus results in a sudden increase in firm operating costs (T1) and correction in asset pricing (T2) [46]. This abrupt transition affects short-term production as carbon-intensive industries adapt, thereby potentially contributing to a macroeconomic contraction (T3). In parallel, it also creates “stranded assets” (i.e., assets such as fossil fuel reserves that are rapidly and substantially devalued) and this specific risk becomes a systematic risk because devaluation affects not only asset owners in the real economy, but also creditors, shareholders, and insurers that fail to properly price transition risks (T4) [47, 48]. These widespread correlated exposures among financial institutions could result in systemic risk as uncertainty amplifies the initial price shock (T5). The rapid deflation of a “carbon bubble”(i.e., the inflated valuation of carbon-intensive firms resulting from the substantial negative carbon emissions externality) could result in reinforcing negative feedback loops among the financial sector (T6) and real economy (T7), mirroring the asset bubbles that have preceded prior financial crises [1, 49].
Thus, evaluating the effects of climate change on financial stability requires identifying how the precipitating climate change risk is transmitted to the financial sector, how the resulting financial risks propagate across the financial system, and how macrofinancial linkages amplify or absorb these risks. Drawing on the discussion of physical and transition risks, climate-related financial risks, and the potential propagation of climate-related systemic risk in this section, the subsequent sections analyze whether and how stress testing can be used to operationalize these dynamics to mitigate the effects of climate change on financial stability.
Stress testing in theory and practice
To analyze whether and how stress testing can be used to measure and manage climate-related systemic risk, it is important to first reflect on the existing uses of stress testing in the financial sector, and the benefits and limitations of stress testing as a macroprudential tool. This section summarizes the emergence, current practice, and scholarly critiques of stress testing, which in turn informs the development of the evaluative framework for climate stress testing presented in the next section.
Stress testing is a form of scenario-based analysis used to evaluate the resilience of an institution or system to hypothetical stressors [50, 51]. Financial stress tests use scenarios to define exogenous shocks and adverse market conditions, and financial models to project how these stressors affect the exposures of financial firms or sectors and the implications for institutional or system resilience [52]. Financial firms have long used scenario-based analysis for internal risk management, and the implementation of bank stress tests was supported by international frameworks such as the Basel Committee for Banking Supervision’s 1996 Market Risk Amendment and 2004 Basel II framework [53, 54]. In addition to these firm-level stress tests, the International Monetary Fund (IMF) and World Bank established the Financial Sector Assessment Program in 1999, which includes macro stress tests of participating countries’ financial systems [54]. During the 2007–2009 global financial crisis, stress testing became an integral component of crisis management, particularly for the FRB and the Committee of European Banking Supervisors. Stress testing was systematically integrated into post-crisis microprudential and macroprudential regulatory and supervisory toolkits [54, 55].
As this brief history demonstrates, supervisory stress testing—i.e., stress tests designed, implemented (or overseen), and analyzed by regulatory agencies, central banks, or supervisory authorities—plays a distinctive role depending on when it is deployed. In “wartime,” stress testing supports crisis management [56]. For example, it can inform recapitalization strategies and restore market confidence by sending a signal about the resilience of financial firms and sectors amidst a crisis and concurrent policy responses. In “peacetime,” stress testing supports risk management [56]. For example, it can be used to monitor risks and inform regulation by providing a forward-looking assessment of institutional and system resilience to potential (severe but plausible) adverse scenarios.
Supervisory peacetime stress tests can be further categorized by their policy objectives [57]. Microprudential stress tests seek to inform firm-level actions, including setting microprudential standards and informing firm risk management practices. For example, solvency stress tests provide a forward-looking dynamic assessment of capital adequacy and are used to calibrate static backward-looking firm-level capital standards in many jurisdictions [54, 55, 58]. Macroprudential stress tests seek to inform system-level actions, including monitoring systemic risk and calibrating policies targeting financial stability. For example, some jurisdictions use stress tests to calibrate countercyclical capital buffers and sectoral capital requirements [59]. While some authorities, such as the European Central Bank (ECB), separate microprudential and macroprudential stress testing exercises, many jurisdictions implement stress tests that have both microprudential and macroprudential objectives [60].
A theme across the extant stress testing literature is that the design of stress tests should be aligned with underlying policy objectives [52, 56]. While this seems fairly intuitive, in practice policy objectives and stress test designs have diverged based on regulatory authorities and resources [57, 61]. A key design consideration is the level at which stress tests are implemented (i.e., firm vs. system), which should align with the risks that are being measured and the potential risk management actions that are being considered, and will determine the metrics, models, and data that are fit for purpose [61]. Firm-level stress tests measure the effects of a given scenario on a firm’s portfolio and generally employ bottom-up models, in which the empirical relationship between the stressor and the outcome of interest is estimated at a high level of granularity using firm-specific data.2 This modeling approach enables an assessment of financial institutions’ idiosyncratic risks and as such is useful for informing firm-level risk management practices and calibrating microprudential regulatory requirements. In contrast, system-level stress tests seek to measure the effects of a given scenario at the sector or subsector (e.g., banking, insurance, investment) level, and are generally conducted by centralized regulatory authorities using top-down models, in which the empirical relationship between the stressor and the outcome of interest is estimated at a lower level of granularity using aggregate macroeconomic and regularly reported macrofinancial data [50–52]. This modeling approach enables an assessment of system-wide risk and as such may inform the calibration of macroprudential standards [43, 52]. In practice, stress tests with both microprudential and macroprudential objectives have combined top-down, bottom-up, and hybrid modeling approaches [57]. For example, the BOE and FRB both conduct stress tests with microprudential and macroprudential objectives but employ different modeling priorities with respect to granularity and severity. In the US, supervisors focus modeling efforts on “forming an independent view of bank level results at a fairly granular level,” often employing more plausible scenarios, while in the UK, supervisors focus modeling efforts on “understanding the system-wide effects involving spillovers into the rest of the financial system and the real economy,” often employing more severe scenarios [56 p.134]. Some scholars have argued that the former enables a more robust coupling of stress testing and microprudential regulation and partially addresses concerns about model instability in more top-down approaches [63]. However, others have argued that the latter plays to the comparative advantage of regulators in conducting peacetime macroprudential stress tests with more severe scenarios to explore emergent risks and system vulnerabilities, which in turn can be used to link stress testing to macroprudential regulation [64].
Notwithstanding the argument that firms have a comparative advantage in understanding institutional risks whereas regulators have a comparative advantage in understanding system-wide risks [56], another key theme in the literature is the analytical complexity associated with measuring and managing financial stability risks, including via macroprudential stress testing. In particular, the literature underscores the challenges of modeling both exogenous and endogenous sources of systemic risk [43]. As discussed above, systemic risk may result from an exogenous shock that is sufficiently widespread and high consequence to be destabilizing. Thus, one concern raised in the literature is whether the shocks introduced in stress test scenarios are appropriately severe to represent potential threats to financial stability [55]. Indeed, scholars have argued that stress tests have performed poorly as early warning devices, calling into question the utility of stress tests in operationalizing inherently difficult to predict tail risks (i.e., low probability, high consequence) [52, 55]. However, systemic risk often results from the amplification of risk throughout the financial system, even if the precipitating shock is relatively localized. A second concern raised in the literature is how completely the propagation of systemic risk is analyzed. In general, microprudential stress tests model first-round effects—i.e., the direct and isolated impacts of a scenario on a firm’s balance sheet—whereas macroprudential stress tests should include both first- and second-round effects—i.e., the indirect and interactive impacts of a scenario on a broad range of financial and non-financial sector agents [57]. To measure these indirect and interactive effects, stress tests need to model the propagation of risk throughout the system, including amplification among interconnected financial institutions and potential spillovers into the real economy [43, 51, 55]. Existing studies, and historical experience, suggest the magnitude of these second-round effects can be quite substantial [65, 66].
For jurisdictions that implement stress tests to fulfill both macroprudential and microprudential objectives, the direct effects captured in firm-level horizontal stress tests may provide the foundation for analysis of second-round effects at the system level [51]. Several authorities have made progress on developing modeling approaches for direct contagion channels (i.e., cross-holdings) and selected macrofinancial feedback loops within stress tests [43]. For example, the ECB’s macroprudential exercise models feedback effects for individual banks and interbank contagion as well as selected macrofinancial linkages [57]. Another example is the IMF, which models amplification mechanisms resulting from interconnectedness and leverage in the interbank market as well as feedback loops between the financial sector and real economy within and across jurisdictions [67]. However, the literature identifies persistent analytical challenges associated with representing indirect contagion effects (i.e., interlinkages), such as fire sales, information asymmetry, strategic complementarities, and herding [43], as well as integrating the full set of potential interactions among the financial sector and real economy [55]. Moreover, existing approaches tend to focus on transmission channels for which models are well-established [57], resulting in incompleteness around novel sources of risk that do not propagate through standard macrofinancial channels. Various strategies to incorporate a broader range of transmission channels into stress tests have been proposed, such as integrating liquidity and solvency stress testing to capture liquidity-solvency interactions as a source of microprudential risk and potential systemic feedback loops (e.g., via fire sales), as well as layering network analysis and agent-based-modeling to analyze the propagation of macroprudential risk via contagion and dynamic firm behavior, respectively [65, 66].
A final theme in the literature is that stress tests should enable not only risk measurement but also risk management, albeit with robust debates about the efficacy of alternative mechanisms [61]. For example, the scope, timing, and granularity of stress test disclosure varies across jurisdictions [57], and while there is scholarly consensus that disclosure of firm-level results is essential for promoting market discipline [55], there are divergent perspectives on the consequences of disclosing models and methodologies (e.g., due to concerns about gaming and model monoculture) [68]. There is also variation in how stress test results are used and debates about the implications for effectively managing risks. Some have argued that for stress tests to be effective they must be tied to remedial actions, such as calibrating regulatory requirements or approving firm strategies [54, 55]. Indeed, one of the BIS’ stress testing principles is that “stress testing should be used as a risk management tool and to inform business decisions,” including “to quantify the capital needs at [a] systemic level during a time of crisis” and “to inform the calibration of macroprudential policies and instruments” [61 p.87]. Others have argued that the primary risk management benefit of stress testing is structuring thinking among various stakeholders, including firms and market regulators with more microprudential perspectives and central banks and financial stability regulators with more macroprudential perspectives [52, 69].
In summary, stress testing is an essential financial regulatory risk measurement and management tool. While the governance, implementation, and outcomes of stress testing have become more robust in the decade since the global financial crisis, the literature raises concerns about the alignment between stress testing objectives and design, the challenges associated with measuring systemic risk, and the ways in which stress testing can support not only risk measurement, but also more effective risk management.
Evaluative framework for climate stress testing
This section develops a framework for analyzing whether and how stress testing can be used to measure and manage climate-related systemic risk, and in turn, mitigate the effects of climate change on financial stability. Drawing on the climate-related financial risk, financial stability, and stress testing literature discussed in the preceding sections, the framework is grounded in two principles.
First, the design and use of climate stress tests should be aligned with the underlying policy objective. Stress tests, including those incorporating climate change, may be used to address financial firm risks, financial system risks, macroeconomic risks, and, in some cases, a central bank’s own balance sheet risks [15]. Given that financial regulators and central banks have largely mobilized around climate change as a potential threat to financial stability, climate stress tests should seek to inform the measurement and management of climate-related systemic risk. With respect to measurement, climate stress tests should provide legibility into the effects of climate change on financial institutions and the financial system in which they are embedded, with attentiveness to how climate-related financial risk can propagate through the financial system and create spillovers into the real economy. To mitigate the effects of climate change on financial stability, stress tests should enable more effective or efficient management of climate-related systemic risk by financial regulators, firms, and markets.
Second, the integration of climate risks into stress tests should be informed by institutional constraints, including the scope of financial stability regulatory authorities and design of existing stress testing regimes. This attentiveness to institutional design is critical to ensure that climate stress testing is feasible in the near-term and climate risk is systematically integrated into financial risk measurement and management practices rather than treated as an ancillary non-financial risk in the longer-term. Given existing policy debates about the utility of macroprudential stress testing, the framework is particularly focused on opportunities to address deficiencies identified in the literature and calibrate existing stress tests to better capture potential systemic risks, including those introduced by climate change. Recognizing the wide range of existing stress test designs and heterogeneity of climate exposures across firms, the framework focuses on how firm-level stress tests can inform system-level analysis as well as risk management actions by regulators, firms, and markets via a two-stage modeling approach, thereby incorporating both microprudential and macroprudential policy objectives.
The framework consists of seven criteria that can be used to evaluate whether and how stress tests are, or could be, used to support the measurement of climate-related systemic risk (“Risk measurement” section) and whether and how the outcomes of stress tests are, or could be, used to improve the management of such risks (“Risk management” section). This framework is applied to comparative case studies of effectiveness and capacity below (“Framework application” section).
Risk measurement
The first set of criteria analyzes the capacity or effectiveness of a given stress testing regime for measuring climate-related systemic financial risk. Notably, the goal of macroprudential climate stress testing should not be to precisely predict the causes and consequences of climate-related systemic financial risks, but rather to explore various transmission channels and propagation pathways within the financial system and identify firm- and system-level vulnerabilities that enable the magnification of these risks. As such, the four risk management criteria—scope, scenario design, metrics, and systems analysis—focus on the extent to which stress tests do, or could, support assessment of the potential effects of climate change on financial stability.
Scope: supervisory stress tests are applied to all systemically important and materially climate-exposed financial firms
The first risk measurement criterion relates to the applicability of scenarios, that is, which financial firms are subject to climate stress tests. Although all firms with material climate exposures should consider employing scenario analysis to measure the resilience of their business strategies to physical and transition risks, initial prioritization of supervisory stress tests should be guided by the regulatory principles of materiality and proportionality.
Regulators should prioritize stress testing firms with the greatest climate-related financial exposures—firms for which the realization of severe physical or transition risks could impair functioning—and firms of the greatest systemic importance—firms whose impairment would pose the greatest risk to financial stability. This dual focus on the materiality of climate exposure and proportionality of systemic importance underscores the criticality of centralizing climate stress testing with regulators empowered to address financial stability rather than those focusing on conduct within a particular subsector (e.g., banking, insurance, investment). A sectoral approach may lead to the systematic underestimation of the effects of climate change on particular types of financial products and firms and an incomplete representation of interactions and amplification mechanisms, thereby limiting the validity of any results vis-à-vis financial stability [57, 70]. There may be reasonable, albeit imperfect, proxies to guide selection; for example, size, interconnectedness, complexity, substitutability, and cross-jurisdiction activity are used as indicators of systemic importance and factors such as relative concentration of lending in climate-exposed regions, financed emissions, or carbon-intensity of investment portfolios may be used as partial indicators of climate-related financial exposures [18, 20, 71, 72].
Scenario design and modeling approach: input variables and models incorporate climate-related financial risks transmitted via physical and transition channels, with appropriate spatial and temporal resolution and robust representation of uncertainty and interdependencies
The second risk measurement criterion relates to the stressors—i.e., the exogenous shocks and adverse market conditions as represented by input variables—applied within a given test and the modeling approaches used to estimate the impacts of stressors on tested firms. In other words, as depicted in Fig. 1, climate stress test scenarios and models provide the translation from climate risk to macroeconomic (or sectoral) risk to financial (asset, firm, or system) risk.
With respect to input variables, stress tests generally include an underlying narrative, transmission channels, and specification of the scope and severity of stressors [57]. For climate stress tests, the underlying narrative should reflect the acute or chronic climate hazards and policy, technology, or social changes that result in physical and transition risk, respectively. For example, physical risk might be operationalized via temperature pathways and the frequency and severity of resulting extreme weather events, while transition risk might be operationalized via carbon prices, emissions, and energy mix [14, 72]. Scenario narratives are essential for both guiding modeling and articulating non-modellable underlying assumptions, which may be particularly relevant for climate risks (e.g., due to political and behavioral dynamics). The transmission channels should specify how the underlying climate risks or responses represented in the narrative cause an exogenous shock or adverse market conditions, and the resulting impacts on key macroeconomic and macrofinancial input variables. For example, macroeconomic variables might include changes in gross domestic product, unemployment, sector profitability, and property prices while macrofinancial variables might include corporate and government bond yields, equity indices, and exchange rates. While existing stress test scenarios are often threat agnostic (e.g., operationalizing recessionary conditions but not defining the causes of the recession [73]), specification of the underlying causes and causal paths informs risk management and enables more robust ex-post validation of scenario and model assumptions, which is particularly important for novel financial risks, including those resulting from climate change. Given the relationship between physical and transition risks—for example, the increasing severity and frequency of climate-driven natural hazards creates more pronounced physical risks, which may in turn increase the likelihood of mitigation policies, thereby producing transition risks and, over the longer-term, reducing physical risks—climate stress tests should address both. Including multiple risks and transmission channels in a single scenario is more computationally intensive than developing parallel scenarios, but also enables more accurate representation of the interdependencies among climate-related risks and responses [36].
Scenarios should include robust representation of uncertainty with respect to the spatial and temporal scope and severity of climate risks and responses, as well as the resulting economic and financial consequences. Scenarios will necessarily have different scopes based on the underlying risk and transmission channel. For example, temperature-based scenarios tend to be longer than event-based scenarios, with policy-based scenarios varying based on a given temperature or emissions target [10]. Given that climate change will have both within- and cross-sector/-border impacts as well as near- and long-term impacts, integrating climate change into existing stress tests will almost certainly necessitate longer time horizons and broader scopes to fully capture risks and impacts. However, early progress on climate stress testing has demonstrated the ability to disaggregate nearer-term (i.e., within the current business cycle) and longer-term (i.e., outside of the current business cycle) assessments of climate risk and strategies to address the uncertainty associated with each [74]. Uncertainty about the magnitude of impacts can be partially captured through scenarios of varying severity, while uncertainty about the spatial or temporal scope of impacts can be partially captured by calibrating models to reflect near-term and localized extreme worst-case scenarios (i.e., tails of predicted climate risks to enable realization of macroeconomic and financial impacts within the scenario scope) [75]. Including at least one tail risk scenario is critical for representing the potential exogenous triggers of climate-related systemic financial risk. Given the probabilistic nature of climate risks and responses, uncertainty regarding the scope and severity of the resulting impacts, and the novelty of modeling these dynamics for financial regulators, early stress tests should account for some degree of model error and assumptions should be well documented and iteratively reviewed [63].
The second component of scenario design is the modeling approach, that is, the methodologies by which input variables are translated into impacts on financial assets, firms, and systems [43, 52]. As discussed in the previous section, stress tests may employ top-down, bottom-up, and hybrid estimation models, which vary in the scope of risks included and granularity of data required (e.g., level of sub-sectoral and regional decomposition). For example, transition risks may be more feasible to represent in top-down models because the hypothesized transmission pathways are operationalized at the macroeconomic and sectoral levels. Physical risks, on the other hand, may be more feasible to represent in hybrid or bottom-up models that can account for regional economic impacts and heterogeneous portfolio-level interactions. The modeling approach will also affect the steps required to conduct systems analysis and translate results into firm- and system-level risk management strategies. To the extent possible, modeling approaches should seek to balance microprudential and macroprudential objectives.
Although developing climate stress tests requires novel data and modeling capabilities [4, 76], there are opportunities to leverage scenarios and methodologies developed by intergovernmental organizations. For example, the Network of Central Banks and Supervisors for Greening the Financial System (NGFS) and the UN Environment Programme’s Finance Initiative have drawn on well-established general circulation models, natural hazard models, and integrated assessment models to develop reference scenarios for various physical and transition risks and, to some extent, translate these climate risks into macroeconomic and macrofinancial input variables [77, 78]. Discussed in more detail in the “Climate stress testing practices” section, leading jurisdictions have leveraged these reference scenarios and underlying climate-economy and macroeconometric models, commercially available scenario analysis tools, in-house financial models, and diverse data sources to implement climate scenario analyses and stress testing pilots [5, 21]. International cooperation among financial regulators on climate scenario design is particularly important because of the cross-border and multi-sector nature of climate-related financial risks and thus the need to compare and aggregate results across jurisdictions and sectors [18, 21]. Additionally, coordination among domestic and international financial and environmental authorities can ensure the accurate translation of climate and natural hazard models into scenarios appropriate for analyzing financial risks, which is essential given the heterogeneity of assumptions in underlying physical and social science climate models (e.g., regarding discount rates and climate sensitivity) [31, 79–81]. Finally, financial regulators and firms alike have flagged the importance of the development of a common set of scenarios and assumptions to provide the foundation for regulatory cooperation in the development of international standards for climate-related financial risk management [18].
Metrics: model output variables quantitatively and qualitatively measure how climate change affects the performance of tested firms and are translated into decision-relevant metrics
The third risk measurement criterion relates to what stress tests measure and whether output variables are relevant to firm or regulatory decision making. As described above, physical and transition channels may result in myriad types of financial risks, including credit, market, liquidity, operational, and reputational risks. Some of these risks may be captured in standard stress test performance metrics, but probabilistic metrics that isolate the consequences of climate risks will enable more accurate performance targets as firms move from climate risk measurement to risk management.
The maturation of metrics for stress tests may be informed by, and may inform, the parallel development of green taxonomies, sustainable accounting methodologies, and climate risk disclosure frameworks, which in turn are driving advancements in climate risk analytics. Green taxonomies categorize sustainable and transition-enabling economic activities and assets and provide standardized definitions and a framework for categorizing contributions to climate change mitigation or adaptation [82]. Sustainable accounting methodologies provide a measurement approach for translating contributions to climate change or climate mitigation and adaptation efforts into financial metrics, which in turn can be integrated into both government reporting and private auditing standards [83–87]. For example, portfolio-level emissions can be used in various metrics that provide useful estimates of climate exposures, such as weighted average carbon intensity—which provides a measure of portfolio exposure to carbon-intensive companies weighted by company revenue—, carbon earnings at risk—the net present value of potential future earnings based on different carbon prices—, and portfolio alignment with various temperature targets. Other metrics focus on transition-enabling efforts, such as the dollar value of “green” transactions metrics. Moreover, academics are increasingly exploring how existing financial metrics can be adapted to better reflect climate risk, for example, scholars have proposed a climate value at risk assessment approach—which applies traditional value at risk methodologies to firms’ balance sheets but isolates the effects of climate risk scenarios of varying severity—as well as methodologies to extend solvency analysis to isolate the effects of climate risk proxies on capital adequacy [26]. Commercial service providers have incorporated approaches from the literature into scenario analysis frameworks, in part to meet the demand emerging from voluntary climate risk disclosure frameworks, such as the Task Force on Climate-related Financial Disclosures (TCFD) [12, 88, 89].
In addition to these quantitative performance metrics, stress tests should include qualitative assessments of firms’ climate risk management practices. A recent survey of central banks and regulators suggests that evaluations of governance practices and risk management frameworks are among the most common approaches for understanding how financial firms are addressing climate change risks [2]. Qualitative and quantitative metrics should be considered in combination to provide a holistic picture of exposures to physical and transition risks and the extent to which firms’ potential responses exacerbate or mitigate these risks. A key benefit of regulatory cooperation in the development of climate stress testing—as with the post-global financial crisis development of solvency stress testing—will be establishing a more complete, consistent, and comparable set of metrics in an increasingly crowded and incompatible voluntary standards space.
Systems analysis: firm-level stress tests enable system-level analysis of the effects of climate change on financial stability, including correlated exposures, counterparty contagion, and the propagation of risks across the financial sector and real economy
The fourth risk measurement criterion relates to the extension of firm-level stress tests to measure the effects of climate change on the financial system and its stability via system-level analysis. As noted in the preceding section, a key tradeoff in the design of stress tests is the level of analysis, which should be driven by the underlying policy objectives and potential policy responses [52, 56]. Because climate-related financial risk may manifest and be managed at both the firm and system levels, climate stress tests should seek to combine microprudential and macroprudential objectives, using horizontal stress tests to measure firms’ idiosyncratic climate risks and systems analysis to understand how such risks aggregate and interact at the financial sector level.
As described above, climate change may trigger both exogenous and endogenous sources of systemic risk, which should be represented in systems analysis. To address the potential for a climate-driven shock that is sufficiently widespread to impede the functioning of the financial system, systems analysis should evaluate potential correlated exposures resulting from the inclusion of such a shock in the firm-level analysis. To address the potential amplification of climate-related financial risk, systems analysis should evaluate counterparty contagion and the propagation of risks within the financial system and across the financial sector and real economy. At the simplest level, systems analysis could consist of integration of firm-level results into aggregate metrics, such as total losses from a given climate change risk for a given set of firms. For example, scholars have recently proposed a systemic climate risk metric, “CRISK,” which is the country-level aggregate expected capital shortfall of banks resulting from climate transition risk scenarios [26]. As described above, it is common for regulators to use horizontal stress testing data and results for systems modeling, and aggregating firm-level results is necessary for understanding total direct effects and may be sufficient if the primary concern is widespread correlated exposures [26]. However, for many of the climate change dynamics that are hypothesized to have systemic implications, the concern is about not only the magnitude of direct effects, but also the potential for propagation throughout the system via contagion and other amplification factors [21–23]. Horizontal firm-level stress testing that incorporates counterparty analyses may account for channels of direct contagion but is insufficient to capture sources of indirect contagion (e.g., fire sales, feedback effects). System-level analysis should therefore seek to model second-round effects, which would reflect the indirect and interactive impacts of scenarios on a range of financial, and potentially non-financial, sector agents [21]. These interactions can be modeled at the financial subsector level (e.g., among banks), financial sector level (e.g., among banks, insurers, and reinsurers), and at the multi-sector level (i.e., among the financial system and sectors in the real economy) [43, 51, 55, 67].
As previously noted, systemic risk modeling is an active area of research and capabilities may need to be developed iteratively depending on the scope and level of sophistication of an implementing authority’s existing stress tests. For example, while many jurisdictions have developed strategies to quantify direct contagion and selected macrofinancial linkages, modeling of indirect contagion and non-traditional macrofinancial linkages is more limited. Developing a system-wide view of climate-related correlated exposures and potential counterparty contagion are critical first steps, and there are several opportunities for incremental expansion. For example, because the effects of climate change are heterogenous across sectors and geographies, improving understanding of macrofinancial linkages among the financial sector and those sectors or regions most exposed to climate-related physical or transition risks is a critical first step [67, 75]. There are also opportunities for learning among leading jurisdictions. As discussed above, the ECB and IMF have developed advanced the modeling of second-round effects in routine stress tests that may be relevant to climate stress testing, and international coordination among early adopters of climate stress testing may enable a more holistic assessment of cross-border effects [5, 90]. Finally, there are emergent methodologies in the literature that may inform practice, for example researchers have begun exploring how complex systems modeling approaches (e.g., network analysis) can be used to better account for direct and indirect climate exposures among financial institutions and the systemic implications [25]. Forward-looking jurisdictions should approach the development of climate stress tests as an opportunity to improve the modeling of second-round effects (i.e., the endogenous drivers of financial instability), which together with the introduction of shocks that better represent tail risks (i.e., the exogenous drivers of financial instability) would considerably advance the quality of systemic risk stress testing more broadly, particularly with respect to representing underlying volatility, uncertainty, complexity, and ambiguity.
Risk management
Effective risk measurement is necessary but insufficient for mitigating the effects of climate change on financial stability. The second set of criteria analyzes the capacity or effectiveness of a given stress testing regime for informing the management of climate-related systemic risk via market discipline, risk mitigation evaluation, and evidence-based regulation. The role and limitations of stress testing as part of a broader climate risk management toolkit are discussed in the conclusion.
Market discipline: stress testing is sufficiently credible and transparent to promote market discipline by enabling more accurate pricing of climate risks and incentivizing improvements in tested firms’ climate risk management practices
The first risk management criterion relates to whether the practice of stress testing and associated disclosures lead to improved climate risk management via market discipline. A key benefit of stress testing is that it enables market participants to better assess firms’ risks and risk management approaches, which in turn incentivizes tested firms to improve performance and processes [55, 69]. For firms subject to climate stress tests, preparing for and participating in stress testing—which requires management articulation of and firmwide visibility into climate risks—may enhance board and management understanding of and accountability for climate risks, particularly when there is an expectation that results will be disclosed. With respect to markets, the disclosure of stress testing results can provide greater transparency around participating firms’—and potentially their major counterparties’—exposures to physical and transition risks and the efficacy of their risk management approaches. If this disclosure facilitates more accurate market pricing of climate risks, it may mitigate the adverse effects of a sudden policy-based price correction and the resulting financial stability consequences of a disorderly transition [24]. Similarly, the disclosure of stress test results may drive shareholder actions to enhance participating firms’ climate risk management practices.
There is a longstanding literature on the role of stress testing results disclosure on market behavior across institutional contexts [91], and more recent empirical work highlights the power of climate disclosures to shape market behavior [92]. However, these market disciplining effects will only be observed if stress tests are credible and transparent, and the resulting disclosures are decision-relevant. Prior analyses suggest that disclosures of climate-related financial risks are “incomplete, inconsistent, and insufficient,” contributing to the dual market failures—emissions externalities and information asymmetries—driving the mispricing of climate risk [86, 93, 94]. Disclosure of stress test results may promote market discipline if it enables more complete, consistent, and comparable accounting of climate-related financial risk. Moreover, while advocating for systems analysis to measure climate-related systemic financial risk, disclosure of such risks must necessarily occur at a more granular level to catalyze stakeholder actions. For example, firm-level reporting may enable market assessments of institutional resilience to climate change whereas reporting at a more granular level would be necessary to drive repricing of climate risks for financial assets [64].
Mitigation evaluation: stress testing enables tested firms to articulate and assess the effectiveness of alternative strategies to increase resilience to climate-related risks
The second risk management criterion relates to the extent to which stress testing enables tested firms to develop more effective and efficient climate risk management strategies. As described above, stress testing may incentivize firms to improve risk management—because participating firms gain a better understanding of their risks and anticipate market responses to the disclosure of results—but it may also enable participating firms to assess the effectiveness of alternative climate risk mitigation strategies. Such analysis may occur as part of routine stress testing via dynamic balance sheets or may be captured through a second-round or parallel analysis of risk management responses.
A wide range of climate-related financial risk management strategies are emerging, and stress testing can serve as a tool to quantitatively and qualitatively evaluate the effectiveness of alternative approaches. Articulating mitigation strategies—ranging from analyzing and pricing exposures (e.g., climate risk premia) to redirecting capital (e.g., divestment) to hedging (e.g., insurance) to capacity building—within the context of a stress test may enable both evaluation and demonstration of the resilience of selected strategies to physical and transition risks. For example, scenarios could be rerun with mitigations to identify their cost effectiveness relative to predetermined performance objectives or to evaluate the costs and benefits associated with a given mitigation [96]. A recent survey (n = 71) suggests that while myriad financial services firms are beginning to incorporate climate change into firm-run scenario analyses, such assessments are seldom used to inform risk management strategies [97]. This finding suggests an unrealized opportunity for firms to use climate stress testing to inform integrated strategies to mitigate their own climate-related financial risks as well as opportunities related to broader climate mitigation, adaptation, and monitoring efforts (e.g., investing in transition-enabling technologies, disclosing climate risk analyses). It should be noted that optimal climate risk management for individual firms may not be optimal for financial stability and thus regulators should evaluate potential mitigations with respect to impacts on both firm and system resilience. For example, recent studies highlight the potential for protection gaps as EU insurers improve climate risk measurement capabilities and climate-related risk transfers in US mortgage markets [98, 99], this shifting of liabilities to households or taxpayers via government insurers of last resort could have systemic implications in a future with increasingly severe and widespread climate shocks. Thus, while financial firms can leverage climate stress testing to develop innovative strategies to manage their own climate-related risks, firm action alone does not enable complete understanding of system-level impacts or identification of socially optimal risk mitigation strategies.
Evidence-based regulation: stress testing provides evidence to regulators that informs the design and implementation of microprudential and macroprudential regulation to better manage the accumulation and propagation of climate-related systemic risk
The third risk management criterion relates to the role of stress testing in the design and implementation of evidence-based regulatory policy. As with stress testing more broadly, climate stress testing can inform risk management by enabling the calibration and enforcement of microprudential and macroprudential standards and supporting monitoring of the accumulation and propagation of climate-related risk throughout the financial system. While the literature suggests that linking stress tests to remedial actions is critical to promoting their effectiveness [55], doing so requires a relatively high degree of confidence in the risk measurement strategy. A debate in the literature is emerging regarding the comparative merits of building precise measurement strategies versus taking a more precautionary approach given the radical uncertainty associated with climate-related financial risk [100, 101].
Various microprudential and macroprudential tools to address climate risks have been suggested, including green-supporting or sectoral risk weights for capital requirements, concentration limits, carbon-countercyclical capital buffers, green supporting margin requirements, and enhanced climate risk disclosure [4, 49, 103, 104]. However, regulators are in the early stages of considering these measures; for example, a 2020 survey of central banks (n = 27) found that the majority had not yet considered including climate-related financial risks in prudential capital frameworks [6]. As microprudential and macroprudential regulations to bolster financial firm and sector resilience to climate change mature, stress testing may both inform design and enable implementation [14]. For example, the ECB has recently noted that while its climate stress test will not have a capital hurdle rate, it may inform the supervisory process by which it sets banks’ capital add-ons pursuant to Basel’s Pillar II [106, 107]. Similarly, the BOE is using the results of its climate stress test to assess alignment with climate risk management supervisory expectations and is also exploring how underlying capabilities and data might be matured to support ongoing work on capital frameworks for climate risks [108].
In addition to informing microprudential and macroprudential regulation, stress testing may facilitate better understanding of the accumulation and propagation of climate-related risk throughout the financial system. Indeed, by incorporating tail risk scenarios, modeling endogenous drivers of financial instability, and exploring strategies to bolster system resilience, macroprudential climate stress testing could support a precautionary approach to climate risk. As described above, stress testing can help structure thinking among various stakeholders, including integrating the more microprudential risk management perspectives of firms and market regulators with the more macroprudential perspectives of central banks and financial stability regulators. The process of developing stress tests—particularly with coordination across financial regulators—could provide legibility into the pathways by which climate-related financial risk becomes systemic, and thus may enable broader macroprudential strategies by elucidating systemically important institutions and activities.
Climate stress testing practices
Stress testing has emerged as one of the most prevalent regulatory tools for addressing climate-related financial risks [6]. As noted in the introduction, myriad governmental and intergovernmental organizations have called for the institutionalization of climate stress tests and a 2019 survey of central banks and regulators (n = 33) found that 84% had already incorporated or planned to incorporate climate change into stress tests within the next three years [2]. In the years since, notwithstanding the pandemic and attendant economic and health crises, regulators have continued to advance climate stress testing efforts [2, 5, 72]. This section takes stock of cross-national policy proposals and early practices and provides a preliminary application of the evaluative framework for climate stress testing in comparative case studies of effectiveness—for a leading jurisdiction, the UK—and capacity—for a lagging jurisdiction, the US.
Climate stress testing proposals and early practices
Central banks and regulators globally are developing scenario analyses and stress tests that incorporate climate-related financial risks [2, 5, 72]. Table 2 provides a summary of progress among these leading authorities, organized by jurisdiction, entity, year of (proposed) exercise, scope of climate risks and financial institutions covered, and exercise type.3 Discussed in more detail below, despite widespread interest in stress testing, progress to date has been uneven across jurisdictions, and even among leading jurisdictions, existing stress testing practices are insufficient to measure and manage the effects of climate change on financial stability. In particular, the institutionalization of systematic stress testing practices is nascent as is the operationalization of systemic risk within climate stress tests.Table 2 Selected climate stress testing practices and proposals
Jurisdiction (authority) Climate-related financial risk scenario analyses and stress tests
United Kingdom (Bank of England/Prudential Regulation Authority) 2017 scenario analysis of transition risk for UK financial markets [109]
2019 pilot stress test to assess physical and transition risks for UK insurers [110, 111]
2021 systematic climate stress test to assess physical and transition risks for large UK banks and insurers [108, 112, 113]
European Union (European Central Bank/European Systemic Risk Board/European Banking Authority/European Insurance and Occupational Pensions Authority/European Securities and Markets Authority) 2020 ESMA scenario analysis of transition risk for EU investors [115]
2021 ECB economy-wide scenario analysis of physical and transition risks for EU banks, insurers, and investors [98, 116]
2021 ESRB/ECB parallel scenario analyses of physical and transition risks for large EU banks and investors [94, 95]
2022 ECB systematic climate stress test to assess physical and transition risks for large EU banks [106, 107, 117, 118]
France (Banque de France/Autorité de Contrôle Prudentiel et de Résolution) 2020 voluntary pilot stress test to assess transition risk for large FR banks and physical and transition risks for large FR insurers [119, 120]
Canada (Bank of Canada/ Office of the Superintendent of Financial Institutions) 2020 economy-wide scenario analysis of physical and transition risks for CA financial markets [121]
2021 voluntary pilot stress test to assess physical and transition risks for selected CA banks and insurers [122]
Australia (Australian Prudential Regulatory Authority/Council of Financial Regulators) 2021 pilot stress test to assess physical and transition risks for large AU banks [123–125]
The Netherlands (De Nederlandsche Bank) 2017 scenario analysis of physical risk for NL banks, insurers, and investors [126]
2018 scenario analysis of transition risk for NL banks, insurers, and investors [127, 128]
Denmark (Danmarks Nationalbank) 2019 scenario analysis of physical risk for DK banks [129]
2020 scenario analysis of transition risk for DK banks [130]
Germany (Deutsche Bundesbank) 2020 scenario analysis of transition risk for DE banks, insurers, and investors [131]
2023 scenario analysis of physical and transition risks for DE banks, insurers, and investors announced [132]
Singapore (Monetary Authority of Singapore) 2018 industry-wide stress test included physical risk for SG insurers [133]
2022 stress test for SG financial firms announced [133]
Hong Kong (Hong Kong Monetary Authority) 2021 voluntary pilot stress test to assess physical and transition risks for large HK banks [134]
China (People’s Bank of China) 2021 pilot stress tests to assess transition risk for CN development and commercial banks [135]
Switzerland (Swiss Federal Office for the Environment/State Secretariat for International Finance) 2017 and 2020 scenario analyses of physical and transition risks for CH banks, insurers, and investors, conducted by external consultant [136]
United States (Federal Reserve) 2023 pilot scenario analysis for large US banks announced [137]
As Table 2 depicts, a growing number of central banks and regulators are conducting climate-related scenario analyses—which consist of a centralized exercise designed and conducted by a regulatory authority without direct participation by firms or results linked to remedial actions [20, 50]. Scenario analyses thus far have tended to focus on financial sector exposures to specific physical risks or financial sector sensitivity to transition risks within a given jurisdiction. For example, Denmark’s central bank conducted a study of mortgage lending risks resulting from projected sea-level rise in 2019 and a scenario analysis of the sensitivity of banks’ capital positions to an abrupt transition, characterized by a substantial impairment charge over a short timeframe, in 2020 [129, 130]. Similarly, the central bank of the Netherlands was an early leader in climate scenario analyses, assessing weather-related physical risks and disorderly energy transition risks via policy and technology shocks for banks, insurers, and pension funds in 2017 and 2018, respectively [127, 128]. Scenario analyses have leveraged diverse data sources (e.g., historical natural disaster data, routine supervisory data, bespoke surveys) to provide a high-level assessment of physical and transition risks [5, 21]. Scenario analyses can complement other climate risk research efforts and serve as a foundation for stress tests; however, tradeoffs in scope and granularity combined with the relatively static nature of these analyses suggest that they are insufficient to measure or manage climate-related financial risk.
Climate stress testing practices—which involve the application of supervisory scenario analysis to test the resilience of firms or sectors, generally with direct participation by tested entities to enable sufficient granularity and confidence to inform institution- or system-level responses or remedial actions [5, 21]—are less developed. Thus far, climate stress tests have consisted of exploratory or pilot exercises, often with voluntary firm participation, although more systematic stress tests are emerging in leading jurisdictions. The design of extant stress tests varies with respect to the physical and transition channels, financial risks, and institutions included. For example, France’s prudential regulator and central bank conducted a voluntary stress test pilot to assess the impacts of physical and transition risks for large French insurers and the impact of transition risk for large French banks in 2020 [119, 120, 140]. Like other climate stress tests, this pilot leveraged external scenarios (from the NGFS and IPCC) as well as prior central bank-led efforts, including a 2019 survey-based analysis of climate-related physical, transition, and liability risks for French banks and insurers [141–143]. The BOE was the first authority to conduct a compulsory climate stress test (discussed in more detail below) and the ECB recently published results from a systematic climate stress test [108, 118]. As with scenario analyses, stress tests have used external reference scenarios and coupled external integrated assessment and macroeconometric models with in-house financial models, but are differentiated by the translation from aggregate financial (sub)sector risk to firm-specific risk via a combination of both supervisory and firm-level modeling, which in several cases accounts for firm behavior (via a dynamic balance sheet) and cross-sector interactions. This combined approach enables these stress tests to provide a more granular and dynamic assessment of climate risk, firm behavior, and potential system implications than climate scenario analyses.
Several gaps emerge from this stocktake of climate stress test proposals and early practices. First, with respect to scope, most exercises have focused on a relatively narrow subset of climate-related financial risks—largely credit and, to a lesser extent, market risks for loan and investment portfolios, respectively—and affected firms—mostly banks, with some insurers and relatively few investment firms. While many exercises have included both physical and transition risks, analysis of risk interactions and interdependencies is limited. Moreover, although voluntary initiatives are critical for capacity building among both authorities and participating firms, there may be an adverse selection problem that limits the generalizability of results and validity of inferences about systemic risk and resilience. Second, with respect to coordination, while many exercises have leveraged common scenarios, the design of exercises has varied considerably across jurisdictions; heterogeneity in the design of stress testing pilots might enable learning across jurisdictions about the different types of climate-related financial risks, but lack of harmonization may also inhibit integration of cross-border impacts and coordination of financial regulatory responses. Third, with respect to financial stability, although many exercises have macroprudential objectives, the operationalization of systemic risk is limited to aggregating (partial) exposures, with inattention to second-round effects. There has been some qualitative assessment of tested firms and large counterparties’ potential mitigations and interactions—including the possibility of insurance protection gaps [112]—but modeling of risk amplification and financial stability implications is nascent [20, 21]. Fourth, exercises to date have largely focused on risk measurement, with most seeking to identify climate risks and assess the nature and magnitude of exposures for firms and sectors, rather than informing regulatory or firm risk management strategies.
Early progress suggests that institutionalizing climate stress testing is an iterative process. Jurisdictions leading the development and preliminary implementation of more systematic climate stress tests have experience conducting climate-related risk research, stakeholder engagement, scenario analyses, and stress test pilots, as well as issuing complementary supervisory guidance. For example, in advance of the Climate Biennial Exploratory Scenario (CBES), the BOE produced extensive research on climate-related financial risks [109, 144–147], conducted a survey of industry climate risk practices [148], issued a supervisory statement on climate risk [149], and launched a pilot climate stress test for the insurance sector [110, 111]. Similarly, in advance of the implementation of climate stress tests [117], European authorities partnered to conduct climate scenario analyses [94, 95, 98], publish supervisory guidance for climate risk analysis and disclosure [150], and review bank-led climate stress testing practices as part of alignment with supervisory expectations [74]. This iterative progress is promising given that many jurisdictions are reportedly engaging in these activities [2, 5]. For example, in addition to the jurisdictions depicted in Table 2, several other central banks are conducting research, collecting data, and engaging regulated entities and commercial service providers to assess climate risks (e.g., Norway [151, 152], Mexico [153], Italy [154–156], Spain [157], Japan [158]). However, it also underscores the importance of distinguishing capacity building exercises from systematic stress tests when evaluating the potential benefits vis-à-vis climate change and financial stability.
In summary, although there have been many stress testing policy proposals and increased use of scenario analyses among leading jurisdictions, the limited stress testing practices to date are insufficient to measure and manage climate-related systemic financial risk. Iterative progress is promising, however, inattention to systemic risk across existing analyses suggests a disconnect between policy motivation and implementation. While some tailoring by jurisdiction is necessary, greater coordination in stress testing design and implementation methodologies would enable a more holistic assessment of climate risks, including spillovers across jurisdictions, as well as coherence in regulatory strategies to address these risks. The detailed comparative case studies presented in the next section explore these themes in greater detail.
Framework application: comparative case studies of effectiveness and capacity
The BOE and FRB serve as illustrative comparative cases because of their similar institutional designs and divergent approaches to climate risk. As central banks of large economies with major financial centers, both the BOE and FRB have financial stability mandates that include regulating systemically important financial institutions. Both central banks conduct annual solvency stress tests for these institutions—which include multiple scenarios, countercyclical elements, and added shocks for particular firms—and use the results of these stress tests to calibrate capital requirements and monitor systemic risk. The scenarios designed by the BOE and FRB are also used in other supervisory and firm-run stress tests. While both stress testing regimes combine microprudential and macroprudential objectives, the BOE’s stress tests reflect a somewhat greater macroprudential focus: it uses stress tests to calibrate macroprudential policy (e.g., countercyclical capital buffer), runs a biennial exploratory scenario to evaluate novel risks, and has developed workstreams on system-wide amplification and feedback [43, 51]. It should also be noted that pursuant to US legislative and regulatory reforms, the design and scope of BOE and FRB stress tests have diverged more in recent years.
This section applies the evaluative framework for climate stress testing developed in the preceding section to the BOE and FRB. To preview the findings, which are also summarized in Table 3, the BOE is a leader in the implementation of climate stress testing and the design of its first climate stress test enables partial measurement of climate-related financial risks and financial stability implications. Noting that measurement is a necessary condition for effective and efficient risk management, this section identifies opportunities for the BOE to improve the measurement and expand to management of climate-related systemic risk. Notwithstanding its central role in global financial markets and advanced prudential regulatory regime, the US lags its international counterparts in confronting climate-related financial risks, including via the development of climate stress tests. However, the FRB has both the authority and the capacity to incorporate climate risks into its existing stress testing regime, and this section identifies potential strategies to do so.Table 3 Climate stress testing effectiveness and capacity
Criteria Bank of England US Federal Reserve
Scope Systemically important and materially climate-exposed financial institutions Large UK banks, building societies, and general and life insurers Largest and most complex US banks; scope could be expanded via FSOC designation and scenarios could inform broader US stress testing regime
Scenario Design and Modeling Approach Physical and transition risks with appropriate spatial and temporal resolution and robust representation of uncertainty and interdependencies Three scenarios combining physical and transition risks over a 30-year time horizon with 5-year reporting intervals and a static baseline balance sheet; separate analysis of litigation risks for general insurers Climate scenarios could be implemented as part of the comprehensive macro scenario or as an added component; necessitates the development of threat-specific narratives for climate risks and transmission pathways
Metrics Performance-based and decision-relevant metrics derived from quantitative and qualitative outputs Quantitative metrics include changes in banking book assets for banks and liabilities and assets for insurers; qualitative metrics include climate risk measurement and management assumptions and approaches for participants and selected counterparties Existing capital-based metrics could be used to assess performance, but disaggregated balance sheet impacts and broader risk management processes may be more informative; additional metrics should be considered as climate risk data, methodologies, and frameworks advance
Systems Analysis Firm-level analysis enables system-level modeling of the effects of climate change on financial stability via risk aggregation and amplification Facilitated by concurrent application and detailed reporting for banks and insurers, as well as follow-on data collection to assess system-wide interactions, inconsistencies, and financial stability implications Requires expansion, but the scope and concurrent application could facilitate analysis of second-round effects and the counterparty analysis could be expanded to improve the measurement of system-wide interactions and financial stability implications
Market Discipline Incentivizes improved risk management practices and enables price corrections via disclosure Increased climate risk awareness among participating firms via individual feedback and integration into supervisory assessment; potential for price corrections limited by aggregate disclosure Granular disclosure could facilitate market discipline among participating firms and their stakeholders
Mitigation Evaluation Firms articulate and assess the effectiveness of alternative strategies to increase resilience to climate risks Static balance sheet in the first round, but articulation of potential management actions in the second round and documentation of assumptions about counterparties’ actions Mitigations could be represented in capital planning via dynamic balance sheets, as well as in expanded use of the qualitative assessment to include climate risk management processes
Evidence-Based Regulation Supports the monitoring of climate-related systemic risk accumulation and propagation and informs the design and implementation of microprudential and macroprudential standards Monitoring the accumulation and propagation of climate-related systemic risk is an explicit purpose; not used for calibrating firm- or system-level capital standards (routine stress tests are used to calibrate these requirements), but will inform interagency and international cooperation on climate-related financial risk regulation Monitoring the accumulation and propagation of systemic risk is an explicit purpose (but routine stress tests are not used to calibrate macroprudential requirements); could be used to calibrate microprudential capital standards with sufficiently robust measurement approaches
Bank of England
The BOE began concurrent solvency stress testing in 2014 and conducts annual stress tests of the largest UK banks and building societies to assess the resilience of these financial institutions and the financial system. The BOE’s stress tests inform both firm and system-wide capital buffers. Via the Prudential Regulatory Authority, the BOE also conducts periodic solvency stress tests for selected insurance firms [51]. Beginning in 2017, it also began conducting a biennial exploratory exercise to evaluate novel risks, particularly those that may be decoupled from the normal financial cycle. The 2021 exploratory exercise, the CBES, focused on climate risk and represents the first systematic climate stress test.4 The stated purpose of the CBES was to “test the resilience of the current business models of the largest banks, insurers and the financial system to the physical and transition risks from climate change” [112 p.1]. Specifically, it was designed to provide an assessment of the magnitude of climate-related financial risks for large UK banks and insurers, the climate-related risk management and governance strategies of these firms, and the implications of climate change and financial sector responses for financial stability. This section assesses the CBES based on the criteria articulated in the evaluative framework for climate stress testing.
The CBES was applied to the UK’s largest banks and building societies as well as large general and life insurers. Participating banks covered 70% of UK residential and commercial lending, life insurers covered 65% of the UK market by asset size (with various business models), general insurers covered 60% of the UK market by premium size, and selected syndicates covered 40% of the Society of Lloyd’s property and liability insurance market by premium size [108]. The BOE integrated the bank and insurance stress tests to provide a more complete assessment of climate-related financial risks for the entire financial system, including potential interdependencies and risk-transfers (e.g., effects of changes in insurance provision on banks’ credit risks). Discussed in more detail below, the structure of the counterparty analysis also necessitated an assessment of climate exposures for many large non-financial UK companies [160].
The CBES included three scenarios with physical and transition risks for all participants, as well a separate analysis of liability risks for general insurers based on seven hypothetical legal cases [108]. The three primary CBES scenarios expanded on NGFS scenarios via the BOE’s work with climate scientists and other subject matter experts, and represented an early and orderly transition with limited physical risks (“early action”), a late and disorderly transition with limited physical risks (“late action”), and severe physical risks in the absence of a transition (“no additional action”). The magnitude of physical and transition risks within these scenarios was driven by the timing and stringency of climate policies, as described by scenario narratives that included temperature, emissions, and mitigation pathways, and resulting physical and transition risks. Physical risks were operationalized as changes in the frequency and severity of chronic and acute climate-driven natural hazards, based on global and regional temperature pathways, with sufficient granularity to represent geographic variation in these risks. Transition risks were operationalized as different combinations of climate policies, technological developments, and consumer preferences, with an emphasis on the effects of carbon pricing. These physical and transition risks were translated into macroeconomic and macrofinancial variables such as gross domestic product, unemployment, bond yields, and equity indices. The scenarios did not include a macroeconomic shock beyond those resulting from physical and transition risks [112]. The CBES employed a 30-year modeling horizon and reporting at five-year intervals, with the impact of climate risks based on comparisons to static 2020 balance sheets. Uncertainty with respect to the timing and severity of impacts was represented within and across scenarios, and via an additional sensitivity analysis in the second round. Although the scenarios were not predictive—but rather represented the potential macroeconomic and macrofinancial risks associated with alternative climate scenarios—the BOE reportedly plans to work with climate scientists to provide probabilities for the realizations of impacts aligned with each scenario and also asked tested firms to provide expectations around the likelihood of alternative scenarios. Moreover, for physical risks, firms were asked to model both the mean and tail of climate risk distributions, and the “no additional action” scenario was calibrated to reflect within-scenario realization of the more material risks anticipated outside of the scenario timeframe (i.e., risks anticipated in the period from 2050 to 2080 occurred by 2050 within the scenario) [112].
The CBES had both microprudential and macroprudential goals, and the scenario design and modeling approach balanced the level of granularity needed to enable firms to meaningfully assess their idiosyncratic climate risks with the level of consistency needed to enable comparability and aggregation to assess system-level climate risk. To achieve this balance, the scenarios were applied concurrently and consistently across participants, with variables focused on the UK and key economies and operationalized at the regional and sectoral levels, and participants expanded the analysis via the inclusion of other economies and more granular counterparty impacts. Participants modeled the impacts of each scenario on their corporate, household, and government exposures. For corporate exposures, participants were also required to engage with counterparties and use publicly available data, such as TCFD disclosures, to analyze how a given scenario would affect the counterparty, the counterparty’s vulnerability (inclusive of any planned mitigations), and the resulting financial impacts; counterparty impacts were compared to scenario averages to ensure coherence or explore divergence. This more detailed counterparty analysis was originally proposed for 80% of nominal exposures to corporates with climate-exposed assets, but was subsequently revised to cover at least the top 100 non-financial corporate exposures, the three largest companies in the sectors most impacted by the scenarios (if not in top 100), and the five largest exposures [161]. A less granular approach was permitted for the remaining counterparties, such as extrapolation from sectoral indicators.
The CBES included quantitative metrics to evaluate the impact of climate scenarios on participants’ balance sheets and business models—accounting for the “direct and indirect impacts of climate-related financial risks, as well as the mitigation and adaptation plans of counterparties” [112 p.9]—and a qualitative questionnaire to assess participants’ climate risk modeling and management approaches. The quantitative metrics used in the CBES are similar to those employed in the BOE’s insurance and banking stress tests—insurers were asked to report on both liabilities and assets (with an emphasis on the value of invested assets and insurance claims), while banks were asked to report on assets (with an emphasis on detailed credit risk analysis for large corporate counterparties) but were not required to provide detailed modeling of liabilities or income statements. In the initial proposal, banks were required to report on assets for both their banking books—via annual and cumulative impairment charges—and their trading books—via the change in fair value of assets—however, the BOE decided to exclude the trading book in light of feedback about the dynamic nature of these risks [160, 161]. While the standard balance sheet metrics are not climate-specific, the comparison to the 2020 baseline, absence of additional macroeconomic shocks, and detailed reporting requirements enable at least partial isolation of climate impacts. For example, exposures in key regions were reported at the sub-national level and sectors with the greatest vulnerability were also reported at more granular levels. Insurers were instructed to disaggregate the values of assets and liabilities by natural hazard and territory within each scenario. For the top 50 counterparties, participants also provided a detailed breakdown of modeling approaches and assumptions, including assumptions about counterparties’ mitigation and adaptation strategies. The original proposal also included an assessment of the temperature alignment of participants’ portfolios based on an aggregation of counterparty assessments, for example, by estimating global temperature increases given the emissions intensity and technology pathways of counterparty activities, however, the revised proposal excluded the temperature alignment assessment in response to concerns about data availability [112, 161]. While the initial analysis quantified results based on static balance sheets, participants also identified how and when they would modify business models in response to each scenario, leveraging a provided menu of management actions to address climate-related risks (e.g., pricing exposures) and opportunities (e.g., redirecting capital). Finally, participants also conducted a qualitative assessment of risks not captured elsewhere, including climate-related operational, reputational, and litigation risks, as well as risks arising from bank-insurer interactions [112].
Following the firm-level exercise, the BOE conducted an analysis of “system-wide impacts and inconsistencies,” which included a second-round data collection. This systems analysis was enabled by the design of the firm-level stress test, including the detailed information collected on participants’ modeling approaches, data, and assumptions about climate risks and resilience [112]. The CBES proposal outlined how the systems analysis would explore climate-related systemic risk and the implications for financial stability via an assessment of the system-level impact of bank and insurer behaviors and interactions as well as an analysis of the plausibility and potential consequences of participants’ mitigation actions, both individually and in aggregate. The CBES proposal stated that this analysis would consider the potential for “spillovers across sectors, for behaviours to amplify the impact of the underlying climate shocks, and for material disruption to the provision of financial services to UK households and businesses” [112 p.9], leveraging firm responses on when, why, and how they would react to scenarios. Although detailed results of this analysis were not published, the CBES proposal outlined four key areas it would explore: changes in the provision/pricing of services to the real economy, inconsistencies in assumptions, fire sales, and capacity to support the transition [112]. The second-round data collection was designed to address inconsistencies identified in the first-round submission and provide participants with the opportunity to “respond to the aggregated results from the first round, and potentially revise parts of their submissions in response to Bank feedback” [112 p.25]. The results indicate that this data collection included firm responses on transition opportunities in various competitive landscapes, the impact and management responses if losses were double the initial projections, the methodologies for assessing the credibility of counterparties’ transition plans, and the envisaged role of public sector support for climate-vulnerable entities [108].
With respect to risk measurement, the results provide some insight into climate-related financial risks for large UK banks and insurers and the potential implications for financial sector stability, but interpretation is limited by the exercise design, modeling challenges, and data gaps. The results of the first-round analysis suggest that climate loss projections would be equivalent to a 10–15% drag on profits annually, but there is substantial uncertainty around the magnitude of these risks due to the limited scope of the exercise and “immaturity of firms’ approaches and the complexity of modelling the impact of these risks” [108 p.12]. For example, BOE staff analysis found that losses could be four times higher than UK insurers’ submitted estimates [108]. A close read of the results also highlights sectoral, geographic, and temporal concentrations that may be troubling from a financial stability perspective and are obscured by the cumulative average loss projections [108]. For example, carbon-intensive sectors account for a third of credit loss provisions in transition scenarios despite representing only 15% of banks’ corporate exposures, mortgage impairments are highly concentrated in regions prone to flooding, and 40% of losses occurred in the first five years of the transition in the late action scenario [108]. The results of the second-round analysis also highlight potential system-wide consequences, including the macroeconomic consequences of fire sales and credit supply shocks associated with a disorderly transition. The BOE appropriately cautioned that the validity of results is limited by the design of the exercise (e.g., exclusion of banks’ trading losses and life insurers’ mortality risks), the immaturity of climate financial risk modeling (e.g., uncertainty around climate loss projections and reliance on external providers), and data gaps (e.g., reliance on counterparties’ transition plans and lack of standardized information on corporate asset locations and value chain emissions) [108].
While the purpose of the CBES is risk measurement, there are several aspects of the exercise design that support risk management. With respect to market discipline, the BOE did not publish stress test results for individual firms, but participants received feedback on the strengths and weaknesses of submissions and the BOE used findings on climate risk management to assess alignment with climate risk supervisory expectations [108, 112, 149]. Thus, while the lack of firm-level disclosure limits the market disciplining effects of the CBES, the results state that it has driven improvements in firms’ risk management approaches by exposing data and modeling gaps [108]. With respect to informing firm risk management strategy, the articulation and analysis of potential mitigations—including management actions provided by the BOE, embedded in participants’ transition plans, and resulting from counterparty assessments—may contribute to better understanding of climate risks and potential responses. With respect to informing regulatory responses to climate risk, the results outline several ways in which the CBES and underlying capabilities will support work on microprudential and macroprudential tools to address climate risks at the national and international levels [108]. For example, the BOE is exploring the role of regulatory capital frameworks in bolstering firm and system resilience to the financial consequences of climate change, and has noted that stress testing could in principle inform the calibration of capital requirements for climate risk, but will require better data and more sophisticated models [102]. The results also describe how the CBES findings will inform ongoing work on climate-related financial stability policy issues (e.g., financial system resilience and real economy spillovers) [108]. Finally, the BOE plans to identify best practices and disseminate lessons learned within the UK government and across international peers to improve cross-sectoral modeling and reduce regulatory fragmentation, respectively [108].
Thus, the CBES is the first systematic climate stress test and represents important progress in the assessment of climate-related financial risk, however, given the identified limitations, BOE’s regulatory authority, and rapidly maturing analytical capabilities, there are opportunities to improve the measurement and expand to management of the effects of climate change on financial stability. The scope of the CBES was fairly broad, covering a substantial portion of financial activity in the banking and insurance sectors; however, the exclusion of investment firms and banks’ trading books is an impediment to assessing market risks [161]. The scenarios integrated physical and transition risks, were attentive to the distinctive temporal and spatial dynamics of climate risk and the attendant uncertainties, included at least some tail risks, and balanced the need for comparability at the system-level and decision-relevance at the firm-level. The included metrics provide a useful, albeit partial, assessment of climate-related financial risk for participants and their major counterparties. The proposed initial framework—which included a more complete accounting of banks’ climate risk (via change in fair value of trading assets) and transition readiness (via temperature-based portfolio alignment)—would have provided a more holistic assessment of participating firms’ resilience to physical and transition risks. As data concerns are addressed, alignment of stress testing metrics with current and emergent climate disclosure frameworks and capital requirements could maximize the utility of the exercise for both the BOE and participating firms [12, 108, 162–164]. The design of the CBES incorporates microprudential and macroprudential elements, using horizontal stress tests to size the climate exposures of banks and insurers and systems analysis to explore how such risks aggregate and interact, including risks resulting from bank-insurer interdependencies and herding in management responses. While a detailed description of the methodology and results of the systems analysis was not published, there are opportunities to leverage the systemic risk amplification modeling approaches in existing stress tests and ongoing development of analytical approaches for quantifying second-round effects and financial sector-real economy spillovers [43, 51]. Finally, as the BOE moves from risk measurement to risk management, stress testing may play a role in calibrating firm- and system-level capital buffers, but could also inform the broader implementation of the UK climate-finance agenda, such as mandatory TCFD-aligned climate risk disclosures [164, 165].
US Federal Reserve
In the US, stress tests were used to restore market confidence during the 2007–2009 global financial crisis and post-crisis legislation institutionalized peacetime stress tests [166]. The Dodd–Frank Wall Street Reform and Consumer Protection Act of 2010 includes stress testing authority as part of the FRB’s enhanced prudential supervision of certain banks and non-bank systemically important financial institutions as designated by the Financial Stability Oversight Council (FSOC) [167]. Legislative and regulatory reforms in recent years have reduced the frequency, scope, and stringency of stress tests, while changes to the FSOC’s designation processes have limited the applicability of enhanced prudential supervision for non-bank financial institutions [168, 169, 193]. Although other US financial regulators engage in stress testing, stress testing to support financial stability is uniquely within the purview of the FRB, as such, the focus of this analysis is on the design and implementation of supervisory stress tests within the Federal Reserve System.
The US lags its international counterparts in confronting climate-driven financial risks, although the regulatory landscape as it relates to these risks is rapidly evolving [9, 170–173]. The FRB has acknowledged climate change as a potential threat to the financial system in the past [174, 175], and more recently has connected climate change to its microprudential and macroprudential mandates in official publications [176–178], joined some 100 other central banks as members of the NGFS [179], formed working groups on supervisory and financial stability climate risk issues [180], and signaled its intention to conduct a pilot scenario analysis exercise for climate risk [180–137].5 This section considers whether and how climate-related systemic risk could be incorporated into the FRB’s stress testing regime, given the FRB’s financial stability regulatory authority and the design of existing stress tests.
The FRB has the authority to vary the scope and frequency of stress tests based on firm size and institutional characteristics, however, only the largest and most complex banks are currently subject to the FRB’s annual (or periodic) supervisory stress tests. While these banks represent a substantial share of the US financial sector, there may be other bank and non-bank financial firms that meet the criteria of material climate exposures and systemic importance [9, 182, 183]. Although these firms may perform company-run stress tests or undergo stress testing under the supervision of other regulatory authorities, their exclusion from the FRB’s stress testing regime means that their potential contributions to financial (in)stability, including vis-à-vis climate-related systemic risk, may be omitted from systems analysis. Notably, FSOC has the authority to designate non-bank financial institutions as systemically important, which brings these firms under the authority of the FRB’s enhanced prudential supervision, including stress testing. Moreover, the FRB has authority over financial market utilities, for which some have suggested climate stress testing, potentially building on derivative markets regulators’ reverse and liquidity stress testing for centralized clearing parties [10, 159, 184].
The FRB’s stress tests evaluate firm- and system-level resilience, with a focus on the capital adequacy of participating firms. The FRB develops scenarios of varying severity for each cycle, which are used in both supervisory and company-run stress tests and represent the effects of hypothetical adverse macroeconomic conditions via input variables related to domestic and international economic activity, asset prices, and interest rates. For the largest and most complex firms, two additional components are incorporated into the most severe scenarios: a global market shock—which simulates deteriorations in global markets that result in “general market distress and heightened uncertainty”—and a counterparty default—which simulates the effects of an “instantaneous and unexpected default” of the tested firm’s largest counterparty [187 p.7]. Existing scenarios are effectively threat agnostic, they reflect how adverse macroeconomic conditions stemming from some unspecified threat would affect key variables, however, the FRB has conducted threat-specific sensitivity analyses in response to COVID-19 [188]. Thus, given the design of current stress tests, climate risks could be incorporated via the comprehensive scenario, the global market shock or counterparty default additional components, or a new additional component. In the comprehensive scenario, the macroeconomic and macrofinancial impacts of climate change could be reflected via the existing input variables—for example, physical and transition risks could be represented by changes in commodity prices for climate-exposed and carbon-intensive sectors, respectively. Although the additional components are similarly threat-agnostic, they could be used to provide more specificity about transmission channels—for example, the global market shock component could be used to assess transition risk resulting from heterogeneity in cross-national climate mitigation policies, while the counterparty default component could be used to operationalize physical risk for counterparties that have the greatest climate exposures. The FRB also has the authority to require covered firms to include added components in stress tests based on “the company’s financial condition, size, complexity, risk profile, scope of operations, or activities, or risks to the US economy” [167]. As such, the FRB could develop a new additional component reflecting specific climate risks and transmission channels. Finally, although outside the scope of this analysis, climate risk might be considered in the context of the comprehensive liquidity assessment review; as noted above, the integration of solvency and liquidity stress testing is an active area of systemic risk research and policy discussion [43].
The FRB’s stress tests include a range of metrics to evaluate how each scenario affects firms’ balance sheets and capital positions over a nine-quarter planning horizon, with capital-based solvency metrics, such as the common equity tier 1 ratio, reported as the primary performance outcomes [73, 186, 187]. As part of stress tests, the FRB previously conducted a qualitative assessment of selected firms’ capital planning practices, which covered “governance, risk management, internal controls, capital policies, incorporating stressful conditions and events, and estimating impact on capital positions,” however, this assessment is now part of the standard confidential supervisory process [73 p.17]. Existing capital adequacy metrics could be used to gauge firm-level resilience to climate risks but would necessitate careful causal attribution to disaggregate the effects of physical and transition risks, as well as a taxonomy to guide the incorporation of climate change into asset risk weightings. A more incremental approach might entail first seeking to measure the impact of scenarios on assets and liabilities and then identifying the appropriate performance metrics to represent climate risks and guide risk management decisions. Prior analyses have noted that the statutory linking of stress testing and capital planning authorities means that any expansion of stress testing to address climate risks would necessarily need to preserve the focus on capital [159]. However, there are a broad range of firm risk management actions that stress tests might inform, only some of which would be captured by capital planning processes. Moreover, to maximize utility for both regulators and regulated entities, the quantitative metrics employed in stress testing could seek to align with those being developed in sustainable accounting methodologies and disclosure frameworks, including potential changes to mandatory disclosure frameworks at the US state and federal levels [12, 189, 190]. Finally, the qualitative assessment previously implemented in stress tests could potentially be expanded to incorporate not only the efficacy of capital planning but also of climate risk planning processes—including firms’ strategies for measuring and managing climate-related credit, market, liquidity, operational, and reputational risks.
Although the FRB’s stress tests are firm-level exercises, the FRB notes three related “macroprudential elements” that facilitate system-level analysis of financial stability issues [40 p.67]. First, the scope and severity of stress tests, which enables “examination of the loss-absorbing capacity of institutions under a common macroeconomic scenario that has features similar to the strains experienced in a severe recession” and reflects “salient risks” [40 p.67]. Second, the horizontal application of stress tests to the largest banks—representing some 80% of banking sector assets—which enables assessment of potential correlated exposures [40, 57]. Third, the inclusion of counterparty shocks, which enables evaluation of the effects of counterparty distress for the largest and most interconnected firms [40]. As such, embedding a sufficiently severe climate scenario in the FRB’s horizontal stress tests could enable analysis of climate-related systemic risk arising from banking sector correlated exposures and selected systemic risk amplification channels. The scope and design of the FRB’s stress tests facilitate estimation of the magnitude of direct effects for the banking subsector, but these aggregate estimates omit correlated exposures among other financial services subsectors. The counterparty default may enable assessment of systemic risk amplification resulting from counterparty contagion, but the scope of the added component is insufficient to fully represent second-round effects [57].
While it is premature to evaluate whether the inclusion of climate risk in the FRB’s stress tests would lead to improved management of those risks, it is possible to assess how existing stress tests may inform firm, regulatory, and market actions. With respect to market discipline, there is evidence that financial markets are responsive to stress test disclosures [4, 91], and the granularity of disclosures under existing stress tests could facilitate greater US market-based transparency about climate risks. However, the extent to which disclosures would promote market discipline depends on the design and credibility of stress tests and whether results transparency addresses deficiencies in existing voluntary climate risk disclosures. With respect to informing firm strategy, given the design of current stress tests, firms’ actions in anticipation of, or in response to, physical or transition risks could be represented in dynamic balance sheets and capital plans, although potential mitigation strategies may go beyond capital planning processes. Aligning stress testing requirements with current and emergent disclosure frameworks could also create incentives for firms to use the underlying scenario analysis to evaluate, and demonstrate, the resilience of their business strategies to climate risks, a key component of TCFD’s framework—of which most large US banks are supportersand for which compliance is “significantly lower” than any of the other recommended disclosures [191 p.4, 192]. With respect to informing regulatory responses, the FRB uses stress tests to calibrate capital requirements via the stress capital buffer (previously, capital distributions could be restricted based on the inadequacy of capital positions or planning processes [193]). As such, following the validation of its measurement strategy, the FRB could potentially use the results of climate stress tests to calibrate microprudential capital standards and to analyze the accumulation and propagation of climate-related systemic risk, which in turn might inform macroprudential responses. Additionally, FSOC, which currently employs an activities-based approach to analyze sources of financial instability, could use information from these stress tests to identify activities that increase climate-related financial risk or system vulnerability to risk amplification. Independent of these regulatory responses, climate stress testing could incentivize and inform more effective firm risk management strategies via market discipline and mitigation evaluation, respectively.
Thus, although the US lags its international counterparts in confronting climate-driven financial risks and developing climate stress tests, US financial regulators have the authority and the capacity to incorporate climate change into the existing stress testing regime. While the scope of current stress tests may be insufficient to fully capture the effects of climate change on financial stability, the FRB could incorporate climate change into stress testing for the largest and most complex banks, which are by definition systemically important. Additionally, the realization of proposals calling for FSOC’s incorporation of climate change into systemic risk monitoring and designation authority could expand the scope of the FRB’s stress tests [185]. Moreover, the FRB’s development of climate scenarios would facilitate the inclusion of climate-related financial risk in other financial regulator- and firm-run stress tests, with potential for coordination via the Federal Financial Institutions Examination Council (FFIEC) and Federal Insurance Office (FIO) [10, 169]. The FRB has several options for incorporating climate risk into existing scenarios, and there are opportunities for incremental progress and learning via interagency and international regulatory coordination. For example, the FRB could partner with other governmental and nongovernmental entities to leverage existing climate, natural hazard, and climate-economic modeling expertise to produce estimates of the effects of climate change on the macroeconomic variables routinely included in stress tests or to develop an additional component focused on climate risk (e.g., leveraging FSOC’s Climate Data and Analytics Hub [194]). Similarly, FRB research on climate change and financial stability and engagement in international regulatory fora—such as the NGFS, Basel Committee on Banking Supervision’s Task Force on Climate-Related Financial Risks, Financial Stability Board, G20 Sustainable Finance Study Group, and Central Banks’ and Supervisors’ Climate Training Alliance—could support the development of climate stress test scenarios. Metrics should be reassessed as US climate risk disclosure frameworks mature, but estimating the effects of climate change on assets and liabilities and incorporating climate risk management into a more holistic qualitative assessment are important first steps. Moreover, the FRB is reportedly collecting supervisory data on firm approaches to evaluating climate exposures, which may inform the development of metrics and scenarios; to maximize utility for macroprudential climate risk measurement, the FRB’s efforts around climate change should be coordinated with FFIEC and FIO to ensure banks and insurance companies regulated and supervised by other state and federal agencies and their specific regional risks are represented in ongoing data collection and analysis [9, 171]. The FRB’s existing approach to systems analysis could enable partial assessment of climate-related systemic risk, but systematic modeling of indirect effects arising from interactions among financial services subsectors and macrofinancial linkages are key gaps. The FRB has undertaken efforts to better represent systemic risks in stress tests—including exploring the incorporation of direct or system-wide liquidity shocks to understand liquidity-solvency interactions—and has developed sophisticated financial stability models for other institutional functions [43, 176, 177]. Moreover, there are opportunities to leverage these existing modeling efforts, as well as international progress and emergent academic research, as discussed above, to improve analyses of financial sector interactions and cross-sector interconnections [25, 90]. Thus, while developing and implementing climate stress testing is a substantial undertaking, this section identifies concrete steps the FRB could take to begin to assess microprudential and macroprudential climate-related financial risks and explores how a sufficiently rigorous risk measurement approach could enable US financial markets, firms, and regulators to improve climate risk management.
Conclusion
Despite widespread recognition among financial regulators and central banks that climate change may threaten financial stability, the causes and consequences of climate-related systemic financial risk remain underexplored. Stress testing has emerged as one of the most prevalent regulatory tools for addressing climate-related financial risks, and this article analyzes the role of stress testing in mitigating the effects of climate change on financial stability.
Stress testing can support the measurement and management of both microprudential and macroprudential climate-related financial risks. Drawing on the climate-related financial risk, financial stability, and stress testing literature, this article argues that stress testing is an essential tool for addressing financial risks, including potential systemic risks resulting from climate change. It identifies how the design of stress tests—including scope, scenarios, metrics, and systems analysis—could support the measurement of climate-related systemic financial risk. Noting that measurement is a necessary condition for effective and efficient risk management, this article also discusses how financial firms, regulators, and market participants could use stress tests to mitigate the effects of climate change on financial stability—via market discipline, mitigation evaluation, and evidence-based regulation. Moreover, by incorporating tail risk scenarios, modeling endogenous drivers of financial instability, and exploring strategies to bolster system resilience, macroprudential climate stress testing could also support a precautionary approach to climate risk. While the framework proposed in this article is necessarily ambitious, iterative approaches to climate stress testing can enable (at least partial) assessment of firm- and system-level climate exposures, provide insight into the propagation pathways and system characteristics that could result in the amplification of climate-related systemic risk, and structure thinking about potential responses among financial policymakers and market participants.
However, the benefits of stress testing vis-à-vis climate change and financial stability are unrealized. Notwithstanding sustained interest in stress testing among policymakers, progress to date has been uneven across jurisdictions, and even among leading jurisdictions, climate stress testing practices are insufficient to mitigate the effects of climate change on financial stability. In particular, the institutionalization of systematic stress testing practices is nascent, with leading authorities conducting one-off scenario analyses and largely voluntary pilot stress tests. The operationalization of systemic risk within existing climate stress tests is also limited, with most exercises focusing on a relatively narrow subset of microprudential risks and financial institutions, suggesting a disconnect between policy motivation and implementation. While the design of financial stability regulatory authorities and existing stress testing regimes shapes approaches to climate stress testing, institutional design alone does not explain the divergence between leading and lagging jurisdictions, as exemplified by the comparative case studies of the BOE and FRB.
Stress testing is an essential tool for better understanding, and potentially mitigating, the effects of climate change on financial stability, but it is not a panacea [57]. The literature highlights that although stress testing practices have become more robust in the decade following the global financial crisis, there continues to be room for improvement in the modeling of systemic risk and the linking of results to macroprudential policies; such challenges will persist with climate stress testing, which will also bring novel challenges. Forward-looking jurisdictions should approach the development of climate stress tests as an opportunity to improve the quality of systemic risk stress testing more broadly, including by evolving exogenous shocks to better reflect tail risks and by improving modeling of second-round effects to better reflect endogenous drivers of financial instability. Moreover, climate stress testing should be considered within the context of a broader regulatory strategy for addressing climate-related financial risks, which might also include other types of climate-related financial risk assessments, supervisory expectations for climate risk management, standardization of climate risk disclosure requirements, and capital frameworks for climate resilience. Together, these policies can mitigate the financial consequences of climate change, but they are not a substitute for macroeconomic and financial policies to address the causes of climate change [34, 102, 195].
Regulatory cooperation—within and across jurisdictions—is needed to advance climate stress testing and climate-related financial risk regulation more broadly. Interagency cooperation can support scenario development by leveraging and enabling translation across government-wide climate, natural hazard, macroeconomic, and financial modeling expertise. Such coordination is particularly important in light of recent research highlighting the misuse and misinterpretation of climate models for financial risk analyses [79]. Moreover, as governments take increasingly holistic approaches to climate policy, there may be opportunities to learn from financial sector experience. For example, stress testing might be extended to non-financial firms with substantial climate-related exposures to assess resilience to, and inform disclosures of, physical and transition risks [96, 196]. International regulatory cooperation will ensure coherence across microprudential and macroprudential responses to climate change, including stress testing [197]. The next phase of this research will explore how interagency coordination could inform the development of stress testing methodologies as well as opportunities for two-way learning as policymakers across all levels of government and sectors prioritize resilience to climate change. Future research will also include an expanded cross-national evaluation of climate stress testing practices and identification of international regulatory cooperation strategies to overcome market failures inherent to the provision of global public goods like environmental sustainability and financial stability.
Declaration
Conflict of interest
The author states that there is no conflict of interest.
1 Consistent with the literature, natural hazard is used to refer to the threat of a climate-driven natural phenomenon that will have a negative effect on humans, while the realization of that effect is referred to as a natural disaster.
2 There is variation in how the terms top-down and bottom-up are used in the context of stress testing. In Europe, they often refer to whether banks (i.e., bottom-up) or their supervisors (i.e., top-down) develop the models and scenarios and manage the implementation, whereas in the US they often refer to the level at which empirical relationships are estimated [52, 62].
3 A range of other entities are also conducting climate scenario analyses, such as the IMF via its Financial Sector Assessment Program, and climate scenario analyses are reportedly forthcoming in an even broader range of jurisdictions, including Austria, Brazil, Hungary, Italy, Japan, Korea, Malta, New Zealand, and Poland [19]. Table 2 is limited to stress testing practices implemented by financial regulators and central banks, and for which publications from the issuing authority were publicly available at the time of writing.
4 The first round of the CBES began in June 2021, the second round began in February 2022, and the results were published in May 2022.
5 US state-level insurance regulators have also used scenario analysis to explore climate risks—in 2018 and 2021, the California State Insurance Commission and New York Department of Financial Services, respectively, worked with an external consultant to conduct scenario analyses of transition risk for insurers within their states [138, 139].
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
1. Carney M Breaking the Tragedy of The Horizon – Climate Change and Financial Stability 2015 London Lloyd’s of London
2. Official Monetary and Financial Institutions Forum Tackling Climate Change: The Role of Banking Regulation and Supervision 2020 London Official Monetary and Financial Institutions Forum
3. Intergovernmental Panel on Climate Change Climate Change 2021: The Physical Science Basis Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change 2021 Geneva United Nations Intergovernmental Panel on Climate Change
4. Campiglio E Dafermos Y Monnin P Ryan-Collins J Schotten G Tanaka M Climate Change Challenges for Central Banks and Financial Regulators Nature Climate Change. 2018 8 6 462 468 10.1038/s41558-018-0175-0
5. Financial Stability Board Stocktake of Financial Authorities’ Experience in Including Physical and Transition Climate Risks as Part of Their Financial Stability Monitoring 2020 Basel Financial Stability Board
6. Basel Committee on Banking Supervision Climate-Related Financial Risks: A Survey on Current Initiatives 2020 Basel Bank for International Settlements
7. Climate Change Financial Risk Act of 2019 (H.R.5194).
8. National Academies of Sciences Engineering and Medicine Accelerating Decarbonization of the US Energy System 2021 Washington, DC The National Academies Press
9. Financial Stability Oversight Council FSOC Report on Climate-Related Financial Risk 2021 Washington, DC US Department of the Treasury
10. Climate-Related Market Risk Subcommittee of the Market Risk Advisory Committee Managing Climate Risk in the US Financial System 2020 Washington, D.C. Commodity Futures Trading Commission
11. European Systemic Risk Board Advisory Scientific Committee Too Late, Too Sudden: Transition to a Low-Carbon Economy and Systemic Risk 2016 Frankfurt European Systemic Risk Board
12. Task Force on Climate-related Financial Disclosures Recommendations of the Task Force on Climate-related Financial Disclosures 2017 Basel Financial Stability Board Bank for International Settlements
13. Task Force on Climate-related Financial Disclosures Technical Supplement: The Use of Scenario Analysis in Disclosure of Climate-Related Risks and Opportunities 2017 Basel Financial Stability Board, Bank for International Settlements
14. Network for Greening the Financial System A Call for Action: Climate Change as a Source of Financial Risk 2019 Paris NGFS Secretariat, Banque de France
15. Network for Greening the Financial System. 2020. Guide for Supervisors Integrating Climate-Related and Environmental Risks into Prudential Supervision. Paris: Banque de France.
16. International Monetary Fund Global Financial Stability Report: Lower for Longer 2019 Washington, DC International Monetary Fund
17. Basel Committee on Banking Supervision. 2022. Principles For the Effective Management and Supervision of Climate- Related Financial Risks. Basel: Bank for International Settlements.
18. BlackRock Financial Markets Advisory for the European Commission Development of Tools and Mechanisms for the Integration of ESG Factors into the EU Banking Prudential Framework and into Banks' Business Strategies and Investment Policies 2021 Brussels European Commission
19. European Systemic Risk Board Advisory Scientific Committee. 2021. Climate-Related Risk and Financial Stability. Frankfurt: European Central Bank.
20. Basel Committee on Banking Supervision. 2021. Climate-Related Financial Risks – Measurement Methodologies. Basel: Bank for International Settlements.
21. Baudino P Svoronos J-P Stress-Testing Banks for Climate Change – A Comparison of Practices 2021 Basel FSI Insights on Policy Implementation Bank for International Settlements
22. Basel Committee on Banking Supervision Climate-Related Risk Drivers and Their Transmission Channels 2021 Basel Bank for International Settlements
23. Battiston S Dafermos Y Monasterolo I Climate Risks and Financial Stability Journal of Financial Stability. 2021 54 100867 10.1016/j.jfs.2021.100867
24. Giglio S Kelly B Stroebel J Climate Finance Annual Review of Financial Economics 2020 13 15 36 10.1146/annurev-financial-102620-103311
25. Battiston S Mandel A Monasterolo I Schütze F Visentin G A Climate Stress-Test of the Financial System Nature Climate Change. 2017 7 283 290 10.1038/nclimate3255
26. Jung H Engle R Berner R Climate Stress Testing 2021 New York City Federal Reserve Bank of New York Staff Report
27. Network for Greening the Financial System Macroeconomic and Financial Stability Implications of Climate Change 2019 Paris NGFS Secretariat, Banque de France
28. Reinders HJ Schoenmaker D Van Dijk MA A Finance Approach to Climate Stress Testing 2020 Basel Bank for International Settlements
29. Auffhammer M Quantifying Economic Damages from Climate Change Journal of Economic Perspectives. 2018 32 3 33 52 10.1257/jep.32.4.33
30. Auffhammer M Hsiang SM Schlenker W Sobel A Using Weather Data and Climate Model Output in Economic Analyses of Climate Change Review of Environmental Economics and Policy. 2013 7 181 198 10.1093/reep/ret016
31. Hsiang S Kopp R An Economist’s Guide to Climate Change Science Journal of Economic Perspectives. 2018 32 3 3 32 10.1257/jep.32.4.3
32. Arregui N Chen R Ebeke CH Frie J-M Macia DG Iakova DM Sectoral Policies for Climate Change Mitigation in the EU 2020 Washington, D.C. International Monetary Fund
33. Environmental Protection Agency. 2020. Greenhouse Gas Emissions. https://www.epa.gov/ghgemissions
34. Krogstrup S Oman W Macroeconomic and Financial Policies for Climate Change Mitigation: A Review of the Literature 2019 Washington, DC International Monetary Fund
35. Dietz S Bowen A Dixon C Gradwell P 'Climate Value at Risk’ of Global Financial Assets Nature Climate Change. 2016 6 676 679 10.1038/nclimate2972
36. Bolton P Despres M Pereira Da Silva LA Samama F Svartzman R The Green Swan: Central Banking and Financial Stability in the Age of Climate Change 2020 Basel Bank for International Settlements
37. Task Force on Climate-related Financial Disclosures. 2020. Risk Management Disclosure: How the Organization Identifies, Assesses, and Manages Climate-Related Risks. https://www.tcfdhub.org/risk-management/.
38. Schwarcz SL Systemic Risk Georgetown Law Journal. 2008 97 1 193 249
39. Board Sustainability Accounting Standards Climate Risk Technical Bulletin 2016 San Francisco, CA Sustainability Accounting Standards Board
40. Federal Reserve. 2016. The Federal Reserve System: Purposes & Functions, 10th ed. Washington, DC: Board of Governors of the Federal Reserve System.
41. Schinasi GJ Defining Financial Stability 2006 Washington, DC International Monetary Fund
42. Adrian T Covitz D Liang N Financial Stability Monitoring 2015 New York Annual Review of Financial Economics. Federal Reserve Bank
43. Anderson R Baba C Danielsson J Das US Kang H Segoviano M Macroprudential Stress Tests and Policies: Searching for Robust and Implementable Frameworks 2018 Washington, DC International Monetary Fund
44. Cortés KR Strahan PE Tracing Out Capital Flows: How Financially Integrated Banks Respond to Natural Disasters Journal of Financial Economics. 2017 125 1 182 199 10.1016/j.jfineco.2017.04.011
45. Ivanov I Macchiavelli M Santos JAC Bank Lending Networks and the Propagation of Natural Disasters Financial Management 2020 80 1 44
46. Bolton P Kacperczyk M Global Pricing of Carbon-Transition Risk 2021 Cambridge National Bureau of Economic Research
47. Atanasova C Schwartz ES Stranded Fossil Fuel Reserves and Firm Value 2019 Cambridge National Bureau of Economic Research
48. van der Ploeg F Rezai A Stranded Assets in the Transition to a Carbon-Free Economy Annual Review of Resource Economics. 2020 12 281 298 10.1146/annurev-resource-110519-040938
49. Steele G Confronting the “Climate Lehman Moment”: The Case for Macroprudential Climate Regulation Cornell Journal of Law and Public Policy, New York 2020 30 109 9
50. Marcelo A Rodríguez A Trucharte C Stress Tests and Their Contribution to Financial Stability Journal of Banking Regulation. 2008 9 65 81 10.1057/jbr.2008.1
51. Dent K Westwood Ben Stress Testing of Banks: An Introduction 2016 London Bank of England
52. Borio C Drehmann M Tsatsaronis K Stress-Testing Macro Stress Testing: Does It Live Up to Expectations? 2012 Basel Bank for International Settlements
53. Committee on the Global Financial System. 2000. Stress Testing by Large Financial Institutions: Current Practice and Aggregation Issues. Basel: Bank for International Settlements.
54. Wall L The Adoption of Stress Testing: Why the Basel Capital Measures Were Not Enough Journal of Banking Regulation. 2014 15 266 276 10.1057/jbr.2014.10
55. Goldstein M Banking's Final Exam: Stress Testing and Bank-Capital Reform 2017 Washington, DC Peterson Institute for International Economics Press
56. Schuermann, T. 2016. Stress Testing in Wartime and in Peacetime. In Stress Testing and Macroprudential Regulation a Transatlantic Assessment. ed. Anderson, R. London: Center for Economic and Policy Research, pp. 125–39.
57. Baudino P Goetschmann R Henry J Taniguchi K Zhu W Stress-Testing Banks – A Comparative Analysis 2018 Basel FSI Insights on Policy Implementation. Bank for International Settlements
58. Bologna P Segura A Integrating Stress Tests within the Basel III Capital Framework: A Macroprudentially Coherent Approach Journal of Financial Regulation. 2017 3 159 186 10.1093/jfr/fjx004
59. Bank of England The Bank of England’s Approach to Stress Testing the UK Banking System 2015 London The Bank of England
60. Ebner A The Financial Stability Aspects of the EU-wide Stress Test Journal of Financial Regulation. 2018 4 2 326 336
61. Basel Committee on Banking Supervision Stress Testing Principles 2018 Basel Bank for International Settlements
62. Kapinos, P.S., C. Martin, and O. A. Mitnik. 2018. Stress Testing Banks: Whence and Whither? Journal of Financial Perspectives 5(1).
63. Hong, H., D. Wu, and Q. Wang. 2018. A Common Pitfall in Bank Stress Testing: Macroeconomic Factors and Model Instability. https://ssrn.com/abstract=2896974
64. Batten, S., R. Sowerbutts, and M. Tanaka. Climate Change: Macroeconomic Impact and Implications for Monetary Policy. In Ecological, Societal, and Technological Risks and the Financial Sector, eds. Walker, T., D. Gramlich, M. Bitar, and P. Fardnia. London: Palgrave Macmillan, pp. 39-68.
65. Battiston S Caldarelli G Risk MDS Leveraging the Network: A Stress-Test Framework Based on DebtRank Statistics & Risk Modeling. 2016 33 117 138 10.1515/strm-2015-0005
66. Battiston S Farmer JD Flache A Garlaschelli D Haldane AG Heesterbeek H Complexity theory and financial regulation Science 2016 351 6275 818 819 10.1126/science.aad0299 26912882
67. Adrian T Morsink J Schumacher L Stress Testing at the IMF 2020 Washington, D.C. International Monetary Fund
68. Bernanke, B. 2013. Stress Testing Banks: What Have We Learned? Speech to Maintaining Financial Stability Holding a Tiger by the Tail Conference Sponsored by the Federal Reserve Bank of Atlanta. Stone Mountain, 8 April.
69. Kohn, D. 2020. Stress Tests: A Policymaker’s Perspective. Speech to the Macroprudential Stress-Testing Conference hosted by the European Central Bank. Frankfurt, 5 February 5.
70. Feldberg, G., A. Metrick. 2021. Stress Tests and Policy, Stress Testing a Discussion and Review. The Journal of Financial Crises 3(1): 1-19.
71. Basel Committee on Banking Supervision Global Systemically Important Banks: Revised Assessment Methodology and The Higher Loss Absorbency Requirement 2018 Basel Bank for International Settlements
72. Network for Greening the Financial System Guide to Climate Scenario Analysis for Central Banks and Supervisors 2020 Paris NGFS Secretariat, Banque de France
73. Federal Reserve. 2019. Comprehensive Capital Analysis and Review 2019: Assessment Framework and Results. Washington, DC: Board of Governors of the Federal Reserve System.
74. European Central Bank Guide on Climate-Related and Environmental Risks: Supervisory Expectations Relating to Risk Management and Disclosure 2020 Frankfurt European Central Bank
75. De Nederlandsche Bank. 2018. An Energy Transition Risk Stress Test for the Financial System of the Netherlands. Amsterdam: De Nederlandsche Bank.
76. Batten, S., R. Sowerbutts, and M. Tanaka. 2016. Let’s Talk About the Weather: The Impact of Climate Change on Central Banks. London: Bank of England.
77. UN Environment Programme Finance Initiative. 2021. UNEP FI TCFD Banking Pilot Projects. https://www.unepfi.org/climate-change/tcfd/tcfd-for-banks/
78. Network for the Greening of the Financial System NGFS Scenarios for Central Banks and Supervisors 2022 Paris Network for the Greening of the Financial System
79. Fiedler T Pitman AJ Mackenzie K Wood N Jakob C Perkins-Kirkpatrick SE Business Risk and the Emergence of Climate Analytics Nature Climate Change. 2021 11 87 94 10.1038/s41558-020-00984-6
80. Newell RG Pizer WA Uncertain discount rates in climate policy analysis Energy Policy 2004 32 4 519 529 10.1016/S0301-4215(03)00153-8
81. Nordhaus WD A Review of the Stern Review on the Economics of Climate Change Journal of Economic Literature. 2007 45 686 702 10.1257/jel.45.3.686
82. EU Technical Expert Group on Sustainable Finance. 2020. Taxonomy: Final Report of the Technical Expert Group on Sustainable Finance. Brussels: European Commission.
83. Value Reporting Foundation. 2021. IFRS Foundation Announces International Sustainability Standards Board. https://www.valuereportingfoundation.org/news/ifrs-foundation-announcement/
84. Carbon Disclosure Project, Climate Disclosure Standards Board, Global Reporting Initiative, International Integrated Reporting Council, Sustainability Accounting Standards Board. 2020. Statement of Intent to Work Together Towards Comprehensive Corporate Reporting. London: Value Reporting Foundation.
85. Institute of International Finance Climate-related Financial Disclosures: Examples of Leading Practices in TCFD Reporting by Financial Firms 2019 Washington, DC The Institute of International Finance
86. Board Sustainability Accounting Standards The State of Disclosure Report 2017 San Francisco Sustainability Accounting Standards Board
87. Science Based Targets Initiative. 2022. Financial Sector Science-Based Targets Guidance Version 1.1. London: Science Based Targets Initiative.
88. Two Degrees Investing Initiative. 2020. Paris Agreement Capital Transition Assessment. https://www.transitionmonitor.com/
89. Carbon Delta. 2019. Climate Value-at-Risk (CVaR) Model. https://www.carbon-delta.com/climate-value-at-risk
90. Aymanns C Farmer JD Kleinnijenhuis AM Wetzer T Hommes C LeBaron B Models of Financial Stability and Their Application in Stress Tests Handbook of Computational Economics 2018 London Elsevier 329 91
91. Flannery M Hirtle B Kovner A Evaluating the Information in the Federal Reserve Stress Tests 2015 New York City Federal Reserve Bank of New York
92. Mésonnier J-S Nguyen B Showing Off Cleaner Hands: Mandatory Climate-Related Disclosure by Financial Institutions and the Financing of Fossil Energy 2021 Paris Banque de France
93. Bingler, J.A., M. Kraus, M. Leippold. 2021. Cheap Talk and Cherry-Picking: What ClimateBert has to say on Corporate Climate Risk Disclosures. https://ssrn.com/abstract=3796152
94. European Systemic Risk Board Advisory Scientific Committee Positively Green: Measuring Climate Change Risks to Financial Stability 2020 Frankfurt European Systemic Risk Board 2020
95. European Central Bank/European Systemic Risk Board Project Team on Climate Risk Monitoring Climate-Related Risk and Financial Stability 2021 Frankfurt European Central Bank
96. DeMenno, M.B., R. Broderick, and R. Jeffers. 2021. From Systemic Financial Risk to Grid Resilience: Embedding Stress Testing in Electric Utility Investment Strategies and Regulatory Processes. Sustainable and Resilient Infrastructure 7(6): 673–694.
97. GARP Risk Institute Second Annual Global Survey of Climate Risk Management at Financial Firms 2020 Jersey City The Global Association of Risk Professionals
98. Giuzio M Krusec D Levels A Melo AS Mikkonen K Radulova P Climate Change and Financial Stability Financial Stability Review 2019 Frankfurt European Central Bank 120 33
99. Ouazad, A., and M. E. Kahn. 2019. Mortgage Finance in the Face of Rising Climate Risk. Cambridge: National Bureau of Economic Research.
100. Cullen J After, “HLEG”: EU Banks, Climate Change Abatement and the Precautionary Principle Cambridge Yearbook of European Legal Studies. 2018 20 61 87 10.1017/cel.2018.7
101. Chenet H Ryan-Collins J van Lerven F Finance, Climate-Change and Radical Uncertainty: Towards a Precautionary Approach to Financial Policy Ecological Economics. 2021 183 106957 10.1016/j.ecolecon.2021.106957
102. Bank of England Climate-Related Financial Risk Management and the Role of Capital Requirements 2021 London Bank of England
103. Ceres. 2020. Addressing Climate as a Systemic Risk: A Call to Action for US Financial Regulators. Boston: Ceres.
104. Lastra, R., and K. Alexander. 2020. The ECB Mandate: Perspectives on Sustainability and Solidarity. Luxembourg: European Parliament Policy Department for Economic, Scientific and Quality of Life Policies.
105. Financial Stability Board. 2020. The Implications of Climate Change for Financial Stability. Basel: Financial Stability Board.
106. European Central Bank. 2021. Information on Participation in the 2022 ECB Climate Risk Stress Test. Frankfurt: European Central Bank.
107. European Central Bank. 2021. In the Spotlight: 2022 Supervisory Climate Stress Test. Frankfurt: European Central Bank.
108. Bank of England Results of the 2021 Climate Biennial Exploratory Scenario (CBES) 2022 London Bank of England
109. Baranova Y Jung C Noss J The Tip of the Iceberg: The Implications of Climate Change on Financial Markets 2017 London Bank of England
110. Bank of England Life Insurance Stress Test 2019: Scenario Specification, Guidelines and Instructions 2019 London Bank of England
111. Bank of England General Insurance Stress Test 2019: Scenario Specification, Guidelines and Instructions 2019 London Bank of England
112. Bank of England Discussion Paper on the 2021 Biennial Exploratory Scenario on the Financial Risks from Climate Change 2019 London Bank of England
113. Bank of England Guidance for participants of the 2021 Biennial Exploratory Scenario: Financial Risks from Climate Change 2021 London Bank of England
114. European Insurance and Occupational Pensions Authority. 2021. Sensitivity Analysis of Climate-Change Related Transition Risks. Frankfurt: European Insurance and Occupational Pensions Authority.
115. European Securities and Markets Authority ESMA Report on Trends, Risks and Vulnerabilities 2021 Paris European Securities and Markets Authority
116. Alogoskoufis S Dunz N Emambakhsh T Hennig T Kaijser M Kouratzoglou C Muñoz M Parisi L Salleo C ECB Economy-Wide Climate Stress Test Methodology and Results 2021 Frankfurt European Central Bank
117. European Central Bank ECB Publishes Final Guide on Climate-Related and Environmental Risks for Banks 2020 Frankfurt European Central Bank
118. European Central Bank 2022 Climate Risk Stress Test 2022 Frankfurt European Central Bank
119. de France Banque Scenarios and Main Assumptions of the ACPR Pilot Climate Exercise 2020 Paris Banque de France
120. Allen T Dees S Boissinot J Graciano CMC Chouard V Clerc L Climate-Related Scenarios for Financial Stability Assessment: An Application to France 2020 Paris Banque de France
121. Ens E Johnston C Scenario Analysis and the Economic and Financial Risks from Climate Change 2020 Ottawa Bank of Canada
122. Bank of Canada. 2022. Using Scenario Analysis to Assess Climate Transition Risk, Final Report of the BoC-OSFI Climate Scenario Analysis Pilot. Ottawa: Bank of Canada.
123. Australian Prudential Regulation Authority Stress Testing Assessment: Findings and Feedback 2020 Sydney Australian Prudential Regulation Authority
124. Australian Prudential Regulation Authority Climate Vulnerability Assessment 2021 Sydney Australian Prudential Regulation Authority
125. Australian Prudential Regulation Authority APRA’s Supervision Priorities 2021 Sydney Australian Prudential Regulation Authority
126. Regelink, M., H. J. Reinders, M. Vleeschhouwer, and I. Van de Wiel, 2017. Waterproof? An Exploration of Climate-Related Risks for the Dutch Financial Sector. Amsterdam: De Nederlandsche Bank.
127. Vermeulen R Schets E Lohuis M Kolbl B Jansen D-J Heeringa W An Energy Transition Risk Stress Test for the Financial System of the Netherlands 2018 Amsterdam De Nederlandsche Bank
128. Vermeulen R Schets E Lohuis M Kolbl B Jansen D-J Heeringa W The Heat Is On: A Framework for Measuring Financial Stress Under Disruptive Energy Transition Scenarios 2019 Amsterdam De Nederlandsche Bank
129. Nationalbank Danmarks Climate Change Can Have a Spillover Effect on Financial Stability 2019 Copenhagen Danmarks Nationalbank
130. Nationalbank Danmarks A Gradual Green Transition Supports Financial Stability 2020 Copenhagen Danmarks Nationalbank
131. Bundesbank Deutsche Financial Stability Review 2019 Frankfurt Deutsche Bundesbank
132. Federal Financial Supervisory Authority Guidance Notice on Dealing with Sustainability Risks 2020 Bonn Federal Financial Supervisory Authority
133. Monetary Authority of Singapore Reply to Parliamentary Question on Including Climate Change-Related Risk in MAS' Annual Industry-Wide Stress Test 2020 Singapore Monetary Authority of Singapore
134. Hong Kong Monetary Authority Pilot Banking Sector Climate Risk Stress Test 2021 Hong Kong Hong Kong Monetary Authority
135. Monetary Policy Analysis Group of the People’s Bank of China China Monetary Policy Report, Q4 202.1 2022 Beijing People’s Bank of China
136. Two Degrees Investing Initiative. 2017. Out of the Fog: Quantifying the Alignment of Swiss Pension Funds and Insurance Funds with the Paris Agreement. Paris: Two Degrees Investing Initiative.
137. Barr, M. 2022. Making the Financial System Safer and Fairer. Speech at the Brookings Institution. Washington, DC, 7 September.
138. California Department of Insurance. 2018. Scenario Analysis: Assessing Climate Change Transition Risk in Insurer Portfolios. https://interactive.web.insurance.ca.gov/apex_extprd/f?p=250:70
139. Two Degrees Investing Initiative An Analysis of New York Domestic Insurers’ Exposure to Transition Risks and Opportunities from Climate Change 2021 New York New York Department of Financial Services
140. de France Banque A First Assessment of Financial Risks Stemming from Climate Change: The Main Results of the 2020 Climate Pilot Exercise 2021 Paris Banque de France
141. de France Banque Climate Change: Which Risks for Banks and Insurers? 2019 Paris Banque de France
142. de France Banque French Banking Groups Facing Climate Change-Related Risks 2019 Paris Banque de France
143. de France Banque French Insurers Facing Climate Change Risk 2019 Paris Banque de France
144. Bank of England The Bank of England’s Response to Climate Change 2017 London Bank of England
145. Bank of England The Impact of Climate Change on the UK Insurance Sector: A Climate Change Adaptation Report by the Prudential Regulation Authority 2015 London Bank of England
146. Garbarino N Guin B High Water, No Marks? Biased Lending After Extreme Weather 2020 London Bank of England
147. Guin B Korhonen P Does Energy Efficiency Predict Mortgage Performance? 2020 London Bank of England
148. Bank of England Transition in Thinking: The Impact of Climate Change on the UK Banking Sector 2018 London Bank of England
149. Bank of England Enhancing Banks’ and Insurers’ Approaches to Managing the Financial Risks from Climate Change (SS3/19) 2019 London Bank of England
150. European Central Bank ECB Report on Institutions’ Climate-Related and Environmental Risk Disclosures 2020 Frankfurt European Central Bank
151. Two Degrees Investing Initiative. 2021. An Analysis of New York Domestic Insurers’ Exposure to Transition Risks and Opportunities from Climate Change. New York: New York Department of Financial Services.
152. Turtveit L-T Goldsack M Technological Advances and Climate Measures Can Influence Banks’ Credit Risk 2019 Oslo Norges Bank
153. de México Banco Climate and Environmental Risks and Opportunities in Mexico’s Financial System: From Diagnosis to Action 2020 Mexico City Banco de México
154. Faiella I Lavecchia L The Carbon Footprint of Italian Loans 2020 Rome Banca d'Italia
155. Faiella I Malvolti D The Climate Risk for Finance in Italy 2020 Rome Banca d'Italia
156. Faiella I Natoli F Climate Change and Bank Lending: The Case of Flood Risk in Italy 2019 Rome Bank of Italy
157. Delgado M Energy Transition and Financial Stability. Implications for the Spanish Deposit-Taking Institutions 2019 Madrid Banco de España
158. Initiative Investing New Partnership with the Japan Financial Services Agency 2019 Paris Investing Initiative
159. Skinner CP Central Banks and Climate Change Vanderbilt Law Review. 2021 47 1301
160. Bank of England Key Elements of the 2021 Biennial Exploratory Scenario: Financial Risks from Climate Change 2021 London Bank of England
161. Bank of England Update on the Bank’s Approach to the Climate Biennial Exploratory Scenario in Selected Areas 2020 London Bank of England
162. Financial Conduct Authority Enhancing climate-related disclosures by asset managers, life insurers and FCA-regulated pension providers (Policy Statement PS21/24) 2021 London Financial Conduct Authority
163. Financial Conduct Authority Enhancing climate-related disclosures by standard listed companies (Policy Statement PS21/23) 2021 London Financial Conduct Authority
164. Treasury HM A Roadmap Towards Mandatory Climate-Related Disclosures 2020 London HM Treasury
165. Department for Business Energy and Industrial Strategy Green Finance Strategy: Transforming Finance for a Greener Future 2019 London Department for Business, Energy and Industrial Strategy
166. The Emergency Economic Stabilization Act of 2008 (P.L. 110–343).
167. The Dodd–Frank Wall Street Reform and Consumer Protection Act (Dodd–Frank Act) of 2010 (P.L. 111–203).
168. The Economic Growth, Regulatory Relief, and Consumer Protection Act (EGRRCPA) of 2018 (P.L. 115–174).
169. DeMenno MB Banking on Burden Reduction: How the Global Financial Crisis Shaped the Political Economy of Banking Regulation Journal of Banking Regulation 2019 21 315 42 10.1057/s41261-019-00120-1
170. Office of the Comptroller of the Currency Principles for Climate-Related Financial Risk Management for Large Banks 2021 Washington, D.C. US Department of the Treasury
171. Department of the Treasury US Department of the Treasury Launches New Effort on Climate-Related Financial Risks in the Insurance Sector 2021 Washington, D.C. US Department of the Treasury
172. Securities and Exchange Commission. 2021. SEC Responses to Climate and ESG Risks and Opportunities’. https://www.sec.gov/sec-response-climate-and-esg-risks-and-opportunities.
173. Biden, J. 2021. Executive Order on Climate-Related Financial Risk (EO 14030). Washington, DC: Executive Office of the President.
174. Brainard, L. 2019. Why Climate Change Matters for Monetary Policy and Financial Stability. Speech to the Economics of Climate Change Conference sponsored by the Federal Reserve Bank of San Francisco. San Francisco, 8 November.
175. Powell, J. 2020. Transcript of Chair Powell’s Press Conference January 29, 2020. Washington, DC: Board of Governors of the Federal Reserve System.
176. Reserve Federal Financial Stability Report 2020 Washington, D.C. Board of Governors of the Federal Reserve System
177. Reserve Federal Federal Reserve Supervision and Regulation Report 2020 Washington, D.C. Board of Governors of the Federal Reserve System
178. Brunetti C Dennis B Gates D Hancock D Ignell D Kiser EK Climate Change and Financial Stability 2021 Washington, D.C. Board of Governors of the Federal Reserve System
179. Federal Reserve. 2020. Federal Reserve Board Announces it has Formally Joined the Network of Central Banks and Supervisors for Greening the Financial System, or NGFS, as a Member. https://www.federalreserve.gov/newsevents/pressreleases/bcreg20201215a.htm
180. Brainard, L. 2021. Financial Stability Implications of Climate Change. Speech to Transform Tomorrow Today Ceres 2021 Conference. Boston, 23 March.
181. Brainard, L. 2021. Building Climate Scenario Analysis on the Foundations of Economic Research. Speech to the 2021 Federal Reserve Stress Testing Research Conference Hosted by the Federal Reserve Bank of Boston. Boston, 7 October.
182. Office of Financial Research Size Alone is Not Sufficient to Identify Systemically Important Banks 2017 Washington, DC US Department of the Treasury
183. Cleary P Harding W McDaniels J Svoronos J-P Yong J Turning Up the Heat – Climate Risk Assessment in the Insurance Sector 2019 Bank for International Settlements Basel
184. Berner, R.B., S.G. Cecchetti, and K. L. Schoenholtz. 2019. Stress Testing Networks: The Case of Central Counterparties. Cambridge: National Bureau of Economic Research.
185. Addressing Climate Financial Risk Act of 2020 (S.5041).
186. Federal Reserve. 2022. Stress Tests and Capital Planning. https://www.federalreserve.gov/supervisionreg/stress-tests-capital-planning.htm
187. Federal Reserve. 2022. 2022 Stress Test Scenarios.. Washington, DC: Board of Governors of the Federal Reserve System.
188. Federal Reserve. 2020. Assessment of Bank Capital during the Recent Coronavirus Event. Washington, DC: Board of Governors of the Federal Reserve System.
189. DeMenno, M.B. 2020. Climate-Related Financial Risk: US Policy Landscape and Implications for Energy and Infrastructure Resilience. Albuquerque: Sandia National Laboratories.
190. Securities and Exchange Commission Proposed Rule on the Enhancement and Standardization of Climate-Related Disclosures for Investors (87 FR 21334) 2022 Washington, D.C. The Federal Register
191. Task Force on Climate-related Financial Disclosures. 2020. 2020 Status Report. Basel: Financial Stability Board.
192. Task Force on Climate-related Financial Disclosures. 2020. Task Force on Climate-related Financial Disclosures, ‘TCFD Supporters. https://www.fsb-tcfd.org/tcfd-supporters.
193. Federal Reserve. 2021. Amendments to Capital Planning and Stress Testing Requirements for Large Bank Holding Companies, Intermediate Holding Companies and Savings and Loan Holding Companies (86 FR 7927). Washington, DC: Federal Register.
194. Office of Financial Research Climate Data and Analytics Hub 2022 Washington, DC US Department of the Treasury
195. Carattini, S., G. Heutel, and G. Melkadze. 2021. Climate Policy, Financial Frictions, and Transition Risk. Cambridge: National Bureau of Economic Research.
196. Ebi, K.L., P. Berry, K. Hayes, and C. Boyer. 2018. Stress Testing the Capacity of Health Systems to Manage Climate Change-Related Shocks and Stresses. International Journal of Environmental Research and Public Health 15(11): 2370.
197. DeMenno, M.B., and C. Kauffmann. 2018. Fostering Better Rules Through International Regulatory Co-Operation. In OECD Regulatory Policy Outlook 2018. Paris: OECD Publishing, pp. 123–146.
| 0 | PMC9703411 | NO-CC CODE | 2022-11-29 23:21:08 | no | J Bank Regul. 2022 Nov 28;:1-29 | utf-8 | null | null | null | oa_other |
==== Front
Wien Med Wochenschr
Wien Med Wochenschr
Wiener Medizinische Wochenschrift (1946)
0043-5341
1563-258X
Springer Vienna Vienna
36441360
988
10.1007/s10354-022-00988-1
Case Report
Onychomadesis in a COVID-19 patient
Ivanova Zlatina Georgieva [email protected]
Aleksiev Teodor Ivanov [email protected]
http://orcid.org/0000-0002-5513-3789
Dobrev Hristo Petrov MD, PhD, DMSc [email protected]
grid.35371.33 0000 0001 0726 0380 Department of Dermatology and Venereology, Medical Faculty, Medical University, 15A V. Aprilov Blvd., 4002 Plovdiv, Bulgaria
28 11 2022
14
2 9 2022
21 10 2022
© The Author(s), under exclusive licence to Springer-Verlag GmbH Austria, ein Teil von Springer Nature 2022
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
We report the case of a 67-year-old woman who developed onychomadesis on 9 of her fingers 2 months after recovering from COVID-19, with subsequent full nail regrowth after 4 months. The development of onychomadesis in COVID-19 is probably related to inhibition of nail proliferation due to fever, direct viral damage, or an inflammatory process associated with endothelial damage and obliterative microangiopathy in the nail matrix area. Clinicians should be aware of nail changes and actively seek them out in patients with COVID-19.
Keywords
Nail changes
Onychomadesis
COVID-19
==== Body
pmcIntroduction
Coronavirus disease 2019 (COVID-19) is a pandemic systemic disease caused by a novel human pathogenic virus of the Coronaviridae family called severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Various nail changes have been observed during and after recovery from COVID-19 and have been sporadically reported in the literature:Red half-moon sign. This represents the appearance of a convex erythematous line over the distal end of the lunula. It was observed in all fingers from 2 days to 3 weeks after the onset of symptoms of COVID-19. The changes were asymptomatic and gradually disappeared over 2–3 months. It is assumed that they are related to the damage to the capillary network of the nail bed from the infectious inflammatory process [1–3]. This is supported by the changes observed by means of videocapillaroscopy of the periungual area. In 64% of the 82 patients with COVID 19 studied, Natalello G. et al. [4] found non-specific capillary changes manifested with pericapillary edema, reduced number of capillaries, enlarged and folded capillaries, slowed blood flow, microhemorrhages, and microthromboses.
Beau’s lines. These represent the appearance of transverse depressions of the nail plate, located at a distance from the proximal nail fold. They were observed in all fingernails and toenails, 2 and 3.5 months after diagnosis of COVID-19, respectively [5, 6].
Mees’ lines (transverse leukonychia). These represent the appearance of transverse, non-disappearing on pressure white lines on all fingernails that migrate forward with their growth. They were observed during illness from COVID-19. They are thought to be caused by abnormal keratinization of the nail plate due to COVID-19 [7].
Transverse orange chromonychia. This represents the appearance of an orange discoloration of the distal parts of the fingernails. It was observed 4 months after recovering from COVID-19. It is assumed that the change in the color of the nails is related to microvascular damage or to closely located and merging hemorrhages (so-called splinter hemorrhages) of the nail bed [8].
Retronychia. This is a condition in which the nail plate grows in the direction of and embeds into the proximal nail fold, leading to the development of chronic perionyxis and paronychia. It was observed on the fingers, and the condition started 12 weeks after the onset of infection [9].
Distal onycholysis and red-white discoloration of the nail bed. This was observed 4 months after recovering from COVID-19 [10].
Nail changes induced by medication used to treat COVID-19. After treatment with favipiravir and under UV irradiation (Woods lamp), a yellow-white fluorescence on the fingernails and lunula was observed [11–13]. The cause of the fluorescence has not been determined with certainty. It is assumed that the drug itself, its metabolites, or additional ingredients of the medicinal tablets are involved.
The only systematic study on nail changes in COVID-19 was performed by Grover C. et al. in 2022 [14]. They studied nail changes in 43 patients with moderately severe disease including 25 men and 18 women with a mean age of 53 years. The study was performed on average 18 days (8–38 days) after disease onset. New nail changes were found in 32 (74%) of the patients. The most common changes observed on the fingernails were as follows: nail bed erythema in 15 (35%) patients; red half-moon sign-type nails in 14 (32%), most often on the thumbs; leukonychia in 11 (25%); distal brownish staining in 6 (14%); and splinter hemorrhages in 4 (9%). On the toes, distal brown staining in 22 (51%) and leukonychia in 20 (46%) were most often observed. According to the authors, although the pathogenesis of most nail changes remains unclear, a role of inflammation and predisposition to intravascular coagulation can be assumed on one hand, and on the other hand, the impact of disturbances in the function of internal organs (liver, kidneys) in the disease can be suggested.
In addition to nail changes, other changes of the nail apparatus such as periungual desquamation, pernio-like periungual erythematous edema (COVID-toe or finger), and acral gangrene have been described in patients with COVID-19 [15].
We had the opportunity to observe and follow up on a case of onychomadesis of the fingernails after recovery from COVID-19.
Case report
A 67-year-old woman was hospitalized in a specialized department with severe general fatigue, muscle pain, shortness of breath, fever, and a positive rapid antigen test for COVID-19, where the diagnosis of COVID-19 was confirmed with a positive PCR test for SARS-Co‑2 and observation of ground glass-type changes on CT examination of the lung. Laboratory tests including complete blood count, ferritin, Lactate dehydrogenase (LDH), Aspartat-Aminotransferase (ASAT, AST), Alanine Aminotransferase (ALAT, ALT), bilirubin, urea, creatinine, C‑reactive protein, and D‑dimer were within normal limits. After treatment with antibiotics, corticosteroids, antithrombotic, and symptomatic agents, the patient was discharged with an improved general condition. Two months later, she noticed the onset of changes in the nails of the fingers—separation of the nail plates from the base of the nail shaft—about which she turned to a dermatologist. On examination, changes were observed in the nails of all fingers except the fifth finger of the right hand (Fig. 1а and 2a). The nail plate consisted of two parts, with a different ratio between them on individual fingers. The proximal portion consisted of a normal newly growing nail, and the distal portion consisted of the old nail that had been proximally detached from the nail bed. There was an arcuate groove between the two parts. The changes were consistent with onychomadesis. The patient was followed for 6 months. Growth of healthy fingernails was observed (Fig. 1b and 2b).Fig. 1 a Fingernails with onychomadesis; b Nails of both finger thumbs with onychomadesis
Fig. 2 a Healthy fingernails after 4 months; b Healthy nails of both finger thumbs after 4 months
Discussion
Onychomadesis is a complete separation of the nail plate from the nail matrix while preserving the attachment to the nail bed. It can affect single nails as a result of local factors such as trauma or inflammation of the periungual area (paronychia). When more than one or all nails are affected, a systemic condition should be sought, most associated with [16]:Infections. Onychomadesis is usually associated with infection by coxsackie viruses, the so-called hand-foot-mouth disease. Onychomadesis is most likely to develop in children younger than 7 years of age and after a latent period of 4 to 10 weeks after infection. It is associated with temporary suppression of the proliferation of the nail matrix because of its direct damage by viruses or as a result of a more severe effect on the general condition in young children. Affected nails fall off and new ones grow spontaneously.
Autoimmune diseases. Onychomadesis is reported in about 7% of patients with pemphigus vulgaris. It is associated with the development of acantholysis and bullous lesions as part of the disease process in the area of the nail matrix, nail bed, and periungual area.
Medication intake. Drug-induced onychomadesis affects several or all 20 nails. It occurs because of acute toxicity on the nail matrix, a few weeks after the start of taking the medication. The most reported medications are chemotherapeutic agents such as capecitabine, doxorubicin, cytosine arabinoside, and etoposide, as well as antiepileptics, especially carbamazepine and valproic acid.
In terms of prognosis, in all cases, the condition is described as a temporary event, with nail regrowth within 12 weeks.
There are isolated reports in the literature of development of onychomadesis in COVID-19. In 2020, Senturk N. and Ozdemir H. [17] were the first to report a 47-year-old woman who developed onychomadesis 3 months after recovering from COVID-19. On dermatologic examination, the nails of her fingers and toes were separated from the nail base and healthy nails were seen growing from the proximal matrix.
The next communication of onychomadesis was by Liu J. et al. in 2021 [18]. They observed a 59-year-old man of African origin who had COVID-19 as an outpatient. About 4 months later, there was a sudden detachment of the nail plates of four of the fingers from the proximal nail fold. The toes were not affected. At the follow-up visit after 3 months, the patient reported a gradual loss of nails on all fingers. There was partial growth of new thin nail plates located on about half of the nail bed area.
Colonna C. et al. in 2022 [19] reported on an 11-year-old girl in whom illness from COVID-19 was accompanied by the development of pernio-like (chilblain-like) lesions on the fingers and hands. About 1 month later and 3 days after the resolution of these lesions, detachment of the fingernails from their base occurred. Dermatological examination revealed onychomadesis on four fingernails.
All authors consider that the observed onychomadesis is most likely related to inhibition of nail proliferation due to SARSCoV‑2 infection due to fever, direct viral damage, or an inflammatory process associated with endothelial damage and obliterative microangiopathy in the nail matrix area.
Conclusion
Nail changes in COVID-19 reflect the systemic nature of the disease. They are associated with changes in the vascular apparatus of the nail bed and suppression of nail matrix growth. Like skin manifestations, nail manifestations can also be helpful in the diagnosis of COVID-19, and clinicians should be aware of these and actively seek them out in patients.
Conflict of interest
Z.G. Ivanova, T.I. Aleksiev, and H.P. Dobrev declare that they have no competing interests.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
1. Mendez Mendez-Flores S Zaladonis A Valdes-Rodriguez R COVID-19 and nail manifestation: be on the lookout for the red half-moon nail sign Int J Dermatol 2020 59 11 1414 10.1111/ijd.15167 32860426
2. Neri I Guglielmo A Virdi A The red half-moon nail sign: a novel manifestation of coronavirus infection J Eur Acad Dermatol Venereol 2020 34 11 e663 e665 10.1111/jdv.16747 32535979
3. Unal E Çakmak S Yorulmaz A The red half-moon nail sign in a COVID-19 patient Our Dermatol Online 2022 13 2 223 224 10.7241/ourd.20222.29
4. Natalello G De Luca G Gigante L Nailfold capillaroscopy findings in patients with coronavirus disease 2019: Broadening the spectrum of COVID-19 microvascular involvement Microvasc Res 2021 133 104071 10.1016/j.mvr.2020.104071 32949574
5. Alobaida S Lam J Beau lines associated with COVID-19 CMAJ 2020 192 36 E1040 10.1503/cmaj.201619 32900765
6. Ide S Morioka S Inada M Ohmagari N Beau’s lines and leukonychia in a COVID-19 Patient Intern Med 2020 59 24 3259 10.2169/internalmedicine.6112-20 33132338
7. Fernandez-Nieto D Jimenez-Cauhe J Ortega-Quijano D Transverse leukonychia (Mees’ lines) nail alterations in a COVID-19 patient Dermatol Ther 2020 33 6 e13863 10.1111/dth.13863 32779847
8. Tammaro A Adebanjo G Erasmus H-P Chello C Transverse orange nail lesions following SARS-CoV-2 infection Dermatol Ther 2021 34 1 e14688 10.1111/dth.14688 33340203
9. Ceccarelli M Nakamura R Canella C Multiple retronychia following COVID-19 infection J Emerg Med 2021 24 101087
10. Demir B Yuksel E Cicek D Turkoglu S Heterogeneous red-white discoloration of the nail bed and distal onycholysis in a patient with COVID-19 J Eur Acad Dermatol Venereol 2021 35 9 e551 e553 10.1111/jdv.17347 33987893
11. Gulseren D Yalıcı-Armagan B Yellow-white fluorescence on the nails: A novel finding of Favipiravir used for the treatment of COVID-19 J Cosmet Dermatol 2021 20 8 2392 2393 10.1111/jocd.14214 33978291
12. Kayıran M Cebeci F Erdemir V Fluorescence of nails and hair on Wood’s lamp examination in Covid pandemic; undefined effect of Favipiravir in humans Dermatol Ther 2021 34 1 e14740 33404148
13. Kutlu O Yılmaz S Fingernail lunula luminescence in COVID-19 patients: Is it a favipiravir-related reaction or a novel manifestation of coronavirus infection Photodermatol Photoimmunol Photomed 2021 37 4 343 344 10.1111/phpp.12660 33480085
14. Grover C Saha S Pandhi D Nail changes in COVID-19: a cross sectional study from India Indian Dermatol Online J 2022 13 326 333 10.4103/idoj.idoj_586_21 36226020
15. Wollina U Kanitakis J Baran R Nails and COVID-19—A comprehensive review of clinical findings and treatment Dermatol Ther 2021 34 5 e15100 10.1111/dth.15100 34398500
16. Hardin J Haber R Onychomadesis: literature review Br J Dermatol 2015 172 3 592 596 10.1111/bjd.13339 25132198
17. Senturk N Ozdemir H Onychomadesis following COVID-19 infection: Is there a relationship? Dermatol Ther 2020 33 6 e14309 10.1111/dth.14309 32935910
18. Liu J Veenstra J COVID-19 associated onychomadesis SKIN 2021 5 3 286 288 10.25251/skin.5.3.11
19. Colonna C Giacalone S Zussino M A case of onychomadesis associated with chilblain-like lesions after COVID-19 infection J Eur Acad Dermatol Venereol 2022 10.1111/jdv.18208 35535454
| 36441360 | PMC9703413 | NO-CC CODE | 2022-11-29 23:21:42 | no | Wien Med Wochenschr. 2022 Nov 28;:1-4 | utf-8 | Wien Med Wochenschr | 2,022 | 10.1007/s10354-022-00988-1 | oa_other |
==== Front
Curr Cardiovasc Risk Rep
Curr Cardiovasc Risk Rep
Current Cardiovascular Risk Reports
1932-9520
1932-9563
Springer US New York
711
10.1007/s12170-022-00711-0
Technology and Cardiovascular Health (E. Muse, Section Editor)
Health Techequity: Opportunities for Digital Health Innovations to Improve Equity and Diversity in Cardiovascular Care
http://orcid.org/0000-0002-6545-3110
Hernandez Mario Funes 1
Rodriguez Fatima [email protected]
2
1 grid.168010.e 0000000419368956 Department of Medicine, Division of Nephrology, Stanford University School of Medicine, Stanford, CA USA
2 grid.168010.e 0000000419368956 Department of Medicine, Division of Cardiovascular Medicine, Stanford University School of Medicine, 453 Quarry Road, Room 332B, Stanford, CA 94305 USA
28 11 2022
120
1 11 2022
© The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2022, Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Purpose of Review
In this review, we define health equity, disparities, and social determinants of health; the different components of digital health; the barriers to digital health equity; and cardiovascular digital health trials and possible solutions to improve health equity through digital health.
Recent Findings
Digital health interventions show incredible potential to improve cardiovascular diseases by obtaining longitudinal, continuous, and actionable patient data; increasing access to care; and by decreasing delivery barriers and cost. However, certain populations have experienced decreased access to digital health innovations and decreased representation in cardiovascular digital health trials.
Summary
Special efforts will need to be made to expand access to the different elements of digital health, ensuring that the digital divide does not exacerbate health disparities. As the expansion of digital health technologies continues, it is vital to increase representation of minoritized groups in all stages of the process: product development (needs findings and screening, concept generation, product creation, and testing), clinical research (pilot studies, feasibility studies, and randomized control trials), and finally health services deployment.
==== Body
pmcIntroduction
Intentional and targeted use of digital health innovations can advance and promote health equity. The COVID-19 pandemic drove an increased uptake of telemedicine services of up to 1.5x that of the pre-pandemic period [1]. However, certain populations experienced barriers to telemedicine access [2, 3]. These disparities are rooted in structural inequities that can in term impact how individuals can benefit from the promise of digital health. However, during the COVID-19, we also witnessed an increased collaboration between technology and health industry to expand access to transportation services, educational programs in digital health, and cloud programs to aid with vaccine distribution to historically marginalized groups [4]; federal funded programs increased access to broadband internet, digital health education, and devices [5]. Special efforts will need to be made to expand access to the different elements of digital health, ensuring that the digital divide (the economic, educational, and social inequalities between those that have or do not have access to information and communication technology) does not exacerbate health disparities [6]. In this review, we define health equity, disparities, and social determinants of health; the different components of digital health; the barriers to digital health equity; and cardiovascular digital health trials and possible solutions to improve health equity through digital health.
Health Equity
In 1985, the Department of Health and Human Services published the Heckler Report on Black and Minority Health, exposing for one of the first times that race and ethnicity may be an independent contributor to health outcomes [7]. Specifically for cardiovascular care, it was clear that Black Americans had fewer office visits, diagnosis and interventions for coronary artery disease than White individuals. Subsequently, ample research has elucidated the inequities in cardiovascular care for racial and ethnic minorities [8]. The National Institute on Minority Health and Health Disparities defines a health disparity as a health difference that adversely affects disadvantaged populations, based on one or more health outcomes: higher incidence, prevalence or earlier onset of disease, higher prevalence of risk factors, higher rates of condition-specific symptoms, premature or excessive mortality, and greater global burden of disease [9]. In order to reduce health disparities, we need to understand the different social determinants of an individual’s health or the conditions in which individuals live work and play that impact their health. Many of these factors are outside of the health care systems and include housing, income, and education. Not everyone has the same opportunity to be healthy [10]. To properly address these disparities, we should strive for health equity, not equality. Equality entails the distribution of the same resources and opportunities to every individual across a population—regardless of achieving the same outcome. Equity, on the other hand, is delivering these resources and opportunities tailored to the specific needs of a group to achieve equal outcome in the population [11]. Black individuals have higher rates of uncontrolled cardiovascular risk factors and higher age-adjusted death from cardiovascular disease compared with the general US population [12]. Improving health equity involves directing resources to communities that are underrepresented in research. For instance, Brewer et al. partnered with Black communities in all phases of product development to create and test a digital health cardiovascular disease prevention program that led to an improvement in the intervention Heart Association (AHA) Life-Simple 7 score (components: smoking, healthy diet, physical activity, BMI, blood pressure, cholesterol, and glucose) [13, 14].
Digital Health Elements
To better understand how digital health could address health inequities, we first need to define the different components involved in the use of information and communication technology for the delivery of health care:Digital health: commonly referred as eHealth, is the use of information and communication technology to manage patients and their health [15]. Digital health includes consumer products such as smart devices or connected equipment that have not undergone rigorous clinical studies [16].
Digital medicine: the subset of digital health that pertains to high-quality hardware and software products developed through evidence-based clinical studies for medical care and treatment [17].
Telemedicine: the specific use of information and communication technologies to deliver health care, clinical and administrative services, and medical education, remotely, from one site to another.
mHealth: the use of mobile communication devices to exchange data or information between doctors and patients [18].
Remote patient monitoring: the use of digital health devices to capture and serially monitor vital signs and biometrics with upload of such patient generated data to a digital platform for review by patients and clinical teams [18].
Wearable and consumer technologies: such as commercially available activity tracking, sleep monitoring devices, smartwatches [18].
Barriers to Digital Health Inclusion
The COVID-19 pandemic propelled the use of digital health throughout the world, predominantly through telemedicine and remote patient monitoring [19]. But due to lack of preparedness, many historically marginalized groups were left with less access to care from the lack of infrastructure to support digital health care delivery in these communities [3]. Telemedicine has the ability to decrease health inequities by surpassing the limitations of access to care such as the cost of transportation, inability to leave the house due to disability, and loss of time at work. But in order for patients to benefit from telemedicine services, patients require internet or broadband access, a device (smartphone, computer, or tablet), applications, and a minimum digital health literacy.
In general terms, broadband refers to a set of networked data transmission technologies which permit internet communication and access to digital information. In the USA, 15% of households do not have internet service [20]. The states with majority Black residents or lowest median incomes had the lowest broadband adoption rates. For example, areas with majority White residents had an average adoption rate of 84% compared with 67% of majority-Black residential areas [20]. Decreased access to broadband internet use has been linked to decreased patient portal access. An observational study from a large tertiary-care center hospital showed that Black and Hispanic individuals had lowers odds of initiating a patient portal account or messaging their providers related to their decreased access to the internet [21]. Federal, state, and local programs specifically targeted at underserved communities have the ability to increase access to broadband. The Rural Digital Opportunity Fund program established in 2019 will award a total of $9.2 billion dollars over 10 years. In its 2021 report, the number of Americans without broadband access decreased from 30% in 2016 to 16% in 2019. Additionally, three-quarters of those with new access to broadband lived in rural areas [22].
Gaps remain in computer ownership, with just 69% of Black adults and 67% of Hispanic adults owning a computer as compared with 80% of White individuals [23]. On the other hand, smartphone ownership has steadily increased over the years and data from a 2021 Pew Research survey demonstrates that gaps in smartphone ownership by race are decreasing. Still, nearly one-half of older adults and 30% of those earning less than $30,000 own a smartphone and many low-income households share devices [24]. As an example, home-based cardiac rehabilitation programs can decrease health disparities by expanding care to patients with difficulties attending specialized rehabilitation centers. But if patients from low socioeconomic status, older adults or minoritized groups continue to lack access to computers, broadband, or digital health literacy, then such interventions will continue to exacerbate the digital divide. We need programs that increase access to device ownership and education. If passed, The Device Access for Every American Act introduced at the end of 2021 has the potential to reduce disparities by increasing access to a connected device (desktop, laptop, tablet) by providing a $400 voucher to low-income individuals [25]. In addition, we should extend federal programs enacted during the COVID-19 pandemic to help schools and libraries increase access to devices and education [26].
With regard to wearable ownership, survey data from the Pew Research Center (65% White participants and annual income above 30,000) revealed that in 2020, 21% of US adults report regular use of a smartwatch [24]. Ownership was similar between participants of different ethnic backgrounds, but Black and Hispanic individuals were less likely to approve sharing of their data for heart disease research. On the other hand, surveys performed in Federally Qualified Health Centers with predominantly underrepresented population (70% nonwhite, 70% learning less than 30,000) report lower wearable ownership of 21% [27]. Cost was the main barriers for those who did not own a fitness tracker but would like to own one. With multiple imbedded sensors (accelerometers, photoplethysmography, electrocardiography, etc.), wearable devices can collect a plethora of data for the prevention, diagnosis, and treatment of cardiovascular diseases. Researchers and device companies should continue to collaborate to improve device accuracy and define meaningful use criteria. To increase access, device companies should lower costs; insurance companies must expand reimbursement for biometric data collection and develop programs that increase access to wearable devices [28].
Universal design principles have been present at least since the 1990s to aid designers in developing products for the widest possible range of individuals [29]. The equitable use principle states that products should “Provide the same means of use for all users: identical whenever possible; equivalent when not” and to “avoid segregating or stigmatizing any users” [30]. Unfortunately, the COVID-19 pandemic provided important insight in the gaps in the design of products for people from different cultural, social, educational backgrounds. Most of the apps were available only in English (65%) and 69% had a readability above 9th grade [31]. Several design standards and guides are now available aimed specifically at digital health tools [32]. Two commonly used design processes include the Universal Design Principles and User-Centered Design. While these approaches possess important differences, both aim at maximizing the usability of products to a diverse group of users [33]. Community-based participatory research is another approach that aims to create partnerships between intended end users from the community, academic, and research institutions across all the design stages [34].
Digital Health Interventions for Cardiovascular Care
Hypertension
Hypertension affects close to a third of the adult population worldwide [35]. Black individuals have a higher prevalence of hypertension and lower levels of hypertension awareness and control [36]. Additionally, low-middle-income countries and racial and ethnic minorities in high-income countries tend to have decreased awareness, control, and worse outcomes [35]. Some of the barriers in hypertension control are related to frequent follow up visits, cost of multiple follow-up appointments, the cost transportation, and in certain places the distance required to reach a health care institution. With telemedicine the health care system has the ability to obtain more blood pressure readings that are transferable to the health care team (nurse, pharmacist, physician, etc.) to allow more frequent medication titration without the need for time or cost spent in transportation or additional follow up appointments.
Randomized trials have tested different ways to improve hypertension treatment and control before and after the internet era, with a few trials specifically targeting low-resource setting and diverse patient populations (Table 1). In the Effects of Nurse-Managed Telemonitoring on Blood Pressure at 12-Month Follow-Up Among Urban African Americans study, a nurse-managed telemonitoring intervention (telephonically transmitted BP measurements, nurse intervention, and counseling) showed greater reduction in systolic blood pressure at 12 months compared with usual care alone (13 vs 7.5 mm Hg) [37]. Blood pressure reduction was also achieved with a similar intervention in a subgroup analysis of uncontrolled patients in the Durham VA internal medicine clinics (49% black) [38]. The “Nurse-led Disease Management for Hypertension Control in a Diverse Urban Community” Randomized Controlled Trial (RCT) looked at the response to different levels of intervention (HBPM alone, HBPM + nurse telephone counseling and usual care). At 9 months, systolic blood pressure was − 7.0 mm Hg lower (confidence interval [CI] − 13.4 to − 0.6) in the nurse management plus home blood pressure monitor arm relative to usual care. There was no statistical difference in the home blood pressure only arm versus usual care [39]. Bove et al. studied a telemedicine plus physician intervention versus usual care in a predominantly African American population with 53% at or below the poverty line. Blood pressure control was similar in both groups at 6-month follow-up (52.3% intervention versus 54.5% usual care, P = 0.43) [40].Table 1 Digital health randomized control trials in hypertension
First author Country Age mean (SD) N Race ethnicity* Duration (months) Inclusion criteria Primary outcome& Results Device Intervention Medication titration
Artinian et al. (2007) [37] US 59.1 (13) vs 60.2 (12.3) 394 100% Black 12 months AA with HTN at community centers Reduction in BP at 12 months − 13 vs − 7.5 mm Hg, P = 0.04 HBPM + LifeLink Monitor •TM** + nurse tele counseling
•Enhanced UC$
No
Green et al. (2008) [41] US 59.1 (8.5) 778 83% Black 12 months 25–75 years of age, uncontrolled BP, no diabetes, no cardiovascular or renal disease Change in SBP, DBP and control at 12 months TM + pharmacist versus UC: 56% vs 31%, P = < 0.001 HBPM
(Omron Hem-705-CP)
3 arms:
•TM + Web
•TM + pharmacist
•Usual care
Yes (Pharmacist)
Parati et al
(2009) [42]
Italy 57.2 (10.7) in IT
58.1
(10.8) in UC
288 Not reported 6 months 18–75 years with office SBP ≥ 140 mm Hg or DBP ≥ 90 mm Hg and ABPM SBP ≥ 130 or DBP ≥ 80 mm Hg Proportion of patients reaching control 62% vs 50% (P < 0.05) IT vs UC HBPM (Tensiomed) + Tensiophone TM + BP data telephone transmitter + Nurse intervention if BP above safety thresholds No (physician contacted if BP above safety threshold, i.e., ≥ 180/110 mm Hg)
McManus (2010) [43] UK 66 (8.8) 480 96% White, 1.5% Black, 2.1% Asian 12 months Age 38–85 years with BP ≥ 140/90 mm Hg SBP difference at 6 and 12 months 6-month SBP difference of 3.7 (0.8–6.6, P = 0.013); 12-month SBP difference 5.5 (2.2 to 8.8) in the IT vs control HBPM (Omron 705IT) TM via telephone device (i-modem, Netmedical) Yes (physician)
Wakefield et al. (2011) [44] US 66 (10) 302 96% White, 3% African American 18 months T2DM and HTN treated by VA PCP SBP at 6 months High IT vs UC: − 6.05 vs + 4.48, P = 0.001
High IT vs low intensity:
− 6.05 vs − 0.29, P = 0.9
HBPM + telephone line for data transfer (Viterion-Bayer Panasonic) 3 arms:
•TM + high intensity education
•TM + low intensity education
Usual care
Yes, for high intensity (Physician based treatment algorithm)
Bosworth et al. (2011) [38] US 64 (10) 591 49% White,
48% Black
18 months Diagnosis of HTN, using BP-lowering meds, BP > 140/90 mmHg BP control at 18 months 12.8%, P = 0.03 (behavioral vs usual care)
12.5%, P = 0.03 (combined intervention vs usual care)
HBPM (UA-767PC, A&D Medical Digital BP) 4 arms:
•TM
•TM + behavioral management
•TM + behavioral management + nurse and software intervention
•Usual
Yes
(Physician aided by nurse from study team with web-based algorithm)
Piette et al. (2012) [45] Mexico and Honduras 57.6 (0.8) 200 Not reported 6 weeks Age 18–80 and SBP ≥ 140 mm Hg or ≥ 130 mm Hg with diagnosis of T2DM SBP difference at 6 weeks SBP difference of − 4.2 mm Hg (− 9.1 to 0.7, P = 0.09) IT vs UC HBPM TM via telephone calls + adherence monitoring and behavioral change No
Hebert et al. (2012) [39] US 60.8 (11.6) 416 59% Black, 37% Hispanic 18 months Self-described black or Hispanic, community dwelling at enrollment, BP ≥ 140/90 mm Hg Change in SBP and DBP at 9 Difference of − 7 mm Hg (− 13.4 to − 0.6 IT vs UC at 9 months HBPM (Omron HEM-712C) TM + nurse counseling and contacting physician to suggest treatment changes Yes (Physician aided by nurse from study team)
Margolis et al. (2013) [46] US 61.1 (12) 450 82% White, 12% Black, 2% Asian 18 months HTN with BP above 140/90 mmHg Proportion of patients controlled at 6- and 12-month visit 6 months: 71.8 vs 45.2% P ≤ 0.001
12 months:
71.2 vs 52.8% P = 0.005
HBPM (A&D Medical 767PC) TM + web services + pharmacist calls Yes (pharmacist with web-based algorithm)
Bove et al. (2013) [40] US 59.6 (13.5) 241 81% African American, 15% White, 2.5% Hispanic 6 months SBP ≥ 140 mmHg Proportion of BP control 54.5% versus 52.3%, P = 0.430 HBPM
(Microlife USA Scale (Taylor Digital) Pedometer (Digi-Walker SW-200)
TM + patients uploaded data via web or telephone + monthly report to physicians No (usual clinical decision)
Magid et al. (2013) [47] US 60 (11) 348 84% White, 8% Black, 6% Hispanic 6 months 18–79 years of age, HTN with SBP ≥ 140 or DBP ≥ 90 mmHg Proportion of BP control 54.1 versus 35.4%, P ≤ 0.05 HBPM (Omron HEM-790IT) TM + automatic upload data to Heart360 Web Account + clinical pharmacist Yes (pharmacist)
Kerry et al. (2013) [48] UK 72 (12) 381 77% White, 14.9% Black, 7.2% Asian 12 months History of stroke or TIA 9 months prior and BP > 140/85 mm Hg or on antihypertension medication Change in SBP at 12 months Difference of 0.3 mm Hg (CI − 3.6 to 4.2) at 12 months HBPM (Omron 705CP) Self-monitoring + nurse counseling No
McKinstry et al. (2013) [49] UK 61 (11) 401 Not reported 6 months ≥ 18 years + SBP > 145 mm Hg or DBP > 85 mm Hg Mean change in SBP at 6 months Difference of − 4.3 (− 2 to − 6.5, P = 0.0002) IT vs UC HBPM (Sabil-O-Graph) TM + automatic transmission of BP readings via smart phone + feedback to patients and HCT*** No
Stewart et al. (2014) [50] Australia 67 (12) 395 71% born in Australia, race/ethnicity not reported 6 months ≥ 18 years + HTN + on HTN meds Change in SBP and DBP at 6 months (2ry outcome) SBP difference of − 5.3 mm Hg (0.0 to 10.6, P = 0.05) IT vs UC HBPM (Omron HEM-790IT) TM + pharmacist medication review and adherence check No (pharmacist could referred to GP)
Leiva et al. (2014) [51] Spain 65 (10.7) 114 Not reported 12 months 18–80 years with SBP ≥ 140 mm Hg or DBP ≥ 90 mm Hg SBP at 12 months SBP 151.3 vs 153.7, P = 0.294 IT vs UC HBPM TM + education and motivational interview + pillbox organized + pharmacist intervention Yes (pharmacist)
McManus et al. (2014) [52] UK 69.3 (9.3) in IT
69.6 (9.7) in UC
552 96.6% White, 1.5% Black, 1.5% Asian, 0.5% others 12 months Age ≥ 35 years, SBP ≥ 130 or DBP ≥ 80 mm Hg plus at least one high risk condition# SBP difference at 12 months SBP difference 9.2 (5.7–12.7) IT vs UC HBPM (MicroLife Watch BP Home) TM plus self-titration (3 step plan prespecified by their physician) Yes (participant)
Ogedegbe et al. (2014) [53] US 56 (12.1) 1039 100% African American 12 months Self-identified black or African American, uncontrolled HTN Rate of BP control at 12 months 50.2 vs 45.3%, P = 0.18 HBPM TM + lifestyle counseling
Physicians received monthly feedback
No (usual clinical decision)
Yi et al. (2015) [54] US 61.3 (12) 900 11% White, 26% Black, 63% Hispanic 6 months ≥ 18 years, HTN ≥ 6 months, last clinic visit SBP ≥ 140 or DBP ≥ 90 mmHg Change in SBP and DBP at 7–10 months Mean change in SBP of 14.7 vs 14.1 mmHg P = 0.70 HBPM TM + education No
Bobrow et al. (2016) [55] South Africa 54.3 (11.5) 1256 57.6% Black, 42.4% other 12 months Age ≥ 21 years and HTN diagnosis Change in SBP at 12 months Difference SBP: interactive vs UC − 1.6 (− 3.7 to 0.6, P = 0.16); information only vs UC − 2.2 (− 4.4 to − 0.04, P = 0.046) SMS text messages 3 arms:
•Interactive two-way SMS messages
•One-way information only SMS messages
•Usual care
No
Frias et al. (2017) [56] US IT: 57.8 (1.1)
UC: 61.6 (1.7)
109 66% White, 47% Hispanic, 16% African American, 14% Asian 1–3 months Adults with HTN and BP ≥ 140/90 mmHg Change in SBP at 4 weeks − 21.8 vs − 12.7 mm Hg* DMO*** Digital medicine + wearable sensor + app No
McManus et al. (2018) [57] UK 66.9 (9.4) 1182 95% White, 1.7% Black, 1.4% Asian, 0.6% mixed 12 months Age ≥ 35 years, SBP ≥ 140 mm Hg or DBP ≥ 90 mm Hg SBP difference at 12 months Adjusted difference: self-monitoring − 3.5 mm Hg, P = 0.0029; TM − 4.7 mm Hg, P = 0.0001 vs usual care. No difference between self- and TM HBPM (Omron M10-IT) 3 arms:
•TM via text or web + physician titration and web interface
•Self-monitoring + mail readings + physician titration
•Usual care
Yes (physician)
McManus et al. (2021) [58] UK 66 (10) 622 94% White, 1.4% Black, 1.1% Asian, 3.4% Other 12 months Age ≥ 18 years, SBP ≥ 140 mm Hg or DBP ≥ 90 mm Hg SBP difference at 12 months Mean SBP difference of − 3.4 (− 6.1 to − 0.8) IT vs UC HBPM (Omron M3) TM + patient and physician integrated online intervention + behavioral and lifestyle education Yes (physician)
$Enhance UC: provided with list of PCPs (if did not have one), enrolled in pharmacy assistance program and AHA pamphlet “Silent Stalker”
*As collected or reported in the original manuscript
**TM = BP Telemonitoring
**DMO = includes digital medicines, the wearable sensor patch, and the mobile device app
***HCT = health care team
&All office BP unless otherwise stated
#Stroke or transient ischemic attack; diabetes; stage 3 chronic kidney disease (estimated glomerular filtration rate, 30–59 ml/min/m2); coronary artery bypass graft surgery; myocardial infarction or angina
The “Counseling African Americans to Control Hypertension” trial enrolled 1039 patients from 30 community health centers in the New York City area. In the intervention arm, patients received computerized education, behavioral counseling sessions, and HBMP-validated devices; clinicians received monthly on-site education, hypertension case rounds, and quarterly chart audits of their patient office blood pressure readings. The BP control rate was similar in intervention (49.3%) versus control (44.5%) groups. In prespecified subgroup analyses, the intervention was associated with greater BP control in patients without diabetes mellitus (intervention 54.0% versus usual care 44.7%; odds ratio, 1.45 [CI, 1.02–2.06]); and small-sized community health centers (intervention 51.1% versus usual care 39.6%; odds ratio, 1.45 [CI, 1.04–2.45]) [53].
Researchers and companies developing patient-facing platforms should collaborate with communities throughout the entire design process, test and validate these technologies, and integrate them into the electronic health record for seamless transfer of data. Insurers should expand coverage of home blood pressure monitors and reimburse clinicians for using digital health technology to deliver care. Developing provider-facing platforms for low-resource settings, imbedding tools into regular clinician workflow, and providing concise and actionable data aiming at improving care with no added cost or effort should be national priorities.
Cardiovascular Health
The use of digital health to improve cardiovascular risk factors such as dyslipidemia by improving adherence or promoting lifestyle changes is an important public health intervention, both for primary and secondary prevention of cardiovascular diseases. Unfortunately, there is paucity of data regarding the benefits of digital health interventions targeted specifically at lower socioeconomic status, elderly, Black, or Hispanic populations for risk factor modification in cardiovascular health (Table 2). Furthermore, studies have shown that patients with a self-reported history of atherosclerotic cardiovascular disease are less likely to use health information technology to manage their health [59].Table 2 Digital health trials in ASCVD and cardiac rehabilitation
First author Country Age mean (SD) N Race ethnicity Duration Inclusion criteria Primary outcome Results Device Intervention
Brath et al. (2013) [60] Austria 69 (4.8) 53 Not reported 40 weeks At least 2 diagnoses: HTN, DM2, HLD Intake rate at 20 weeks Significant difference in Metformin adherence. No difference in the other 3 medications Electronic blister + NFC capable smartphone Adherence text reminders to participants and adherence information to physicians
Petrella et al. (2014) [61] Canada 56.7 (9.4) 149 100% Caucasian 12 weeks At least 2 risk factors* SBP at 12 weeks. Secondary outcome: waist circumference, HBA1c, HDL, LDL SBP mean change greater in IT vs control. No difference in secondary outcomes Smartphone, app, glucometer, HBPM, weight scale, pedometer Individualized exercise program + home monitoring kit
Chow et al. (2016) [62] Australia 58 (9.2) 710 66.6% European, 10.7% South Asian, 10.1% other Asian, 9.9% Arab 6 months ≥ 18 years of age and documented CHD** LDL-C level at 6 months Significant difference in LDL-C of − 5 mg/dL (− 9 to 0, P = 0.4) Text messages Semi-personalized text messages with motivation to improve diet, exercise, and smoking cessation
Anand et al. (2016) [63] Canada 50.6 (11.4) 343 100% South Asian
(90% India, 2.3% Pakistan, 5.2% Sri Lanka)
1 year South Asian ≥ 30 years of age MI scores at 12 months Relative change between IT and control was not significant (− 0.27, − 1.12 to 0.58, P = 0.53) Email messages Change-oriented motivational, diet, and physical activity messages
Salisbury et al. (2016) [64] UK 67.4 (4.8) 641 99% White 1 year 40–74 years of age + QRISK2 10-year risk score of ≥ 20% and modifiable diseases*** Maintaining or decreasing QRISK2 score at 12 months Proportion that maintained or improved was not significantly different in IT vs control 50 vs 42%
(OR 1.3, 1.0–1.9, P = 0.08)
Telephone calls + web portal Health advisor plus computerized behavioral management program
Skobel et al. (2017) [65] UK, Germany, Spain 59 (14) 132 Not reported 6 months Hx of acute MI or CAD s/p PCI, LVEF ≥ 30% Peak VO2 max at 6 months in HBCR$ vs CBCR# national standards Peak VO2 max change 1.76 ± 4.1 ml/min/kg in HBCR vs − 0.4 ± 2.7 ml/min/kg in CBCR, P = 0.005 •Smartphone
•ECG
•Vest
•Vital sign senor
•Physician-facing platform
Asynchronous home-based cardiac rehabilitation
Hwang et al. (2017) [66] Australia 67 53 92% Caucasian 12 weeks ≥ 18 years of age and recent heart failure admission, diagnosis confirmed by echocardiogram Non-inferiority: change in 6-min walk distance HBCR vs CBCR At 12 weeks, there was no between-group difference 15 m (95% CI − 28 to 59); F = 1.39, P = 0.24 •Laptop
•Mobile broadband
•HBPM
•Pulse oximeter
•Weight and resistance bands
Synchronous videoconference home-based cardiac rehabilitation
Harzand et al. (2018) [67] US 65 (5) 18 50% African American 12 weeks ≥ 18 years with coronary heart disease plus on indication for cardiac rehabilitation BP and functional capacity (single arm feasibility study) Improvement in metabolic equivalent from 5.3 to 6.3, P = 0.008; mean BP at rest decreased from 1401 to 130.5, P = 0.039 •Smartphone platform
•Hospital-facing dashboard
Asynchronous home-based cardiac rehabilitation
Peng et al. (2018) [68] China 66.3 (10.5) 98 Not reported 4 months ≥ 18 years, heart failure for at least 3 month and NYHA class I–III Primary: QoL, secondary: 6-min walking distance, LVEF and heart rate Statistically significant changes in QoL scores, 6-min walk distance and heart rate Web-based platform Synchronous videoconference home-based cardiac rehabilitation
Maddison et al. (2019) [69] New Zealand 61 (13) 162 75.3% NZ European, 4.3% NZ Maori, 2.5% Pacific, 8% Asian 12 weeks ≥ 18 years with coronary heart disease within 6 months Non-inferiority outcomes: VO2 max at 12 weeks Adjusted mean VO2 max difference = 0.46, 95% CI − 0.92 to 1.84 ml/kg/min, P = 0.51 •Smartphone
•Chest-word wearable sensor
•Apps and Web Platform
Synchronous home-based cardiac rehabilitation
Tekkesin et al. (2021) [70] Turkey Mean:59 (53–64) 283 Not reported 1 year 20–79 years of age with 10 years ASCVD score ≥ 7.5% ASCVD scores at one year IT vs UC reduced ASCVD score by difference of − 2.7% (− 2.2 to − 3.3, P ≤ 0.0001) Smartphone, weight scale, smart wrists band and HBPM Daily upload of data with motivational messages and feedback
Bae et al. (2021) [71] Korea 60.4 (10.5) 879 Not reported 6 months CHD and underwent PCI LDL-C, SBP and BMI change at 6 months No significant difference in any outcome: LDL-C, SBP, and BMI Text messages Semi-personalized text messages with motivation to improve diet, exercise, and smoking cessation
*Waist circumference ≥ 88 cm (women) or 102 cm (men); SBP ≥ 135 mmHg and/or DBP ≥ 85 mmHg; fasting plasma glucose ≥ 6.1 mmol/l; fasting triglycerides ≥ 1.7 mmol/l; fasting HDL ≤ 1.29 mmol/l (women) or 1.02 mmol/l (men)
**Defined as documented prior myocardial infarction, coronary artery bypass graft surgery, percutaneous coronary intervention, or 50% or greater stenosis in at least 1 major epicardial vessel on coronary angiography
***Systolic blood pressure ≥ 140 mm Hg, body mass index ≥ 30, being a current smoker, or any combination of these
$Home-based cardiac rehabilitation
#Center-based cardiac rehabilitation
Bae et al. randomized 879 patients with a history of coronary heart disease who underwent percutaneous coronary intervention to a semi-personalized support website and a short message service (SMS) with lifestyle modifications versus usual care (regular clinic follow-up without text messages). At 6 months, there was no significant difference in the cardiometabolic risk profiles between the groups [71]. A higher intensity intervention in a Turkish population that remotely monitored patients’ diet, weight, steps, and blood pressure with additional motivational messages to improve healthy lifestyle showed a significant reduction in ASCVD score of − 2.7% (adjusted treatment effect − 2.7, 95% CI − 2.2 to − 3.3, P < 0.0001) [70]. Brewer et al. randomized churches with predominantly African American adults to test an app-based cardiovascular health promotion intervention. The FAITH! App provided educational models, diet, and physical activity self-monitoring and social networking. Educational material focused on all American Heart Association (AHA) Life-Simple 7 components: smoking, healthy diet, physical activity, BMI, blood pressure, cholesterol, and glucose. The primary outcome was the average change in mean AHA Life-Simple 7 score between the immediate and delayed intervention groups. At 6 months, the mean AHA Life-Simple 7 score of the intervention group increased by 1.9 (SD 1.9) points compared with 0.7 (SD 1.7) point in the control group (P < 0.0001) [14].
Cardiac rehabilitation programs are important components in cardiovascular health and secondary prevention. It is a Class Ia recommendation for secondary prevention after myocardial infarction (MI), percutaneous coronary intervention (PCI), coronary artery bypass graft (CABG), stable angina, or symptomatic peripheral arterial disease [72–75]. Uptake continues to be low, especially in minoritized groups. An observational study from the Veterans Affairs Health Care System and Medicare administrative data showed that cardiac rehabilitation after MI, PCI, or CABG in Medicare patients was 16.3% and in VA patients was 10.3%. In Medicare, participation rates were 17.6% Whites, 7.3% Blacks, and 3.8% Hispanics, whereas in VA, participation rates were 10.4% Whites, 8.9% Blacks, and 12.0% Hispanics [76]. Some of the barriers to access cardiac rehabilitation include lack of insurance coverage or high co-payments [77], language barriers [78], and transportation [79]. However, home-based cardiac rehabilitation has the potential to improve participation across all population by addressing these barriers.
Virtual world technology can support home-based cardiac rehabilitation with programs not only tailored to the patient’s comorbidity but also to the patient’s social, cultural, and language background [80]. In China, Peng et al. randomized 98 participants to receive home-based cardiac rehabilitation versus usual care that included education and regular clinic follow-up. At 4 months, there was a statistically significant change in QoL scores, 6-min walk distance, and heart rate [68]. In a single-arm feasibility study in a US Veterans Affairs Center, Harzand et al. evaluated the change from baseline in blood pressure metabolic equivalent of an asynchronous home-based cardiac rehabilitation program. At 12 weeks, participants (50% Black) showed improvement in metabolic equivalent (5.3 to 6.3, P = 0.008) and mean systolic blood pressure (140.1 to 130.5, P = 0.039). Studies have demonstrated non-inferiority when compared to center-based cardiac rehabilitation, but with limited representation of minorities groups [65, 66, 69, 81]. Home-based cardiac rehabilitation shows potential for the future of cardiac rehabilitation; nevertheless, data in minoritized patient populations are limited and we should strive for increased representation in future trials.
Heart Failure
Heart failure is one of the leading causes of death worldwide [82]. There is a higher prevalence and increased mortality in Black and Hispanic individuals compared with White individuals [83, 84]. Racial and ethnic minoritized groups are less likely to receive appropriate medical therapy and to be included in cardiovascular trials [85, 86]. Digital health tools have the ability to improve heart failure management by obtaining important health-related information such as blood pressure, heart rate, EKGs, weight, and symptoms. It can also address health disparities by remotely monitoring data, thereby decreasing time lost from work or travel, and expenses in travel or follow up visits. Unfortunately, only a few clinical trials report or include diverse populations and even fewer have been done specifically in minoritized groups (Table 3).Table 3 Digital health randomized control trials in heart failure
First author Country Age mean (SD) N Race ethnicity Duration Inclusion criteria Primary outcome Results Device Intervention Health care team intervention
Koehler et al. (2011) [87] Germany 66.9 (10.8) 710 Not reported 18 months Stable HF with LVEF ≤ 35% and admission in previous 2 yrs OR LVEF ≤ 25% Death from any cause 8.4 in RTM group vs 8.7 in UC (HR 0.97, P = 0.87) 3-lead EKG, HBPM, weight scale, smartphone Daily transmission of biometric data Cardiologist or family doctor
Dendale et al. (2012) [88] Belgium 76 (10) 160 Not reported 6 months Decompensated HF All-cause mortality Reduced all-cause mortality in the IT vs usual care Bluetooth scale and HBPM Automatic transfer of data to website + emails with alerts above threshold to clinicians GPs, Cardiologist and Nurse follow-up
Villani et al. (2014) [89] Italy 58 (12) 94 Not reported 6 months HF with LVEF < 35%, NYHA ≥ 2 Number of HF-related hospital days No difference in HF-related hospital days Weight scale, HBPM, mobile phone Upload of measurements and survey to software app that provides machine-based feedback + weekly nurse evaluation Nurse
Vuorinen et al. (2014) [90] Finland 55 (13.7) 100 62% Caucasian
9% African Canadian
7% Asian
12% other
6 months HF with LVEF < 40% BNP, self-care, and quality of life measured by MLHFQ* Significantly improved self-care score Weight scale, HBPM, single-lead ECG recorder, MLHFQ* Automatic upload of readings and questionnaire by email to cardiologist Cardiologist
Dang et al. (2017) [91] US 53 (9.4) 61 76% White Hispanics, 21% AA 3 months HF, smartphone ownership, survival expected > 6 months Self-care efficacy Improved self-care Smartphone (provided by study team) Daily surveys including weight + feedback to physicians Study coordinator providing data to Heart Failure Clinic
Koehler et al. (2018) [92] Germany 70 (10) 1571 Not reported 393 days HF with LVEF ≤ 45% plus hospital admission in last 12 months Percentage of days lost due to a cardiovascular admission or death 4.88 versus 6.64% lost days (P = 0.046), all cause death 7.9 vs 11.3 100 person years (P = 0.028) EKG device, HBPM, weight scales and oximeter, smartphone Daily transmission of biometric data and surveys plus nurse or physician intervention Physician or nurse
Melin et al. (2018) [93] Sweden 75 (8) 72 Not reported 6 months Admitted HF patients Self-care behavior based on 9-item European HF Self-care Behavior Scale Better self-care behaviors in the intervention (16.5 versus 23.5 P ≤ 0.5) Weight scale and tablet computer Patient education, transmission of surveys and weight NA
Chen et al
(2019) [94]
China 61 (15) 767 Not reported 180 days Decompensated CHF, mobile phone ownership, life expectancy > 1 year All-cause mortality SMS and STS significantly reduced the composite endpoint and readmission in 180 days Smartphone Structured telephone support (STS) vs short message service (SMS) vs usual care No
*Minnesota Living with Heart Failure Questionnaire
Koehler et al. randomized 1571 participants in Germany to a telemedicine intervention versus usual care. Participants in the intervention arm received an EKG device, HBPM, weight scales, oximeter, and smartphone that provided daily transmission of data to a telemedicine center where algorithms aided clinicians in patient care. The primary outcome, the percentage of days lost due to unplanned cardiovascular hospital admissions, and all-cause death was 4.88% (95% CI 4.55–5.23) in the remote patient management group and 6.64% (6.19–7.13) in the usual care group (ratio 0.80, 95% CI 0·65–1.00; P = 0.0460) [92]. Chen et al. evaluated mortality 180 days after discharge in participants randomized to a two-telemedicine telephone support system versus usual care. The first-level intervention group received education and reminders via a SMS; the second group received SMS plus structured telephone support system managed by research nurses who called patients every 30 days and allowed patients to call nurses on an as needed basis. The 180-day composite event rate was significantly lower in the SMS and STS groups (50.4 vs 41.3% and 36.5%, both P < 0.05) than in the usual care group, but no difference was observed between the two phone-based intervention groups (P = 0.268) [94].
Unfortunately, there are a lack of on the distribution of racial and ethnic populations in heart failure RCTs. There are limited data on Hispanic, Black, and other minoritized populations and the few studies available did not evaluate hard clinical outcomes. The “Mobile Phone Intervention for Heart Failure in a Minority Urban County Hospital Population” was a feasibility study in Hispanic (76%) and Black (21%) individuals that evaluated a mobile phone intervention to test the system’s usability (ease of use, navigation, readability, confidence, and motivation). The study team provided patients with a telemonitoring program (mobile phone, data usage and free 30-min calls per month) for participants to provide daily heart failure symptoms and weight. At 3 months, participant satisfaction scores was excellent, with a mean score of 6.84 ± 0.46 (rating scale of 1–7); 94% of participants thought that the program was easy to use and 84% thought that navigating the system was not complicated [91].
Arrhythmia Detection
Over the last few years, there has been a growing interest in the use of wearables or home ECG devices as an adjunct to usual care for the detection of arrythmias, most commonly atrial fibrillation. Wearables have the ability to detect irregular rhythm such as atrial fibrillation through continuous monitoring of irregular pulse variation with the use of photoplethysmography or on-demand ECG recording. There is a growing interest in the medical field to test commercially available remote patient monitors in multiple clinical or real-world settings for the detection of cardiac arrhythmias (Table 4).Table 4 Digital health randomized control trials in arrhythmia detection or management
First author Country Age mean (SD) Design N Race ethnicity Duration Inclusion criteria Primary outcome Results Device Intervention Health care team interpretation
William et al
(2018) [95]
US 68 years Single-Center Non-Randomized 52 Not reported NA AF admitted for anti-arrhythmic drug initiation, 38–85 yrs, hx of paroxysmal AF Sensitivity and specificity 96.6% sensitivity and 94.1% specificity of 225 recordings Kardia Mobile Cardiac Monitor coupled to Wi-Fi smart device (iPod) 30 s recordings of lead I ECG + automatic analysis by KMCM algorithm I-lead ECG reviewed by blinded electrophysiologist
Steinhubl et al. (2018) [96] US 72.4 (7.3) RCT + prospective matched cohort 2659 Not reported 4 months ≥ 75 yrs, male > 55 yrs or female > 65 yrs* Incidence of new AF diagnosis at 4 months immediate vs delayed group 3.9% in the immediate versus 0.9% delayed group iRhythm ZioXT Stored data analyzed by an FDA approved algorithm I-lead ECG adjudication by blinded committee
Guo et al. (2019) [97] China 54 Prospective cohort 187,912 Not reported At least 14 days of monitoring ≥ 18 yrs and smartphone ownership AF detection efficacy PPV of 91.6% (91.5–91.8) Smartphone plus smart wrist band AF detection using PPG in the smartphone or wrist band Confirmed by patient’s provider with use of ECG or 24-h Holter monitoring
Perez et al. (2019) [98] US 41 (13) Prospective single group pragmatic study 219,297** 68% White
12% Hispanic
7.7% Black
6.2% Asian
Median 117 days of monitoring ≥ 22 years without prior AF diagnosis or AC use Proportion of notified participants with AF on ECG patch and PPV of irregular pulse intervals PPV 84% (0.76–0.92) Apple Watch + iPhone AF detection by app with irregular pulse notification algorithm Confirmed by ECG patch worn for 7 days
Goldenthal et al. (2019) [99] US 61 (12) RCT 238 77% White, 3% Black
1% Asian
20%, 9% Hispanic
6 months AF patients who underwent RFA or DCCV AF recurrence detection Likelihood of recurrent significantly greater IT*** vs control (HR = 1.56, 1.06–2.3) KardiaMobile + iPhone + cellular servce plan Record ECG daily or with symptoms plus motivational texts Confirmation was determined by patient’s provider
Koh et al. (2021) [100] Malaysia 65.3 (7.4) RCT 203 Not reported 30 days ≥ 55 years without AF and ischemic stroke or TIA within the preceding 12 months AF detection 30-day monitor KardiaMobile vs 24-h Holter 9.5 vs 2% IT vs control (P = 0.024) KardiaMobile Use KardiaMobile monitor 3 times a day for 30 days I-lead ECG adjudicated by blinded electrophysiologist
Lubitz et al. (2022) [101] US 74 (7) Cluster RCT 30,715 82.4% White, 5.3% Black, 2.2% Hispanic 1 year ≥ 65 years New AF diagnosis at 1 year 1.72% vs 1.59% IT vs control P = 0.38 KardiaMobile + iPad Screening AF at primary care clinic with tracings reviewed by cardiologist I-lead ECG reviewed by independent cardiologist
Confirmation with 12-lead ECG determined by patient’s PCP
*And any of the following diagnosis: prior stroke or TIA, heart failure, diabetes and hypertension, mitral valve disease, left ventricular hypertrophy, COPD on home O2, sleep apnea, history of pulmonary embolism, history of myocardial infarction or obesity
**450 returned patches
***IT = intervention
The Huawei Heart Study screened for atrial fibrillation using a PPG monitoring app on a smartphone and or smartwatch in adults across China (mean age 35 years, 86.7% male). In total, 424 participants (mean age 54 years, 87.0% male) received a “suspected AF” notification, which was confirmed in 227 individuals (positive predictive value of PPG signals being 91.6%, CI 91.5 to 91.8%) [97]. The Apple Heart Study enrolled 219,297 participants with average age of 41 years (± 13 years) (68% White, 12% Hispanic, 7.7% Black, and 6.2% Asian) to evaluate the efficacy of the Apple Watch detect atrial fibrillation in a real-world setting. Over a median monitoring time of 117 days, irregular pulse notifications were received by 2161 participants (0.52%), ranging from 3.1% of those 65 years of age or older to 0.16% of those 22 to 40 years of age. Of the 2089 irregular tachograms sampled from participants who had received a notification for analysis, 1489 showed simultaneous atrial fibrillation on ECG patch monitoring, resulting in a positive predictive value of the individual tachogram of 0.71 (97.5% CI, 0.69 to 0.74) [98].
Currently, there are less data in the Hispanic or Black populations. Lubitz et al. looked at the ability of the KardiaMobile device to screen for atrial fibrillation in patients older than 65 years of age without prevalent atrial fibrillation attending a primary care visit [101]. The KardiaMobile was not superior to usual care in the detection of new onset atrial fibrillation. Data are also lacking for other racial or ethnic populations. The “Mobile Health Intervention for Rural Atrial Fibrillation” study aims to test the efficacy of a mobile health application virtual coach coupled to a heart rhythm monitor (Kardia) in patients with atrial fibrillation to improve adherence to oral anticoagulation in a rural population of Western Pennsylvania. This will be one of the first studies using commercially available heart rate and rhythm monitors to be tested in rural underserved populations. The Fitbit Heart Study will be the largest study to date, enrolling 450,000 participants across the USA [102]. Part of the inclusion criteria include ownership of a Fitbit and smartphone device. Large-scale studies that include a greater number of diverse participants could provide important information regarding the accuracy of wearable devices for the detection of atrial fibrillation in across all racial and ethnic populations. If a difference exists, further studies should be aimed at determining how we can improve the accuracy in hardware or software.
There has been an increase in the use of wearable devices such as fitness trackers and smart watches (e.g., Fitbit, Apple Watch) to track activity, sleep, oxygenation, and heart rate. As we expand the use of wearables into clinical practice, it is crucial that we provide access to wearable devices to all communities who may benefit from early arrythmia detection and guideline directed treatments.
Technology companies and the medical community developing and testing these devices with its algorithms should ensure they provide reliable information in patients of all skin tones and age groups. The accuracy of wearables to detect certain metrics, such as oxygenation or heart rate, continues to be an issue. Oxygen saturation and heart rate in fitness trackers are measured by photoplethysmography green light signaling. Research in the last decade revealed that dark brown skin type showed significantly lower modulation, perhaps due to absorption of the light by melanin [103]. Wearables devices have been shown to be less accurate in darker skin tones [104]. Unfortunately, recent data show that this issue is not restricted to fitness trackers. A retrospective multi-center study in the Veterans Health Administration showed that Black patients had higher probability of having occult hypoxemia in the inpatient setting when oxygen saturation is measured by pulse oximeter [105]. It is important for fitness tracking companies to be forthcoming with the limitations present to measure certain metrics in population with darker skin tones [106].
Digital Inclusion
The National Digital Inclusion Alliance (NDIA) defines digital inclusion as the “activities necessary to ensure that all individuals and communities, including the most disadvantaged, have access to and use of Information and Communication Technologies.” It is clear that the elderly, certain racial and ethnic, and lower socioeconomic populations continue to be underrepresented in clinical trials of digital health interventions in cardiovascular disease. There is a continued need to test digital health solutions in underrepresented communities to better understand which interventions provide the most benefit. Beyond clinical trials, there also continues to be a gap in the access to key digital infrastructure in these populations. Attaining digital health equity in cardiovascular care not only requires increased representation of vulnerable populations in clinical trials but also ensured access to the different component of information and communication technologies. In order to improve our current state, proposed solutions should involve multilevel interventions at the individual, family, community, services, and policy level (Fig. 1) [107].Fig. 1 Digital literacies and social determinants of health. Digital literacy and access, including skills, connectivity, devices, and training and technical support, relate to all other domains of social determinants of health. With permission from Sieck et al., with no changes made [91]. https://creativecommons.org/licenses/by/4.0/
At the individual level, patients require access to affordable devices (smartphone, computer, tablets, etc.), continued access to digital health literacy education, and applications to be created to their level of education, cultural, racial, and ethnic background. The Affordable Connectivity Program, which replaces the Emergency Broadband Benefit, provides the Federal Communications Commission with funds to provide broadband monthly discounts to eligible households in the hopes to improve internet access to communities in need [108]. It is crucial to provide continued, affordable, and easy to access digital education in settings where an individual feels comfortable to address the skills and knowledge gaps in digital health. The American Library Association Digital Literacy Task Force has provided recommendations and online resources to aid schools, academic, and public libraries increase the access to digital health education in their respective communities [109].
Technology companies should partner with communities starting with the inception of the design process. Involvement of these key stakeholders from underrepresented communities in product design and software development will ensure that products are made for and used by a broader audience [110]. Additionally, there needs to be increased representation of diverse employees in the tech sector; in many if the big technology companies, less than 5% identify as Hispanic or Black [111, 112]. Increased representation can not only increase trust when engaging underrepresented communities but enrich the design team’s knowledge of the environment where these products are expected to be deployed. In order to advance digital health equity, we should design and develop products aimed at including underrepresented communities with consideration to social determinant of health—based on where people are born, grow, live, work and age.
Perhaps one of the most important domains in order to effect change across all levels is policy changes. Existing programs have expanded the access to devices and broadband internet coverage in low-income communities. The Emergency Connectivity Fund (ECF) is a $7 billion program targeted at helping schools and libraries acquire devices and broadband equipment. Since its first cycle in 2021, the ECF has funded close to 12 million devices and over 7 million broadband connections [26]. Additionally, the Digital Inclusion Act programs not only aim to increase internet access, but support programs to provide underrepresented communities with skills and training necessary to successfully use the internet [113]. To accelerate health equity, we should continue to collaborate with policy makers to renew and expand available programs [5].
Conclusion
Digital health interventions show incredible potential to improve cardiovascular disease detection, prevention and management by obtaining longitudinal, continuous, and actionable patient data; increasing access to care; and decreasing delivery barriers and cost. As the expansion of digital health technologies continues, it is vital to increase representation of minoritized groups in all stages of the process: product development (needs findings and screening, concept generation, product creation and testing), clinical research (pilot studies, feasibility studies, and randomized control trials), and finally health services deployment.
Author Contribution
MFH drafted the manuscript. FR contributed to draft, edit, and supervisory role of the manuscript.
Funding
Dr. Rodriguez was funded by grants from the NIH National Heart, Lung, and Blood Institute (1K01HL144607), the American Heart Association/Harold Amos Faculty Development program, and the Doris Duke Charitable Foundation (Grant #2022051).
Declarations
Conflict of Interest
Dr. Fatima Rodriguez reports equity from HealthPals and Carta and consulting fees from Novartis, Novo Nordisk, and Amgen outside the submitted work. Dr. Mario Funes Hernandez has no conflict of interest to disclose.
This article is part of the Topical Collection on Technology and Cardiovascular Health
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
1. Koonin LM Hoots B Tsang CA Leroy Z Farris K Jolly T Trends in the use of telehealth during the emergence of the COVID-19 pandemic - United States, January-March 2020 MMWR Morb Mortal Wkly Rep 2020 69 43 1595 1599 10.15585/mmwr.mm6943a3 33119561
2. Eberly LA Kallan MJ Julien HM Haynes N Khatana SAM Nathan AS Patient characteristics associated with telemedicine access for primary and specialty ambulatory care during the COVID-19 pandemic JAMA Netw Open 2020 3 12 e2031640 10.1001/jamanetworkopen.2020.31640 33372974
3. Jaffe DH Lee L Huynh S Haskell TP Health inequalities in the use of telehealth in the United States in the lens of COVID-19 Popul Health Manag 2020 23 5 368 377 10.1089/pop.2020.0186 32816644
4. Durocher K Boparai N Jankowicz D Strudwick G Identifying technology industry-led initiatives to address digital health equity Digit Health 2021 7 20552076211056156 35340562
5. Policy - National Digital Inclusion Alliance 2022 [Available from: https://www.digitalinclusion.org/policy/.
6. Merriam-Webster.com Dictionary sv. “Digital Divide”: @MerriamWebster; 2022 [Available from: https://www.merriam-webster.com/dictionary/digital+divide.
7. United States. Department of Health and Human Services. Task Force on Black and Minority Health. Report of the Secretary’s Task Force on Black & Minority Health. Washington, D.C.: U.S. Dept. of Health and Human Services; 1985. <v. 1, 3–4, pts. 1–2 5–8 in > p.
8. Peterson ED, Shaw LK, DeLong ER, Pryor DB, Califf RM, Mark DB. Racial variation in the use of coronary-revascularization procedures. Are the differences real? Do they matter? N Engl J Med. 1997;336(7):480–6.
9. Borrell LN Vaughan R An AJPH supplement toward a unified research approach for minority health and health disparities Am J Public Health 2019 109 S1 S6 S7 10.2105/AJPH.2019.304963 30699024
10. National Center for Health Statistics (U.S.). Healthy People 2010: final review. Hyattsville, MD: U.S. Dept. of Health and Human Services, Centers for Disease Control and Prevention, National Center for Health Statistics; 2012.
11. Equity vs. equality: what’s the difference? 2021 [Available from: https://onlinepublichealth.gwu.edu/resources/equity-vs-equality/.
12. Carnethon MR Pu J Howard G Albert MA Anderson CAM Bertoni AG Cardiovascular Health in African Americans: a scientific statement from the American Heart Association Circulation 2017 136 21 e393 e423 10.1161/CIR.0000000000000534 29061565
13. Brewer LC Hayes SN Caron AR Derby DA Breutzman NS Wicks A Promoting cardiovascular health and wellness among African-Americans: community participatory approach to design an innovative mobile-health intervention PLoS ONE 2019 14 8 e0218724 10.1371/journal.pone.0218724 31430294
14. Brewer LC Jenkins S Hayes SN Kumbamu A Jones C Burke LE Community-based, cluster-randomized pilot trial of a cardiovascular mobile health intervention: preliminary findings of the FAITH! trial Circulation 2022 146 3 175 190 10.1161/CIRCULATIONAHA.122.059046 35861762
15. Omboni S Benczur B McManus RJ Editorial: Digital health in cardiovascular medicine Front Cardiovasc Med 2021 8 810992 10.3389/fcvm.2021.810992 34957268
16. Kario K, Harada N, Okura A. Digital therapeutics in hypertension: evidence and perspectives. Hypertension. 2022:HYPERTENSIONAHA12219414.
17. Coravos A Goldsack JC Karlin DR Nebeker C Perakslis E Zimmerman N Digital medicine: a primer on measurement Digit Biomark 2019 3 2 31 71 10.1159/000500413 32095767
18. Krishnaswami A Beavers C Dorsch MP Dodson JA Masterson Creber R Kitsiou S Gerotechnology for older adults with cardiovascular diseases: JACC State-of-the-Art Review J Am Coll Cardiol 2020 76 22 2650 2670 10.1016/j.jacc.2020.09.606 33243384
19. Uscher-Pines L Sousa J Jones M Whaley C Perrone C McCullough C Telehealth use among safety-net organizations in california during the COVID-19 pandemic JAMA 2021 325 11 1106 1107 10.1001/jama.2021.0282 33528494
20. Tomer A, Fishbane, L., Siefer, A., & Callahan, B. How broadband can deliver health and equity to all communities. Brookings Metro; 2020.
21. Perzynski AT Roach MJ Shick S Callahan B Gunzler D Cebul R Patient portals and broadband internet inequality J Am Med Inform Assoc 2017 24 5 927 932 10.1093/jamia/ocx020 28371853
22. Sophia Campbell JRC, and David Wessel. The benefits and costs of broadband expansion: @BrookingsInst; 2021 [updated 2021–08–18. Available from: https://www.brookings.edu/blog/up-front/2021/08/18/the-benefits-and-costs-of-broadband-expansion/.
23. Atske S PA. Home broadband adoption, computer ownership vary by race, ethnicity in the U.S.: @pewresearch; 2022 [Available from: https://www.pewresearch.org/fact-tank/2021/07/16/home-broadband-adoption-computer-ownership-vary-by-race-ethnicity-in-the-u-s/.
24. E V. Digital divide persists even as Americans with lower incomes make gains in tech adoption. 2022.
25. Warnock RG. S.2729 - 117th Congress (2021–2022): device access for every American Act [webpage]. 2021 [updated 09/14/2021. Available from: https://www.congress.gov/bill/117th-congress/senate-bill/2729.
26. Emergency Connectivity Fund Federal Communications Commission. 2021 [updated 11-16-2022]. Retrieved October 24, 2022 from: https://www.fcc.gov/emergency-connectivity-fund.
27. Holko M Litwin TR Munoz F Theisz KI Salgin L Jenks NP Wearable fitness tracker use in federally qualified health center patients: strategies to improve the health of all of us using digital health devices NPJ Digit Med 2022 5 1 53 10.1038/s41746-022-00593-x 35469045
28. Bayoumy K Gaber M Elshafeey A Mhaimeed O Dineen EH Marvel FA Smart wearable devices in cardiovascular care: where we are and how to move forward Nat Rev Cardiol 2021 18 8 581 599 10.1038/s41569-021-00522-7 33664502
29. Story MF Maximizing usability: the principles of universal design Assist Technol 1998 10 1 4 12 10.1080/10400435.1998.10131955 10181150
30. The 7 principles. Centre for excellence in universal design. (n.d.). Retrieved October 24, 2022, from https://universaldesign.ie/What-is-Universal-Design/The-7-Principles/.
31. Blacklow SO Lisker S Ng MY Sarkar U Lyles C Usability, inclusivity, and content evaluation of COVID-19 contact tracing apps in the United States J Am Med Inform Assoc 2021 28 9 1982 1989 10.1093/jamia/ocab093 34022053
32. Lyles CR, Adrian Aguilera, Oanh Nguyen, Urmimala Sarkar. Bridging the digital health divide: how designers can create more inclusive digital. 2022.
33. Astbrink G Beekhuyzen J The synergies between universal design and user-centred design 2003 Brisbane, Australia Griffith University School of Computing and Information Technology
34. Unertl KM Schaefbauer CL Campbell TR Senteio C Siek KA Bakken S Integrating community-based participatory research and informatics approaches to improve the engagement and health of underserved populations J Am Med Inform Assoc 2016 23 1 60 73 10.1093/jamia/ocv094 26228766
35. (NCD-RisC) NRFC. Worldwide trends in hypertension prevalence and progress in treatment and control from 1990 to 2019: a pooled analysis of 1201 population-representative studies with 104 million participants. Lancet. 2021;398(10304):957–80.
36. Mills KT Stefanescu A He J The global epidemiology of hypertension Nat Rev Nephrol 2020 16 4 223 237 10.1038/s41581-019-0244-2 32024986
37. Artinian NT Flack JM Nordstrom CK Hockman EM Washington OG Jen KL Effects of nurse-managed telemonitoring on blood pressure at 12-month follow-up among urban African Americans Nurs Res 2007 56 5 312 322 10.1097/01.NNR.0000289501.45284.6e 17846552
38. Bosworth HB Powers BJ Olsen MK McCant F Grubber J Smith V Home blood pressure management and improved blood pressure control: results from a randomized controlled trial Arch Intern Med 2011 171 13 1173 1180 10.1001/archinternmed.2011.276 21747013
39. Hebert PL Sisk JE Tuzzio L Casabianca JM Pogue VA Wang JJ Nurse-led disease management for hypertension control in a diverse urban community: a randomized trial J Gen Intern Med 2012 27 6 630 639 10.1007/s11606-011-1924-1 22143452
40. Bove AA Homko CJ Santamore WP Kashem M Kerper M Elliott DJ Managing hypertension in urban underserved subjects using telemedicine–a clinical trial Am Heart J 2013 165 4 615 621 10.1016/j.ahj.2013.01.004 23537980
41. Green BB Cook AJ Ralston JD Fishman PA Catz SL Carlson J Effectiveness of home blood pressure monitoring, Web communication, and pharmacist care on hypertension control: a randomized controlled trial JAMA 2008 299 24 2857 2867 10.1001/jama.299.24.2857 18577730
42. Parati G Omboni S Albini F Piantoni L Giuliano A Revera M Home blood pressure telemonitoring improves hypertension control in general practice. The TeleBPCare study J Hypertens 2009 27 1 198 203 10.1097/HJH.0b013e3283163caf 19145785
43. McManus RJ Mant J Bray EP Holder R Jones MI Greenfield S Telemonitoring and self-management in the control of hypertension (TASMINH2): a randomised controlled trial Lancet 2010 376 9736 163 172 10.1016/S0140-6736(10)60964-6 20619448
44. Wakefield BJ Holman JE Ray A Scherubel M Adams MR Hillis SL Effectiveness of home telehealth in comorbid diabetes and hypertension: a randomized, controlled trial Telemed J E Health 2011 17 4 254 261 10.1089/tmj.2010.0176 21476945
45. Piette JD Datwani H Gaudioso S Foster SM Westphal J Perry W Hypertension management using mobile technology and home blood pressure monitoring: results of a randomized trial in two low/middle-income countries Telemed J E Health 2012 18 8 613 620 10.1089/tmj.2011.0271 23061642
46. Margolis KL Asche SE Bergdall AR Dehmer SP Groen SE Kadrmas HM Effect of home blood pressure telemonitoring and pharmacist management on blood pressure control: a cluster randomized clinical trial JAMA 2013 310 1 46 56 10.1001/jama.2013.6549 23821088
47. Magid DJ Olson KL Billups SJ Wagner NM Lyons EE Kroner BA A pharmacist-led, American Heart Association Heart360 Web-enabled home blood pressure monitoring program Circ Cardiovasc Qual Outcomes 2013 6 2 157 163 10.1161/CIRCOUTCOMES.112.968172 23463811
48. Kerry SM Markus HS Khong TK Cloud GC Tulloch J Coster D Home blood pressure monitoring with nurse-led telephone support among patients with hypertension and a history of stroke: a community-based randomized controlled trial CMAJ 2013 185 1 23 31 10.1503/cmaj.120832 23128283
49. McKinstry B Hanley J Wild S Pagliari C Paterson M Lewis S Telemonitoring based service redesign for the management of uncontrolled hypertension: multicentre randomised controlled trial BMJ 2013 346 f3030 10.1136/bmj.f3030 23709583
50. Stewart K George J Mc Namara KP Jackson SL Peterson GM Bereznicki LR A multifaceted pharmacist intervention to improve antihypertensive adherence: a cluster-randomized, controlled trial (HAPPy trial) J Clin Pharm Ther 2014 39 5 527 534 10.1111/jcpt.12185 24943987
51. Leiva A Aguiló A Fajó-Pascual M Moreno L Martín MC Garcia EM Efficacy of a brief multifactorial adherence-based intervention in reducing blood pressure: a randomized clinical trial Patient Prefer Adherence 2014 8 1683 1690 10.2147/PPA.S66927 25525344
52. McManus RJ Mant J Haque MS Bray EP Bryan S Greenfield SM Effect of self-monitoring and medication self-titration on systolic blood pressure in hypertensive patients at high risk of cardiovascular disease: the TASMIN-SR randomized clinical trial JAMA 2014 312 8 799 808 10.1001/jama.2014.10057 25157723
53. Ogedegbe G Tobin JN Fernandez S Cassells A Diaz-Gloster M Khalida C Counseling African Americans to control hypertension: cluster-randomized clinical trial main effects Circulation 2014 129 20 2044 2051 10.1161/CIRCULATIONAHA.113.006650 24657991
54. Yi SS Tabaei BP Angell SY Rapin A Buck MD Pagano WG Self-blood pressure monitoring in an urban, ethnically diverse population: a randomized clinical trial utilizing the electronic health record Circ Cardiovasc Qual Outcomes 2015 8 2 138 145 10.1161/CIRCOUTCOMES.114.000950 25737487
55. Bobrow K Farmer AJ Springer D Shanyinde M Yu LM Brennan T Mobile phone text messages to support treatment adherence in adults with high blood pressure (SMS-text adherence support [StAR]): a single-blind, randomized trial Circulation 2016 133 6 592 600 10.1161/CIRCULATIONAHA.115.017530 26769742
56. Frias J Virdi N Raja P Kim Y Savage G Osterberg L Effectiveness of digital medicines to improve clinical outcomes in patients with uncontrolled hypertension and type 2 diabetes: prospective, open-label, cluster-randomized pilot clinical trial J Med Internet Res 2017 19 7 e246 10.2196/jmir.7833 28698169
57. McManus RJ Mant J Franssen M Nickless A Schwartz C Hodgkinson J Efficacy of self-monitored blood pressure, with or without telemonitoring, for titration of antihypertensive medication (TASMINH4): an unmasked randomised controlled trial Lancet 2018 391 10124 949 959 10.1016/S0140-6736(18)30309-X 29499873
58. McManus RJ Little P Stuart B Morton K Raftery J Kelly J Home and Online Management and Evaluation of Blood Pressure (HOME BP) using a digital intervention in poorly controlled hypertension: randomised controlled trial BMJ 2021 372 m4858 10.1136/bmj.m4858 33468518
59. Nwokeji U Spaulding EM Shan R Turkson-Ocran RA Baptiste D Koirala B Health information technology use among persons with self-reported atherosclerotic cardiovascular disease: analysis of the 2011–2018 National Health Interview Survey J Med Internet Res 2021 23 8 e23765 10.2196/23765 34397391
60. Brath H Morak J Kästenbauer T Modre-Osprian R Strohner-Kästenbauer H Schwarz M Mobile health (mHealth) based medication adherence measurement - a pilot trial using electronic blisters in diabetes patients Br J Clin Pharmacol 2013 76 Suppl 1 47 55 10.1111/bcp.12184 24007452
61. Petrella RJ Stuckey MI Shapiro S Gill DP Mobile health, exercise and metabolic risk: a randomized controlled trial BMC Public Health 2014 14 1082 10.1186/1471-2458-14-1082 25326074
62. Chow CK Redfern J Hillis GS Thakkar J Santo K Hackett ML Effect of lifestyle-focused text messaging on risk factor modification in patients with coronary heart disease: a randomized clinical trial JAMA 2015 314 12 1255 1263 10.1001/jama.2015.10945 26393848
63. Anand SS Samaan Z Middleton C Irvine J Desai D Schulze KM A digital health intervention to lower cardiovascular risk: a randomized clinical trial JAMA Cardiol 2016 1 5 601 606 10.1001/jamacardio.2016.1035 27438754
64. Salisbury C O’Cathain A Thomas C Edwards L Gaunt D Dixon P Telehealth for patients at high risk of cardiovascular disease: pragmatic randomised controlled trial BMJ 2016 353 i2647 10.1136/bmj.i2647 27252245
65. Skobel E Knackstedt C Martinez-Romero A Salvi D Vera-Munoz C Napp A Internet-based training of coronary artery patients: the Heart Cycle Trial Heart Vessels 2017 32 4 408 418 10.1007/s00380-016-0897-8 27730298
66. Hwang R Bruning J Morris NR Mandrusiak A Russell T Home-based telerehabilitation is not inferior to a centre-based program in patients with chronic heart failure: a randomised trial J Physiother 2017 63 2 101 107 10.1016/j.jphys.2017.02.017 28336297
67. Harzand A Witbrodt B Davis-Watts ML Alrohaibani A Goese D Wenger NK Feasibility of a smartphone-enabled cardiac rehabilitation program in male veterans with previous clinical evidence of coronary heart disease Am J Cardiol 2018 122 9 1471 1476 10.1016/j.amjcard.2018.07.028 30217377
68. Peng X Su Y Hu Z Sun X Li X Dolansky MA Home-based telehealth exercise training program in Chinese patients with heart failure: a randomized controlled trial Medicine (Baltimore) 2018 97 35 e12069 10.1097/MD.0000000000012069 30170422
69. Maddison R Rawstorn JC Stewart RAH Benatar J Whittaker R Rolleston A Effects and costs of real-time cardiac telerehabilitation: randomised controlled non-inferiority trial Heart 2019 105 2 122 129 10.1136/heartjnl-2018-313189 30150328
70. Tekkeşin A Hayıroğlu M Çinier G Özdemir YS İnan D Yüksel G Lifestyle intervention using mobile technology and smart devices in patients with high cardiovascular risk: a pragmatic randomised clinical trial Atherosclerosis 2021 319 21 27 10.1016/j.atherosclerosis.2020.12.020 33465658
71. Bae JW Woo SI Lee J Park SD Kwon SW Choi SH mHealth interventions for lifestyle and risk factor modification in coronary heart disease: randomized controlled trial JMIR Mhealth Uhealth 2021 9 9 e29928 10.2196/29928 34559058
72. Hillis LD Smith PK Anderson JL Bittl JA Bridges CR Byrne JG 2011 ACCF/AHA guideline for coronary artery bypass graft surgery: executive summary: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines J Thorac Cardiovasc Surg 2012 143 1 4 34 10.1016/j.jtcvs.2011.10.015 22172748
73. O’Gara PT Kushner FG Ascheim DD Casey DE Jr Chung MK de Lemos JA 2013 ACCF/AHA guideline for the management of ST-elevation myocardial infarction: executive summary: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines: developed in collaboration with the American College of Emergency Physicians and Society for Cardiovascular Angiography and Interventions Catheter Cardiovasc Interv 2013 82 1 E1 27 10.1002/ccd.24776 23299937
74. Yancy CW Jessup M Bozkurt B Butler J Casey DE Jr Drazner MH 2013 ACCF/AHA guideline for the management of heart failure: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines J Am Coll Cardiol 2013 62 16 e147 239 10.1016/j.jacc.2013.05.019 23747642
75. Amsterdam EA Wenger NK Brindis RG Casey DE Jr Ganiats TG Holmes DR Jr 2014 AHA/ACC guideline for the management of patients with non-ST-elevation acute coronary syndromes: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines J Am Coll Cardiol 2014 64 24 e139 e228 10.1016/j.jacc.2014.09.017 25260718
76. Beatty AL Truong M Schopfer DW Shen H Bachmann JM Whooley MA Geographic variation in cardiac rehabilitation participation in medicare and veterans affairs populations: opportunity for improvement Circulation 2018 137 18 1899 1908 10.1161/CIRCULATIONAHA.117.029471 29305529
77. Mead H Ramos C Grantham SC Drivers of racial and ethnic disparities in cardiac rehabilitation use: patient and provider perspectives Med Care Res Rev 2016 73 3 251 282 10.1177/1077558715606261 26400868
78. Brady S Purdham D Oh P Grace S Clinical and sociodemographic correlates of referral for cardiac rehabilitation following cardiac revascularization in Ontario Heart Lung 2013 42 5 320 325 10.1016/j.hrtlng.2013.07.001 23998380
79. Shanmugasegaram S Oh P Reid RD McCumber T Grace SL Cardiac rehabilitation barriers by rurality and socioeconomic status: a cross-sectional study Int J Equity Health 2013 12 72 10.1186/1475-9276-12-72 23985017
80. Content VG Abraham HM Kaihoi BH Olson TP Brewer LC Novel virtual world-based cardiac rehabilitation program to broaden access to underserved populations: a patient perspective JACC Case Rep 2022 4 14 911 914 10.1016/j.jaccas.2022.05.027 35912322
81. Thomas RJ Beatty AL Beckie TM Brewer LC Brown TM Forman DE Home-based cardiac rehabilitation: a scientific statement from the American Association of Cardiovascular and Pulmonary Rehabilitation, the American Heart Association, and the American College of Cardiology J Am Coll Cardiol 2019 74 1 133 153 10.1016/j.jacc.2019.03.008 31097258
82. Bragazzi NL Zhong W Shu J Abu Much A Lotan D Grupper A Burden of heart failure and underlying causes in 195 countries and territories from 1990 to 2017 Eur J Prev Cardiol 2021 28 15 1682 1690 10.1093/eurjpc/zwaa147 33571994
83. Virani SS Alonso A Benjamin EJ Bittencourt MS Callaway CW Carson AP Heart disease and stroke statistics-2020 update: a report from the American Heart Association Circulation 2020 141 9 e139 e596 10.1161/CIR.0000000000000757 31992061
84. Glynn P Lloyd-Jones DM Feinstein MJ Carnethon M Khan SS Disparities in cardiovascular mortality related to heart failure in the United States J Am Coll Cardiol 2019 73 18 2354 2355 10.1016/j.jacc.2019.02.042 31072580
85. Prasanna A Miller HN Wu Y Peeler A Ogungbe O Plante TB Recruitment of Black adults into cardiovascular disease trials J Am Heart Assoc 2021 10 17 e021108 10.1161/JAHA.121.021108 34431310
86. Vilcant V Ceron C Verma G Zeltser R Makaryus AN Inclusion of under-represented racial and ethnic groups in cardiovascular clinical trials Heart Lung Circ 2022 31 9 1263 1268 10.1016/j.hlc.2022.06.668 35850910
87. Koehler F Winkler S Schieber M Sechtem U Stangl K Böhm M Impact of remote telemedical management on mortality and hospitalizations in ambulatory patients with chronic heart failure: the telemedical interventional monitoring in heart failure study Circulation 2011 123 17 1873 1880 10.1161/CIRCULATIONAHA.111.018473 21444883
88. Dendale P De Keulenaer G Troisfontaines P Weytjens C Mullens W Elegeert I Effect of a telemonitoring-facilitated collaboration between general practitioner and heart failure clinic on mortality and rehospitalization rates in severe heart failure: the TEMA-HF 1 (TElemonitoring in the MAnagement of Heart Failure) study Eur J Heart Fail 2012 14 3 333 340 10.1093/eurjhf/hfr144 22045925
89. Villani A Malfatto G Compare A Della Rosa F Bellardita L Branzi G Clinical and psychological telemonitoring and telecare of high risk heart failure patients J Telemed Telecare 2014 20 8 468 475 10.1177/1357633X14555644 25339632
90. Vuorinen AL Leppänen J Kaijanranta H Kulju M Heliö T van Gils M Use of home telemonitoring to support multidisciplinary care of heart failure patients in Finland: randomized controlled trial J Med Internet Res 2014 16 12 e282 10.2196/jmir.3651 25498992
91. Dang S Karanam C Gómez-Orozco C Gómez-Marín O Mobile phone intervention for heart failure in a minority urban county hospital population: usability and patient perspectives Telemed J E Health 2017 23 7 544 554 10.1089/tmj.2016.0224 28051761
92. Koehler F Koehler K Deckwart O Prescher S Wegscheider K Kirwan BA Efficacy of telemedical interventional management in patients with heart failure (TIM-HF2): a randomised, controlled, parallel-group, unmasked trial Lancet 2018 392 10152 1047 1057 10.1016/S0140-6736(18)31880-4 30153985
93. Melin M Hägglund E Ullman B Persson H Hagerman I Effects of a tablet computer on self-care, quality of life, and knowledge: a randomized clinical trial J Cardiovasc Nurs 2018 33 4 336 343 10.1097/JCN.0000000000000462 29369123
94. Chen C Li X Sun L Cao S Kang Y Hong L Post-discharge short message service improves short-term clinical outcome and self-care behaviour in chronic heart failure ESC Heart Fail 2019 6 1 164 173 10.1002/ehf2.12380 30478888
95. William AD Kanbour M Callahan T Bhargava M Varma N Rickard J Assessing the accuracy of an automated atrial fibrillation detection algorithm using smartphone technology: The iREAD Study Heart Rhythm 2018 15 10 1561 1565 10.1016/j.hrthm.2018.06.037 30143448
96. Steinhubl SR Waalen J Edwards AM Ariniello LM Mehta RR Ebner GS Effect of a home-based wearable continuous ECG monitoring patch on detection of undiagnosed atrial fibrillation: the mSToPS randomized clinical trial JAMA 2018 320 2 146 155 10.1001/jama.2018.8102 29998336
97. Guo Y Wang H Zhang H Liu T Liang Z Xia Y Mobile photoplethysmographic technology to detect atrial fibrillation J Am Coll Cardiol 2019 74 19 2365 2375 10.1016/j.jacc.2019.08.019 31487545
98. Perez MV Mahaffey KW Hedlin H Rumsfeld JS Garcia A Ferris T Large-scale assessment of a smartwatch to identify atrial fibrillation N Engl J Med 2019 381 20 1909 1917 10.1056/NEJMoa1901183 31722151
99. Goldenthal IL Sciacca RR Riga T Bakken S Baumeister M Biviano AB Recurrent atrial fibrillation/flutter detection after ablation or cardioversion using the AliveCor KardiaMobile device: iHEART results J Cardiovasc Electrophysiol 2019 30 11 2220 2228 10.1111/jce.14160 31507001
100. Koh KT Law WC Zaw WM Foo DHP Tan CT Steven A Smartphone electrocardiogram for detecting atrial fibrillation after a cerebral ischaemic event: a multicentre randomized controlled trial Europace 2021 23 7 1016 1023 10.1093/europace/euab036 33782701
101. Lubitz SA Atlas SJ Ashburner JM Lipsanopoulos ATT Borowsky LH Guan W Screening for atrial fibrillation in older adults at primary care visits: VITAL-AF randomized controlled trial Circulation 2022 145 13 946 954 10.1161/CIRCULATIONAHA.121.057014 35232217
102. Lubitz SA FA, Atlas SJ, McManus DD, Singer DE, Pagoto S, Pantelopoulos A, Foulkes AS. Rationale and design of a large population study to validate software for the assessment of atrial fibrillation from data acquired by a consumer tracker or smartwatch: The Fitbit heart study. Am Heart J2021. p. 16–26.
103. Fallow BA Tarumi T Tanaka H Influence of skin type and wavelength on light wave reflectance J Clin Monit Comput 2013 27 3 313 317 10.1007/s10877-013-9436-7 23397431
104. Shcherbina A, Mattsson CM, Waggott D, Salisbury H, Christle JW, Hastie T, et al. Accuracy in wrist-worn, sensor-based measurements of heart rate and energy expenditure in a diverse cohort. J Pers Med. 2017;7(2).
105. Valbuena VSM Seelye S Sjoding MW Valley TS Dickson RP Gay SE Racial bias and reproducibility in pulse oximetry among medical and surgical inpatients in general care in the Veterans Health Administration 2013–19: multicenter, retrospective cohort study BMJ 2022 378 e069775 10.1136/bmj-2021-069775 35793817
106. Colvonen PJ, DeYoung PN, Bosompra NA, Owens RL. Limiting racial disparities and bias for wearable devices in health science research. Sleep. 2020;43(10).
107. Sieck CJ Sheon A Ancker JS Castek J Callahan B Siefer A Digital inclusion as a social determinant of health NPJ Digit Med 2021 4 1 52 10.1038/s41746-021-00413-8 33731887
108. Affordable Connectivity Program 2021 [updated 2021–12–15. Available from: https://www.fcc.gov/acp.
109. Force DLT. Digital literacy, libraries, and public policy 2013 [Available from: http://hdl.handle.net/11213/16261.
110. Equity-centered design framework — Stanford d.school 2022 [Available from: https://dschool.stanford.edu/resources/equity-centered-design-framework.
111. D M. Diversity in tech — information is beautiful: @infobeautiful; 2022 [Available from: https://informationisbeautiful.net/visualizations/diversity-in-tech/.
112. Fry R, Kennedy, B., Funk, C. STEM jobs see uneven progress in increasing gender, racial and ethnic diversity. 2021.
113. Digital Equity Act Programs | Internet for All 2022 [Available from: https://www.internetforall.gov/program/digital-equity-act-programs.
| 36465151 | PMC9703416 | NO-CC CODE | 2022-11-29 23:21:42 | no | Curr Cardiovasc Risk Rep. 2022 Nov 28;:1-20 | utf-8 | Curr Cardiovasc Risk Rep | 2,022 | 10.1007/s12170-022-00711-0 | oa_other |
==== Front
Wetlands (Wilmington)
Wetlands (Wilmington)
Wetlands (Wilmington, N.c.)
0277-5212
1943-6246
Springer Netherlands Dordrecht
1639
10.1007/s13157-022-01639-2
Applied Wetland Science
Restoration Contributes to Maintain Ecosystem Services and Bio-Cultural Linkages Between Wetlands and Local Communities: a Case from a Botanical Diversity Hotspot in Japan
http://orcid.org/0000-0003-3252-6819
Saeki Ikuyo [email protected]
12
Li Yanuo 3
1 grid.20515.33 0000 0001 2369 4728 Faculty of Life and Environmental Sciences, University of Tsukuba, 1-1-1 Tennodai, Tsukuba, Ibaraki 305-8571 Japan
2 grid.265074.2 0000 0001 1090 2030 Makino Herbarium, Tokyo Metropolitan University, Hachioji, Japan
3 grid.20515.33 0000 0001 2369 4728 Graduate School of Comprehensive Human Sciences, University of Tsukuba, Tsukuba, Japan
28 11 2022
2022
42 8 11728 10 2022
10 11 2022
© The Author(s), under exclusive licence to Society of Wetland Scientists 2022
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
The Circum-Ise Bay region in central Japan is characterized by a high concentration of species-rich seepage wetlands that provide various ecosystem services to local communities. However, the non-native conifers Cryptomeria japonica and Chamaecyparis obtusa have been widely introduced to the wetlands and compete with native plants. Here, we report the results of a 4-year restoration experiment that involved removing the conifers from a seepage wetland and observing the effects on plant composition, diversity, and ecosystem services to local communities. The experiment was conducted at a seepage wetland in Nakatsugawa city, Japan. The wetland includes many threatened and endemic plants but is also dominated by the conifers. We established three experimental plots within the wetland and removed the conifers from two of them. The stem density of overstory (i.e., canopy-tree) and understory (i.e., sub-canopy to shrub) layers in the conifer-removal plots decreased by 50% while simultaneously increasing the proportion of threatened woody plants by 14.3–50.0%. Despite these changes, plant species diversity in the groundcover layer remained high, and threatened and culturally important species became more concentrated on removal plots than on the control. We did not observe any negative regime shift, such as the establishment of introduced species. The restoration appeared to promote the occurrence of plants associated with bio-cultural linkages between the seepage wetland and local communities and that supply multiple ecosystem services.
Supplementary Information
The online version contains supplementary material available at 10.1007/s13157-022-01639-2.
Keywords
Bio-cultural diversity
Conservation
Seepage
Threatened species
Vegetation
http://dx.doi.org/10.13039/100007684 Asahi Glass Foundation 2020-9 Saeki Ikuyo issue-copyright-statement© The Author(s), under exclusive licence to Society of Wetland Scientists 2022
==== Body
pmcIntroduction
Wetlands provide a variety of ecosystem services to local communities (Meli et al. 2014; Mitsch et al. 2015; Behailu et al. 2016; Sims et al. 2019; Tomscha et al. 2021). However, they frequently suffer high pressure from anthropogenic impacts, including land development, eutrophication and water pollution (Brinson and Malvárez 2002; Hájek et al. 2002; Mayers et al. 2009). In particular, wetlands sustained by seeping water, or seepages, are one of the ecosystems most vulnerable to anthropogenic impacts. The geomorphological process delivering the seeping water is usually complex and invisible underground (Kløve et al. 2011). This makes it difficult to predict the responses of seepages to the effects of human activities. Nevertheless, adequate management of these ecosystems is important for retaining their multiple ecosystem services, especially to people living nearby.
The Circum-Ise Bay region in central Honshu, Japan, is known to have a high concentration of seepage wetlands (Ueda 1994). According to the latest survey, there are more than 1600 seepage wetlands in this region (Study Group of Seepage Marsh 2019). The wetlands are in low to hilly areas, typically at elevations from 100 to 500 m a.s.l. The size of seepage wetlands is usually less than 1 ha. They are characterized by the absence of peat deposits and low nutrient availability. The seepage wetlands have been repeatedly formed in the Circum-Ise Bay region for a long time, continually appearing for at least 1 million years (Ueda 1994; Makinouchi 2001). Substrates deposited since the Pliocene contain impermeable silt and clay layers. When the layers become excessively wet, this leads to frequent landslides in mountainous terrain, and this geomorphic process promotes the formation of seepages. The duration of seepage wetlands is variable; small ones can disappear through succession within only 100 years. However, new wetlands sporadically form at low elevations. This geological process has occurred throughout the period of Quaternary climatic oscillations, thereby providing refugia for many native plants. This process has contributed to establishing a floristic group, known as “Tokai Hilly Land Elements” (Ueda 1989, 1994).
One of the major risks threatening the seepage wetlands is the alternation of wetlands to conifer plantations. In the 1960s and 1970s, the Japanese government promoted the expansion of conifer plantations because there was a severe shortage of timber resources for residential construction (Yamaura et al. 2012). The commonly planted species were Cryptomeria japonica (L.f.) D.Don and Chamaecyparis obtusa (Siebold et Zucc.) Endl. The presence of conifers in the canopy layer reduces the light reaching the forest floor, which inhibits the growth of understory and groundcover plants.
Based upon this situation, we performed a restoration experiment in which we removed non-native conifers from seepage wetlands and then evaluated the effects of this treatment by monitoring plant species composition and diversity for 4 years. In general, the biodiversity of wetlands is highly sensitive to abrupt changes in the physical environment (Hobbs and Huenneke 1992; Davis et al. 2000; Saeki 2007). Therefore, even though changes might be made for conservation purposes, we should be careful to avoid strong impacts on the wetlands, and long-term monitoring is recommended. In the present study, we aimed to examine the changes of plant species composition and diversity after the removal of non-native conifers at a seepage wetland in the Circum-Ise Bay region. We focused in particular on the occurrence of threatened and culturally important plants because the seepage wetlands in this region are known to contain a large number of threatened species, and they have a strong cultural association with local communities (Li and Saeki 2018). One of the reasons why biodiversity should be conserved is that it is a basis of our cultural identity, or bio-cultural diversity (Maffi 2001). This concept is especially important for ecosystems which have a close linkage with local communities, such as seepage wetlands.
Methods
Study Area
The experiment was conducted at a seepage wetland in Iwayado village, Nakatsugawa city, Gifu prefecture, Japan (Appendix S1). The area of Iwayado village is approximately 1 km2, and there are several seepage wetlands in the Satoyama, a traditional agricultural landscape in Japan (Takeuchi 2001). The landscape is characterized by a mosaic of rice paddies, agricultural ponds, vegetable-farming fields, deciduous forests, and conifer plantations. We selected one of seepage wetlands in Iwayado privately owned by a local family for this restoration project because the landowner wished to conserve the wetland even though it is not legislatively designated for preservation. The seepage wetland covers about 0.11 ha, which is typical of the seepage wetlands in the Circum-Ise Bay region (Ueda 1994). The site has a gentle slope of 15%. Part of the seepage wetland is forested but the rest is relatively open (Appendix S2). Both forested and open areas were a target of the present restoration project. Vegetation of the study site is characterized by woody and herbaceous vascular plants with occasionally high dominance of sphagnum moss. There is a fine-scale difference in microtopography created by seeping water. No peat deposits were observed. Ground water level was relatively stable through all the seasons; conspicuous fluctuation was not observed during the study period.
Data Collection
In July 2016, we established three 10 m × 20 m plots within the seepage wetland and recorded plant species and diameter at breast height (DBH) for all the stems of woody plants with DBH ≥ 1.5 cm. Both live and dead woody plants were recorded. We then classified each stem as belonging to overstory (≥ 9.0 cm; i.e., canopy-tree) or understory (1.5–9.0 cm; i.e., sub-canopy to shrub) layers. We also established one 5 m × 5 m plot (hereafter, “subplot”) within each 10 m × 20 m plot for investigating groundcover vegetation. This layer includes vascular plants, including woody plants with DBH < 1.5 cm. The location of each subplot was randomly chosen, but avoiding irregular objects on the ground such as large rocks and woody debris. In each subplot, we recorded occurrences of all vascular plants and their coverage. Coverage was recorded using a 10-class scale (class 1, ≤ 0.5%; 2, 0.5–1%; 3, 1–3%; 4, 3–5%; 5, 5–8%; 6, 8–12%; 7, 12–16%; 8, 16–40%; 9, 40–70%; 10, 70–100%). The vegetation data for the groundcover layer prior to the conifer-removal treatment were recorded on 14 July, 10 August, and 30 September 2016. The survey was repeated several times because some plants germinate in different seasons. The results of the 2016 investigation were partially reported by Li and Saeki (2018).
On 26 February 2017, non-native conifers on the two of the three plots described above were cut. We did not do any cutting on the third plot as a control. One of the treatment plots was relatively open with few canopy-size trees, whereas the control plot had a higher density of canopy trees. The other treatment plot was in between these two, with relatively dense canopy-size trees. The target conifer species to be removed were C. japonica and C. obtusa, which were either planted in the 1960s and 1970s or established from dispersed seeds from adjacent plantation forests. During the removal treatment, we measured the DBH of stems removed from the two treatment plots. To examine the effects of conifer removal on groundcover species, we performed groundcover vegetation surveys of the three subplots every year from 2017 to 2020 and compared them with the data taken in 2016 before the removal treatment. The exception was the year 2020 when we could only visit the site once because of the COVID-19 pandemic. The surveys after the removal were conducted on 10 July, 12 August, and 6 October 2017, 8 August, 1 September, and 29 September 2018, 27 August and 30 September 2019, and 29 September 2020.
Data Analysis
For the overstory and understory layers, we used the data recorded in 2016 to calculate the stem density (stems/ha) and proportion of non-native conifers of the total number of stems before the treatments. Stem density and proportion of non-native conifers were also estimated after the removal treatment by extracting the stem numbers and basal areas for the removed conifers from the 2016 data. For the groundcover layer, the coverage of each plant species recorded in each of the subplots was input in table format (Appendix S3), and the mean coverage of each species was calculated by year. To quantify plant diversity, we calculated species richness (S) and Shannon index (H′; Magurran 1988) for each of the plots and subplots. For the overstory and understory layers we used number of stems as an indicator of abundance, whereas for the groundcover layer we used coverage.
To illustrate differences in groundcover species composition before and after the treatment, we performed non-metric multidimensional scaling (NMDS) analysis using the data recorded by subplot for each year. We checked for the occurrence of threatened and near-threatened species using prefectural and national Red Lists (Gifu Prefectural Government 2014; Ministry of Environment 2020) and monitored their numbers before and after the treatment. The NMDS was performed with the package “vegan” (Oksanen et al. 2020) in R ver. 4.1.2 (R Core Team 2021).
An interview survey in a previous study in the Iwayado area (Li and Saeki 2018) showed that the wetland landowners had a variety of knowledge and experiences regarding the plants in and around their wetlands. We first identified the plants to which landowners referred in the interviews as culturally important species and then examined their occurrences in the study plots before and after the treatment. The species defined as culturally important were those used for certain purposes (e.g., food, play, traditional events, and horticulture) or recognized as symbols (i.e., representatives) of the wetlands (Appendix S4).
Results
The number of living stems in the overstory and understory layers decreased markedly in the conifer-removal plots; the change from before to after the treatment was 50% in the overstory and 34–38% in the understory (Table 1; Fig. 1). In these plots, we were able to remove almost all of the non-native conifers. Simultaneously, the proportion of stems of threatened (Acer pycnanthum K. Koch) and near-threatened (Magnolia stellata [Siebold et Zucc.] Maxim.) woody plant species increased (Table 1). S and H′ of the control plot were 9 and 1.87, respectively. Those of conifer-removal plots were much smaller than the control.Table 1 Comparison of vegetational characteristics and diversity of overstory and understory layers before and after removal of non-native conifers. DBH, stem diameter at breast height
Control plot Conifer-removal plot (closed) Conifer-removal plot (open)
Before Before After Before After
Overstory (DBH ≥ 9.0 cm)
No. of living stems (stems/plot)a 23 14 7 2 1
Proportion of stems of non-native conifer species (%)b 43.5 57.1 14.3 50.0 0
Proportion of stems of threatened endemic species (%)c 34.8 14.3 28.6 50.0 100.0
Species richness 9 5 5 2 1
H′d 1.87 1.25 1.55 0.69 NS
Understory (9.0 > DBH ≥ 1.5 cm)
No. of living stems (stems/plot)a 71 122 80 87 54
Proportion of stems of non-native conifer species (%)b 11.3 32.0 0 35.6 0
Proportion of stems of threatened endemic species (%)c 25.4 5.7 8.8 11.5 14.8
Species richness 13 18 16 16 14
H′d 2.14 2.30 2.29 2.35 2.27
aPlot size: 200 m2
bNon-native conifers: Cryptomeria japonica, Chamaecyparis obtusa
cThreatened endemic species: Acer pycnanthum, Magnolia stellata
dH′: Shannon index based on no. of stems
Fig. 1 Comparisons of diameter at breast height (DBH) distribution with and without conifer-removal treatment on a seepage wetland in Iwayado, Nakatsugawa city, Gifu prefecture, Japan. DBH distribution on (a) a control plot without conifer removal, (b) a forested plot with conifer removal, and (c) an open plot with conifer removal. For (b) and (c), top and bottom charts show distributions before and after the treatment, respectively
In the understory layer, S of the forested and open conifer-removal plots decreased from 18 to 16 and 16 to 14, respectively, reflecting the complete removal of the two conifer species (Table 1). H′ of conifer-removal plots did not change much, ranging from 2.30–2.35 before and 2.27–2.29 after the treatment. These values are slightly higher than that in the control (2.14). In both the control and removal plots, the stem numbers in the understory were much higher than in the overstory, which are well described by the inverse-J shape of DBH class distributions (Fig. 1).
In the groundcover layer, we recorded a total of 88 vascular plant species/taxa (Appendix S3). Among the 88 taxa, at least 10 are listed in either national or local Red Lists, and 13 were referred to by landowners as culturally important. NMDS analysis of groundcover vegetation demonstrates marked differences in species composition among the three subplots (Fig. 2). The control plot (P1 in Fig. 2) is plotted at the low end of axis 1. The forested (P2) and open (P3) conifer-removal plots are plotted at the high end of axis 1, and at the low and high ends, respectively, of axis 2. Of the 10 threatened and near-threatened species, 9 were placed on the positive side of axis 1 (Fig. 2). Regarding culturally important species, 9 of 13 were placed on the positive side of axis 1.Fig. 2 Comparisons of species composition of groundcover vascular plants in three experimental plots on a seepage wetland in Iwayado, Nakatsugawa city, Gifu prefecture, Japan, based on non-metric multidimensional scaling (NMDS). Each plot was 5 m × 5 m. Plot 1 (P1) was a control without removal of non-native conifers. Plots 2 (P2) and 3 (P3) were experimental plots where non-native conifers were removed. Conifers were removed in winter 2017. The numbers after decimal point indicates the chronological information: 1–3, 2016; 4–6, 2017; 7–9, 2018; 10–11, 2019; 12, 2020. Names of recorded plants are abbreviated; see Appendix S3 for full scientific names. The green, bold labels indicate threatened or near-threatened plant species listed in the national and prefectural Red Lists. Characters within ellipses indicate culturally important species identified in interviews with local landowners in a previous study (Li and Saeki 2018). See Appendix S4 for details of culturally important species
S and H′ of the groundcover layer were consistently high across the three subplots before and after the treatment (Fig. 3). H′ was 3.42 in the control plot in the first study year (i.e., 2016), and it did not change much after the removal. The numbers of threatened and near-threatened species listed in the national and prefectural Red Lists were higher in the conifer-removal plots than in the control (Fig. 3). The number of culturally important species also remained high in the conifer-removal plots compared to the control. In the forested conifer-removal plot (P2), however, the numbers were slightly lower because of the disappearance of Triantha japonica (Miq.) Baker, A. pycnanthum, and Drosera rotundifolia L. On the other hand, Habenaria radiata (Thunb.) Spreng. (labeled “Hara” in Fig. 2) was newly recorded on the open conifer-removal plot (P3) after the treatment. This species is designated near-threatened, and it was also noted as culturally important by landowners because of its beautiful, heron-like flowers (Appendix S4).Fig. 3 Changes in (a) species richness (S), (b) Shannon index (H′), (c) number of Red List (RL) species, and (d) number of culturally important species of groundcover vascular plants in three experimental subplots on a seepage wetland in Iwayado, Nakatsugawa city, Gifu prefecture, Japan
Discussion
Owing to a high concentration of threatened plants, the Circum-Ise Bay region has been selected as one of the hotspots of plant diversity (Yahara 2002). To our knowledge, this is the first attempt to restore seepage wetlands in this region by removing non-native conifers. We perceive that our conifer-removal project, working with private landowners, was fruitful in restoring native plant diversity in the wetland and conserving its cultural association with local people. In the overstory and understory layers, A. pycnanthum and M. stellata increased in relative dominance after the conifer removal (Table 1; Fig. 1). Both of these species are listed in the national and prefectural Red Lists and also members of the Tokai Hilly Land Elements (Ueda 1989). Seepage wetlands are characterized by high numbers of shrub species (Saeki 2007). In our project, the conifer-removal plots contained 14–16 species within an area of only 200 m2 after the treatment (Table 1). The removal of conifers will likely contribute to the long-term conservation of species richness of the understory layer as well.
Prior to this restoration project, we were concerned about the possibility of a negative regime shift, such as the introduction of exotic species or a marked increase in dominance of a particular common native species. However, the diversity indices of the groundcover remained high, and threatened and near-threatened species were continuously present after the conifer removal (Fig. 3). In an experiment restoring a fen in Sweden (Hedberg et al. 2012), sedges, grasses, sphagnum, and wetland vascular plants and mosses all showed a positive response to clear-cutting, with increases in their coverage. We did not observe such remarkable changes for 4 years after conifer removal. One reason for that might be the poor nutrient conditions of the seepage wetland. The electrical conductivity (EC) of seeping water in the Circum-Ise Bay region is usually around < 30 μS/cm (Study Group of Seepage Marsh 2019). The actual EC values of seeping water at the experimental site after the conifer removal have been 17–25 μS/cm, and total N and P were 1.2 mg/L and < 0.05 mg/L, respectively (I. Saeki, unpublished data). Exotic plants are known to favor nitrogen-rich sites (Chatterjee and Dewanji 2019). It is often difficult to restore wetland vegetation when nitrogen and phosphorus levels are high because this can increase the productivity of invasive species (van der Hoek and Braakhekke 1998; Zedler 2000). Furthermore, there were limited seed sources for exotic and invasive plants within and around our research plots. Except for the two non-native conifers, there were no non-native or invasive plants, such as dwarf bamboo, within the experimental site (Fig. 2; Appendix S3).
Landowners living near the wetlands have rich cultural associations with a wide range of plants (Li and Saeki 2018; Appendix S4). Conifer-removal treatment helps with conserving these associations because culturally important species remain after the treatment. One of the symbolic species, H. radiata, newly occurred in one of the subplots after the treatment (Fig. 2, Appendix S4). According to interviews with landowners, there used to be no conifers in the seepage wetlands when they were young, and wetlands were more open then than today and held H. radiata. The landowners wanted to restore the wetlands to match those in the past, which motivated their agreement for this restoration project. For long-term conservation of the seepage wetlands, positive actions by local communities are essential because it is often difficult to pass legislation to conserve small but local-scale biodiversity hotspots like the seepage wetlands. Note that culturally important species linked with local people are not necessarily threatened species (Fig. 2), which are often targeted for conservation by scientists, conservation organizations, and governments. We argue the importance of paying attention to local perceptions of the value of biodiversity. When trying to conserve the plant communities of these seepage wetlands, we should focus not only on species with scientific and conservation importance as monitoring indicators, such as threatened, local-endemic species, but also on those having cultural value to local people.
In the conifer-removal plots, the number of species appearing or disappearing was relatively high, and thus plant species composition may be changing over the short term. The rapid change in species composition after the treatment is typical in similar restoration projects (e.g., Glennemeier et al. 2020), and this implies that long-term monitoring is necessary. We conclude that conifer removal on the biodiversity-rich seepage wetlands in the Circum-Ise Bay region can be a prioritized option for managers to conserve their unique plant composition, diversity, and cultural association with local people.
Supplementary Information
Below is the link to the electronic supplementary material.Supplementary file1 (DOCX 1770 KB)
Supplementary file2 (XLSX 16.3 KB)
Acknowledgements
We sincerely thank the landowners and local community residents for allowing us to perform this restoration project. This research was partially funded by the Asahi Glass Foundation (Grant no. 2020-9).
Author Contributions
All authors contributed to the study conception and design. Data collection and analysis were performed by Ikuyo Saeki and Yanuo Li. The first draft of the manuscript was written by Ikuyo Saeki and all authors approved the final manuscript.
Funding
This work was partially supported by the Asahi Glass Foundation (Grant no. 2020–9).
Data Availability
Vegetation data we used for the analyses are available from Appendix S3.
Declarations
Competing Interests
The authors have no relevant financial or non-financial interests to disclose.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
Behailu BM Pietilä PE Katko TS Indigenous practices of water management for sustainable services: case of Borana and Konso, Ethiopia SAGE Open 2016 6 2158244016682292 10.1177/2158244016682292
Brinson MM Malvárez AI Temperate freshwater wetlands: types, status, and threats Environmental Conservation 2002 29 115 133 10.1017/S0376892902000085
Chatterjee S Dewanji A Soil nutrients can influence exotic species richness in urban areas: a case study from the city of Kolkata American Journal of Plant Sciences 2019 10 2052 2069 10.4236/ajps.2019.1011145
Davis MA Grime JP Thompson K Fluctuating resources in plant communities: a general theory of invasibility Journal of Ecology 2000 88 528 534 10.1046/j.1365-2745.2000.00473.x
Gifu Prefectural Government (2014) Red data book of Gifu prefecture (Plant). https://www.pref.gifu.lg.jp/page/11109.html. Accessed 31 March 2022
Glennemeier K Packard S Spyreas G Dramatic long-term restoration of an oak woodland due to multiple, sustained management treatments PLoS ONE 2020 15 e0241061 10.1371/journal.pone.0241061 33095813
Hájek M Hekera P Hájková P Spring fen vegetation and water chemistry in the Western Carpathian flysch zone Folia Geobotanica 2002 37 205 224 10.1007/BF02804232
Hedberg P Kotowski W Saetre P Mälson K Rydin H Sundberg S Vegetation recovery after multiple-site experimental fen restorations Biological Conservation 2012 147 60 67 10.1016/j.biocon.2012.01.039
Hobbs RJ Huenneke LF Disturbance, diversity, and invasion: implications for conservation Conservation Biology 1992 6 324 337 10.1046/j.1523-1739.1992.06030324.x
Kløve B Ala-aho P Bertrand G Boukalova Z Ertürk A Goldscheider N Ilmonen J Karakaya N Kupfersberger H Kvœrner J Lundberg A Mileusnić M Moszczynska A Muotka T Preda E Rossi P Siergieiev D Šimek J Wachniew P Angheluta V Widerlund A Groundwater dependent ecosystems. Part I: hydroecological status and trends Environmental Science & Policy 2011 14 770 781 10.1016/j.envsci.2011.04.002
Li Y, Saeki I (2018) Values of seepage wetlands: relationship between wetlands and local communities in Iwayado, Nakatsugawa-city, Japan. Wetland Research 8:81–97. 10.24785/wetlandresearch.WR008010 (In Japanese with English summary)
Maffi L (2001) On Biocultural Diversity: Linking Language, Knowledge and the Environment. Smithsonian Institution Press, Washington, D.C
Magurran AE Ecological Diversity and Its Measurement 1988 New Jersey Princeton University Press
Makinouchi T Stratigraphy of the Tokai Group and evolutional process of the Lake Tokai sedimentary basin Science Report of Toyohashi Museum of Natural History 2001 11 33 39
Mayers J, Batchelor C, Bond I, Hope RA, Morrison E, Wheeler B (2009) Water ecosystem services and poverty under climate change: Key issues and research priorities. In: Natural Resource Issues. International Institute for Environment and Development, London
Meli P Rey Benayas JM Balvanera P Martínez Ramos M Restoration enhances wetland biodiversity and ecosystem service supply, but results are context-dependent: a meta-analysis PLoS ONE 2014 9 e93507 10.1371/journal.pone.0093507 24743348
Ministry of Environment (2020) Red list of vascular plants in Japan. http://www.env.go.jp/press/files/jp/114457.pdf. Accessed August 13, 2020
Mitsch WJ Bernal B Hernandez ME Ecosystem services of wetlands International Journal of Biodiversity Science, Ecosystem Services & Management 2015 11 1 4 10.1080/21513732.2015.1006250
Oksanen J, Blanchet FG, Friendly M, Kindt R, Legendre P, McGlinn D, Minchin PR, O’Hara RB, Simpson GL, Solymos P, Stevens MHH, Szoecs E, Wagner H (2020) Vegan: Community Ecology Package (Version 2.5–5). URL https://cran.r-project.org/web/packages/vegan/vegan.pdf
R Core Team (2021) R: a language and environment for statistical computing. R Foundation for Statistical Computing. https://www.R-project.org/
Saeki I Effects of tree cutting and mowing on plant species composition and diversity of the wetland ecosystems dominated by the endangered maple, Acer pycnanthum Forest Ecology and Management 2007 242 733 746 10.1016/j.foreco.2007.02.009
Sims S, Hilberg L, Reynier W, Kershner J (2019) Seeps and Springs: Northern California Climate Change Vulnerability Assessment Synthesis. Version 1.0. EcoAdapt. http://ecoadapt.org/programs/adaptation-consultations/norcal. Accessed 8 Aug 2021
Study Group of Seepage Marsh (2019) Seepage marshes in Tokai area, Japan. (In Japanese)
Takeuchi K Nature conservation strategies for the ‘SATOYAMA’ and ‘SATOCHI’, habitats for secondary nature in Japan Global Environmental Research 2001 5 193 198
Tomscha SA Bentley S Platzer E Jackson B de Roiste M Hartley S Norton K Deslippe JR Multiple methods confirm wetland restoration improves ecosystem services Ecosystems and People 2021 17 25 40 10.1080/26395916.2020.1863266
Ueda K (1989) Phytogeography of Tokai hilly land element I. Definition. Acta Phytotaxonomica Geobotanica 40:190–202. 10.18942/bunruichiri.KJ00001078646 (In Japanese with English summary)
Ueda K Okada H Ueda K Kadono Y The Origin and evolution of the Tokai hilly land element Natural History of Plants: Evolutionary Studies of Diversity 1994 Sapporo Hokkaido University Press 3 18
van der Hoek D Braakhekke WG Joyce CB Wade PM Restoration of soil chemical conditions of fen-meadow plant communities by water management in the Netherlands European wet grasslands: biodiversity, management and restoration 1998 Wiley 265 275
Yahara T (2002) Evaluation of extinction risk in the plant red data book and its application. In: Yahara T, Kawakubo N (eds) Biology of Conservation and Restoration. Bun-ichi Sogo Shuppan, Tokyo. pp
Yamaura Y Oka H Taki H Ozaki K Tanaka H Sustainable management of planted landscapes: lessons from Japan Biodiversity and Conservation 2012 21 3107 3129 10.1007/s10531-012-0357-4
Zedler JB Progress in wetland restoration ecology Trends in Ecology & Evolution 2000 15 402 407 10.1016/S0169-5347(00)01959-5 10998517
| 36467510 | PMC9703418 | NO-CC CODE | 2022-11-29 23:21:08 | no | Wetlands (Wilmington). 2022 Nov 28; 42(8):117 | utf-8 | Wetlands (Wilmington) | 2,022 | 10.1007/s13157-022-01639-2 | oa_other |
==== Front
Curr Phys Med Rehabil Rep
Curr Phys Med Rehabil Rep
Current Physical Medicine and Rehabilitation Reports
2167-4833
Springer US New York
36466557
373
10.1007/s40141-022-00373-4
Cancer Rehabilitation (C Kline-Quiroz, Section Editor)
Establishing a Cancer Rehabilitation Service in a Middle-Income Country: an Experience from Brazil
http://orcid.org/0000-0002-9287-4025
Leite Victor F. [email protected]
1
http://orcid.org/0000-0001-5675-6665
Cecatto Rebeca Boltes 12
http://orcid.org/0000-0001-5275-0733
Battistella Linamara Rizzo 34
http://orcid.org/0000-0003-3775-6533
de Brito Christina May Moran 15
1 grid.411074.7 0000 0001 2297 2036 Rehabilitation Department, Instituto Do Cancer, Hospital das Clinicas da Faculdade de Medicina da Universidade de São Paulo, Av. Dr. Arnaldo, 251 - Cerqueira César, São Paulo, SP 01246-000 Brazil
2 grid.412295.9 0000 0004 0414 8221 Health Sciences Postgraduate Program, Universidade Nove de Julho - UNINOVE, São Paulo, Brazil
3 grid.11899.38 0000 0004 1937 0722 Faculdade de Medicina FMUSP, Universidade de São Paulo, São Paulo, SP Brazil
4 grid.411074.7 0000 0001 2297 2036 Instituto de Medicina Física E Reabilitação, Hospital das Clinicas da Faculdade de Medicina da Universidade de São Paulo, São Paulo, Brazil
5 grid.413471.4 0000 0000 9080 8521 Rehabilitation Service, Hospital Sírio-Libanês, São Paulo, Brazil
28 11 2022
2022
10 4 339344
28 10 2022
© The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2022, Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Purpose of Review
Our aim is to provide a historical review of the implementation of a cancer rehabilitation center in Brazil, active since 2008. We expect this data to support the implementation of other centers both in Brazil and worldwide.
Recent Findings
Cancer rehabilitation delivery is fragmented and punctuated in most cases, and cancer rehabilitation centers are rare. Data on how to establish rehabilitation centers could facilitate the implementation of new centers. We provide data on what was our strategy for hiring, establishing treatment protocols, barriers, and facilitators. We also provide figures on the number of each rehabilitation specialist, as well as the general standard operating procedures of our rehabilitation center, among other features.
Summary
Establishing cancer rehabilitation centers in a middle-income country is feasible. We expect that our experience may facilitate the establishment of new cancer rehabilitation services and the improvement of current ones.
Keywords
Cancer rehabilitation
Brazil
Rehabilitation center
Rehabilitation program
Middle-income country
Health facilities
issue-copyright-statement© Springer Science+Business Media, LLC, part of Springer Nature 2022
==== Body
pmcIntroduction
Cancer and its treatments can cause several impairments, which rehabilitation has the potential to mitigate and treat [1•, 2, 3]. Unfortunately, comprehensive cancer rehabilitation programs are the exception rather than the rule since the great majority of cancer care centers have punctuated and fragmented rehabilitation care [4–8]. Many barriers hinder the implementation of such programs, for instance, lack of funding, paucity of trained rehabilitation specialists, and lack of a cohesive survivorship care plan [1•, 6, 7]. The knowledge and lessons learned by established cancer rehabilitation centers may serve as valuable tools for the implementation of new centers. Our objective is to provide a narrative review of how we were able to implement and maintain a cancer rehabilitation center in a middle-income country since 2008, and to provide data on rehabilitation usage, personnel, and physical structure of our rehabilitation center. We expect this report will assist with the implementation or improvement of other cancer rehabilitation centers globally.
Brazilian Healthcare
Understanding Brazilian healthcare and how rehabilitation and cancer care are structured in this system is a necessary context for appreciating our experience. Brazil has a universal government-funded public healthcare system, Sistema Único de Saúde (SUS), as well as a supplementary paid system operated by several private operators. The entire population has access to the SUS free of charge, and currently over 25% of Brazilians pay for a supplementary health system [9], as some treatments are not covered by SUS, and also to avoid the delay that may result from the high demand in the public healthcare system. However, about a quarter of those who hire a private health insurance still use the SUS for their cancer treatment [10]. Healthcare is provided by the SUS using three escalating levels of complexity [11]. Primary care is delivered throughout the country by teams responsible for over 3000 citizens each. As the complexity of care grows, more specialized care is provided by fewer institutions.
Rehabilitation services are offered in all levels of healthcare. At the primary level, care is provided by Family Health Teams that provide education, orient the use of assistive technology, and provide community-based rehabilitation, among a spectrum of different services [12]. At the secondary level, rehabilitation care is provided at dedicated rehabilitation facilities, where more specific rehabilitation interventions are provided, including provision of assistive technologies [12]. The primary and secondary levels are responsible for longitudinal rehabilitation care, caring for the subacute/chronic rehabilitation phase, occupational and social reintegration, supported by multi-professional teams and home-based activities among others.
At the tertiary level, rehabilitation care is provided in a more specialized manner, for instance, in cancer or stroke centers [12]. Some of the challenges for providing rehabilitation in the SUS are the chronic public underfunding in rehabilitation care and the need for more rehabilitation specialists [12].
Cancer care is provided in the secondary and tertiary levels of the SUS. There is disparity in the geographical availability of resources for healthcare, as 70% of the specialized cancer care units are concentrated in the south and southeast regions of Brazil, which account for about 22% of the population [10]. Consequently, median length of distanced traveled for cancer care can reach about 500 km (310 mi) in several states, compared to 100 km (62 mi) in São Paulo [13].
Our Experience
Our Institution
The Cancer Institute of the State of Sao Paulo (ICESP—Instituto do Câncer do Estado de São Paulo) was launched in 2008 and is a tertiary-level public university-based hospital dedicated to adult cancer care. This cancer center is part of the biggest university-based public hospital in Latin America, the Hospital das Clínicas of the University of São Paulo’s School of Medicine. ICESP is a tall vertical structure, with a total area of 84,000 m2 (approximately 900,000 ft2), distributed in 28 floors. It has 476 beds (85 of them located at Intensive Care Units), 124 ambulatory offices, 18 surgery rooms, 100 chemotherapy infusion rooms, 6 linear accelerators, 1 for brachytherapy, 4 magnetic resonances, 6 computed tomographies, 1 single photon emission computed tomography, high-intensity focused ultrasound, endoscopic clinic, and robotic surgery. By the end of 2020, it had 5229 employees and had attended 116,000 patients, with 44,000 under current care.
Setting Up the Practice
Since the beginning of ICESP’s activities, rehabilitation services were available at the institution. That was facilitated by a national regulation published in 2005 that mandated the provision of rehabilitation care for all high-complexity oncology services [14]. This determination was a landmark and promoted an important advance for the greater provision of rehabilitation care for cancer patients in the country.
Hiring
As very few professionals with experience in the rehabilitation care of oncological patients were available in 2008, we sought rehabilitation professionals that had experience in specific needs that were prevalent in the cancer population (e.g., musculoskeletal impairments, neurologic rehabilitation, lymphedema, chronic pain, cardiopulmonary rehabilitation). Currently, the Rehabilitation Service has 132 professionals that were progressively hired during the first 2 years of the Institute and now are composed as follows: 100 physical therapists, 8 speech therapists, 6 physiatrists, 5 occupational therapists, 4 neuropsychologists, 3 exercise physiologists, 3 coordinators, 1 manager, 1 nurse assistant, and 1 administrative assistant.
Guidelines and Procedures
Due to the scarcity of cancer rehabilitation guidelines and standard operating procedures (SOPs) at the time when ICESP was launched in 2008, the rehabilitation and clinical teams mobilized to elaborate evidence-based guidelines for rehabilitation protocols and SOPs, which lead to published articles and a Cancer Rehabilitation Manual [15–20]. We also adopted international healthcare quality standards to structure, procedures, and outcome measures (e.g., Functional Independence Measure (FIM), Six-Minute Walk Test, hand grip strength). Our Rehabilitation Service was accredited by the National Accreditation Organization (ONA), followed by the Joint Commission International (JCI) and the Commission on Accreditation of Rehabilitation Facilities (CARF). The Rehabilitation Service of the ICESP was the first one outside of the USA to receive a specific Accreditation for Cancer Rehabilitation Services, at the end of 2014.
Our Rehabilitation Service supports the patients being currently cared for at ICESP, that is, beginning at the initial cancer diagnosis, following active cancer treatment and active surveillance, and ending after cancer care is finished. Cancer survivorship is not usually performed at ICESP. Post-cancer transition of care and rehabilitation are provided by primary- and secondary-level institutions.
In our institution, the physiatrists act as a gatekeeper of outpatient rehabilitation care, managing all rehabilitation referrals using our electronic healthcare system. Currently, we receive about 50 weekly referrals to outpatient Physiatry/Rehabilitation. A Physiatrist of our team will then assess the health records of the referred patient and will schedule the eligible patients for an initial in-person visit with either a physiatrist or with our rehabilitation team. Most patients are directed to first undergo a comprehensive initial visit with a physiatrist, who will then determine the rehabilitation needs of the patient. However, there are specific situations in which patients will be assigned directly to a rehabilitation program (“fast track”) before an initial visit with a physiatrist occurs, for instance, post-operative care and range of motion limitation that restricts radiotherapy initiation, among others. These protocols aim to reduce the waiting time for time-sensitive conditions.
The outpatient rehabilitation program may take two forms in our institution: (a) a comprehensive, individualized rehabilitation program including one or all of the following: physical therapist, speech therapist, occupational therapist, neuropsychologist, and/or exercise physiologists; or (b) orientation groups—in which education and exercise orientation are provided to a small group of patients, in a few visits to the rehabilitation clinic, by one or more of the rehabilitation professionals, depending on the groups’ needs. There are specific orientation groups for the more common impairments, such as lymphedema prevention, lymphedema follow-up after discharge from the rehabilitation program, compliance to physical activity after discharge from the rehabilitation program, and chronic pain long-term care. Orientation groups are also an option for patients that cannot travel to ICESP. Weekly, the rehabilitation team and each physiatrist convene to assess the progress of selected patients, including their progress during therapies, and possible next steps that may be therapy adaptation, extension, or discharge.
Rehabilitation of hospitalized patients follows specific protocols that include at least one of the abovementioned rehabilitation specialists. In case the clinical/rehabilitation team following the inpatient identifies a more complex rehabilitation need, a physiatrist is consulted.
Rehabilitation Center—Structure and Outcomes
Our Rehabilitation Center is located at the ground floor of ICESP, in an area of 190 m2 (around 2000 ft2) (Fig. 1), comprising offices for the delivery of rehabilitation therapies, a gymnasium, and several resources such as treadmills, stationary bicycles, rowing machine, a walking track (for the Six-Minute Walk Test), parallel bars, therapeutic beds and platforms, orthostatic bed and table, spaces that simulate daily activities, and virtual reality resources. On the 17th floor, there is a smaller rehabilitation room (of around 30 m2 or 300 ft2) for inpatient rehabilitation activities.Fig. 1 Rehabilitation Center of the Cancer Institute of the State of São Paulo of the University of São Paulo School of Medicine
Our rehabilitation team provides around 8500 therapy sessions/month for inpatients and 2500 therapy sessions/month for outpatients. The current weekly hours of therapy provided by each rehabilitation professional are as follows: 3000 h of physical therapy (300 h for outpatient, 2700 h for inpatient), 240 h for speech and language pathology, 150 h for occupational therapy, 120 h for neuropsychologists, and 90 h for physical educational professional/kinesiologist. Physiatrist care is provided for 120 h/week.
Demographics, primary cancer site, and main reason for referral to rehabilitation therapies in 2021 are expressed in Table 1. Most common reasons were pain (28%) and range-of-motion limitation (20%). Although most patients presented with more than one rehabilitation need, those data relate to the primary reason. Criteria for referral to therapy have not changed over the years.Table 1 Characteristics of outpatients referred to rehabilitation therapies in our service in 2021
Prevalence
Primary cancer sites
Breast 48%
Head and neck 8%
Hematologic 7%
Gastrointestinal 6%
Urogynecological 6%
Nervous system 6%
Bone and soft tissue 5%
Thoracic 4%
Female 69%
Age
18–40 14%
41–65 62%
66–85 22%
> 85 2%
Main reason for referral
Pain 28%
ROM limitation 20%
Lymphedema 19%
Neurological deficits 12%
Post-COVID syndrome 11%
Cancer-related fatigue 8%
Post-surgical care 3%
ROM range of motion
Characteristics and outcomes of the rehabilitation program can be found in Table 2. Since our outpatient rehabilitation center was closed for several months during 2020 and 2021 due to the COVID-19 pandemic, data from 2019 was reported. The duration of the rehabilitation program is about 3 months. We have a high satisfaction rate (98%), and about three quarters of patients met our rehabilitation goals either partially or completely.Table 2 Outpatient rehabilitation data from our service in 2019
Average time from referral to 1st rehabilitation session 16.9 days
Duration of rehabilitation program 84.1 days
Percentage of missing rehabilitation sessions 19%
Patient satisfaction (NPS) 98%
Rehabilitation goals at the end of the program
Fulfilled (76–100%) 68%
Partially fulfilled (51–75%) 9%
Partially unfulfilled (26–50%) 8%
Unfulfilled (0–25%) 6%
N/A 9%
Reason for unmet goals
Nonattendance 41%
Clinical complications 36%
Patient requested to be discharged 11%
No improvement with therapies 6%
Non-compliant with orientations 3%
Deceased 3%
Social issues 2%
NPS net promoter score
Over the 14 years of the Rehabilitation Service, we observed a growing complexity of cancer patients treated in our Rehabilitation Center. Despite that, we were fortunate to have a history of no severe adverse events occurring with patients during rehabilitation care throughout these years. We documented the safety of rehabilitating patients with metastatic bone disease in a study published in 2021 [21]. In this retrospective study, we assessed outcomes of patients with bone metastasis who underwent rehabilitation therapy at ICESP and found only one pathological fracture during the rehabilitation period, which was unrelated to rehabilitation therapy. We also observed two other skeletal-related events, resulting in a total event rate of 11.8 per every 10,000 h of therapy.
Academic Activities
ICESP’s cancer rehabilitation service is part of the fourth-year medical students’ curricula. We are also a mandatory 1-month rotation for third-year PM&R residents of the University of São Paulo (10 per year), frequently receiving PM&R residents from other institutions [22]. A Cancer Rehabilitation fellowship is being organized and it will be the first one in the country.
ICESP has a multidisciplinary residency program in oncology for physical therapists, occupational therapists, nurses, and case managers.
Our Rehabilitation Service has also played an important role in international societies. We are participants of the Special Interest Group in Cancer Rehabilitation of the International Society of Physical and Rehabilitation Medicine (ISPRM) since its formation. We have collaborated in the promotion of the area and in the creation of scientific content since then [15–21, 23]. In 2022, a Cancer Rehabilitation Manual was published as a result of this collaboration with ISPRM [24].
Our main areas of research include pain management, impact of supervised exercise programs, breast cancer, and cognitive rehabilitation [15–21, 23, 25–28].
Rowing Boat Team
In 2013, we started a rowing program for breast cancer patients, called REMAMA, inspired by the international rowing movement for breast cancer patients that started in Canada (the International Breast Cancer Paddlers Commission—IBCPC), which was present in 12 countries at that time. It is now present in 32 countries, comprising 250 teams. Each team has 22 participants that practice in dragon boats. The international festivals occur every year and attract more than 4000 participants. We have two teams from REMAMA (both called REMAMA Dragão Rosa—Pink Dragon) over a total of 16 currently existing teams from Brazil [29].
Barriers and Facilitators
We experienced barriers to the provision of cancer rehabilitation that are common to those reported by other countries, including low awareness of the benefits of rehabilitation services among patients and providers, under-identification of rehabilitation needs, uncertainty around referral pathways, lack of infrastructure that streamlines the referral process, lack of funding, paucity of trained rehabilitation specialists in the area, need of greater inclusion of rehabilitation intervention in cancer care treatment guidelines, and an under-recognition of potential cost savings and reduced complications with the use of early rehabilitation programs [3, 5, 19].
One of our facilitators is that we are located in a large cancer center with strong interdisciplinary vision. We also observed a growing perception, both in the medical community and in other stakeholders, about the need for and importance of rehabilitation care for cancer patients.
Conclusion
In this report, we aimed to narrate our experience in structuring and maintaining a Cancer Rehabilitation facility in a middle-income county for the past 14 years. Our trajectory was facilitated by the support of the local government and by a national ordinance mandating the provision of rehabilitation services in specialized cancer centers. Our service provides interdisciplinary rehabilitation coordinated by physiatrists, and in contrast to our foundation where we had difficulty recruiting cancer rehabilitation specialists, we now have an important role in the formation of those professionals in Brazil.
Since the prevalence of cancer survivors will increase significantly over the next decades, it is crucial that rehabilitation services organize their resources to respond adequately to the growing need of rehabilitating cancer-associated impairments. Several organizations and governments have established resources and efforts to improve cancer rehabilitation worldwide, including several guidelines [1•, 30•]. We expect that our report assists in the implementation or improvement of cancer rehabilitation services.
Declarations
Conflict of Interest
The authors declare no competing interests.
Human and Animal Rights and Informed Consent
This article does not contain any studies with human or animal subjects performed by any of the authors.
This article is part of the Topical Collection on Cancer Rehabilitation
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
Papers of particular interest, published recently, have been highlighted as: • Of importance
1. • Smith SR, Zheng JY, Silver J, Haig AJ, Cheville A. Cancer rehabilitation as an essential component of quality care and survivorship from an international perspective. Disabil Rehabil. 2020;42(1):8–13. 10.1080/09638288.2018.1514662. Important review on the current state of cancer rehabilitation globally, identifying several international initiatives for cancer rehabilitation and survivorship. Authors also identified several barriers to the implementation of cancer rehabilitation services worldwide.
2. Cheville AL Beck LA Petersen TL Marks RS Gamble GL The detection and treatment of cancer-related functional problems in an outpatient setting Support Care Cancer 2009 17 1 61 67 10.1007/s00520-008-0461-x 18478275
3. Stout NL Silver JK Raj VS Rowland J Gerber L Cheville A Toward a national initiative in cancer rehabilitation: recommendations from a subject matter expert group Arch Phys Med Rehabil 2016 97 11 2006 2015 10.1016/j.apmr.2016.05.002 27237580
4. Stubblefield MD Hubbard G Cheville A Koch U Schmitz KH Dalton SO Current perspectives and emerging issues on cancer rehabilitation Cancer 2013 119 Suppl 11 2170 2178 10.1002/cncr.28059 23695929
5. Silver JK Baima J Mayer RS Impairment-driven cancer rehabilitation: an essential component of quality care and survivorship CA Cancer J Clin 2013 63 5 295 317 10.3322/caac.21186 23856764
6. Fukushima T Tsuji T Watanabe N Sakurai T St AM St KK Cancer rehabilitation provided by designated cancer hospitals in Japan: the current state of outpatient setting and coordination after discharge Prog Rehabil Med 2022 7 20220006 10.2490/prm.20220006 35274061
7. Cho S, Chung SH, Kang M, Jo A, Sim SH, Kim YJ, et al. Underutilisation of physical rehabilitation therapy by cancer patients in Korea: a population-based study of 958,928 Korean cancer patients. J Korean Med Sci. 2021;36(46):0.
8. Thorsen L Gjerset GM Loge JH Kiserud CE Skovlund E Fløtten T Cancer patients' needs for rehabilitation services Acta Oncol 2011 50 2 212 222 10.3109/0284186x.2010.531050 21231783
9. Brazil's National Supplementary Health Agency [Agência Nacional de Saúde Suplementar]: Situation room. https://www.ans.gov.br/images/stories/Materiais_para_pesquisa/Perfil_setor/sala-de-situacao.html (2022). Accessed September 28th, 2022 2022.
10. da Silva MJS O’Dwyer G Osorio-de-Castro CGS Cancer care in Brazil: structure and geographical distribution BMC Cancer 2019 19 1 987 10.1186/s12885-019-6190-3 31647005
11. Massuda A Hone T Leles F Castro M Atun R The Brazilian health system at crossroads: progress, crisis and resilience BMJ Glob Health 2018 3 e000829 10.1136/bmjgh-2018-000829
12. da Cunha MAO Santos HF de Carvalho SMEL Miranda GMD de Albuquerque MdSV de Oliveira RS Health care for people with disabilities in the unified health system in Brazil: a scoping review Int J Environ Res Public Health 2022 19 3 1472 10.3390/ijerph19031472 35162497
13. Saldanha R, Xavier D, Carnavalli K, Lerner K, Barcellos C (2019)Estudo de análise de rede do fluxo de pacientes de câncer de mama no Brasil entre 2014 e 2016. Cad saúde pública 35 10.1590/0102-311x00090918.
14. Brazil's Ministry of Health. Brazil's Ministry of Health, Ordinance number 741, of December 19th, 2005.
15. Almeida EMP AR, Cecatto RB, Brito CMM, Camargo FP de, Pinto CA, Yamaguti WP dos S, Imamura M, Battistella LR. Exercício em pacientes oncológicos: reabilitação. Acta Fisiatr. 2012;19(2):82–9.
16. Brito CMMLM Saul M Bazan M Otsubo PPS Imamura M Battistella LR Câncer de mama: reabilitação Acta Fisiatr 2012 19 2 66 72 10.5935/0104-7795.20120013
17. Cecatto R Almeida E Saul M Brito C Andrade R Imamura M Câncer de pulmão: reabilitação Acta Fisiatr 2013 20 63 67 10.5935/0104-7795.20130011
18. D'Alessandro E de Brito C Cecatto R Saul M Atta JA Lin CA Evaluation of acupuncture for cancer symptoms in a cancer institute in Brazil Acupunct Med 2013 31 1 23 26 10.1136/acupmed-2012-010206 23117345
19. de Brito C, Bazan M, Pinto C, Baia WRM, Battistella LR. . Manual de Reabilitação em Oncologia do ICESP. 1 ed. São Paulo: Manole; 2014.
20. Municelli L Cecatto R Brito C Battistella L Chemotherapy-induced peripheral neurotoxicity: approach to rehabilitation Crit Rev Phys Rehabil 2013 25 261 274 10.1615/CritRevPhysRehabilMed.2013010265
21. Tabacof L Delgado A Dewil S Reis F Velar CM Corteline MED Safety and feasibility of outpatient rehabilitation in patients with secondary bone cancer: a preliminary study Rehabilitation Oncology 2021 39 3 E42 E50 10.1097/01.reo.0000000000000241
22. Ferrão GC Cecatto RB de Brito CM Battistella LR Re: Cancer rehabilitation education during physical medicine and rehabilitation residency Am J Phys Med Rehabil 2015 94 5 e35 e36 10.1097/phm.0000000000000276 25802963
23. de Sousa M Bueno C Mendoza Lopez R de Almeida E Cecatto R de Brito C Postbreast cancer surgery outpatient rehabilitation program: analysis of clinical profile, impact, and direct medical costs J Int Soc Phys Rehabil Med 2019 2 1 22 29 10.4103/jisprm.jisprm_17_18
24. Brito CMM et al. Cancer rehabilitation manual. 1 ed. São Paulo: Manole; 2022.
25. D’Alessandro EG Nebuloni Nagy DR de Brito CMM Almeida EPM Battistella LR Cecatto RB Acupuncture for chemotherapy-induced peripheral neuropathy: a randomised controlled pilot study BMJ Support Palliat Care 2022 12 1 64 10.1136/bmjspcare-2018-001542
26. D'Alessandro EG da Silva AV Cecatto RB de Brito CMM Azevedo RS Lin CA Acupuncture for climacteric-like symptoms in breast cancer improves sleep, mental and emotional health: a randomized trial Med Acupunct 2022 34 1 58 65 10.1089/acu.2021.0073 35251438
27. de Almeida EPM de Almeida JP Landoni G Galas F Fukushima JT Fominskiy E Early mobilization programme improves functional capacity after major abdominal cancer surgery: a randomized controlled trial Br J Anaesth 2017 119 5 900 907 10.1093/bja/aex250 28981596
28. Bagatini O, Bertin C, Hong F, Guarita ML, Shinzato G, Imamura M, et al (2018) Uso da terapia por ondas de choque para o tratamento do linfedema associado ao câncer de mama. Acta Fisiatr. 25 10.11606/issn.2317-0190.v25i4a163839.
29. International Breast Cancer Paddlers Commission - IBCPC. https://www.ibcpc.com/ (2022). Accessed.
30. Stout NL Santa Mina D Lyons KD Robb K Silver JK A systematic review of rehabilitation and exercise recommendations in oncology guidelines CA: A Cancer J Clin 2021 71 2 149 75 10.3322/caac.21639
| 36466557 | PMC9703421 | NO-CC CODE | 2022-12-07 23:16:02 | no | Curr Phys Med Rehabil Rep. 2022 Nov 28; 10(4):339-344 | utf-8 | Curr Phys Med Rehabil Rep | 2,022 | 10.1007/s40141-022-00373-4 | oa_other |
==== Front
Contemp Jew
Contemp Jew
Contemporary Jewry
0147-1694
1876-5165
Springer Netherlands Dordrecht
9466
10.1007/s12397-022-09466-7
Original Research
“I’m Like a Chameleon”: Coping Strategies Used by Haredi Women Doctoral Students Reconciling Their Religious and Academic Identities
http://orcid.org/0000-0002-0791-1590
Binhas Adi [email protected]
grid.443013.1 0000 0004 0468 6046 Beit Berl College, Kfar Sabba, Israel
28 11 2022
118
27 3 2022
3 10 2022
© The Author(s), under exclusive licence to Springer Nature B.V. 2022, Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
This study examined Jewish ultra-Orthodox (Haredi) women doctoral students to analyze the shaping of their religious and academic identities, and particularly the coping strategies they use to reconcile them. It is informed by theories on the definition of social and collective identities and the way individuals assimilate upon encountering a new collective, as well as by actual processes of Haredi integration in Israeli academia over the years. The study concludes that in their academic development, these women challenge their traditional social worlds and enter the world of learning, which in their community is exclusively reserved for men.
Keywords
Haredi women
Higher education
Social identity
Collective
Intersectional identity
==== Body
pmcIntroduction
The king’s daughter is all glorious within (Psalms 45:13, KJV).
The motivation for this study came from my participation in Challenging Gender Inequality in Science and Research (CHANGE, n.d.)—a European Union Horizon 2020 project led in Israel by Dr. Hana Himi.1 My role in the project was as a researcher and change agent, studying barriers and facilitators in the integration of Jewish women doctors of Ethiopian descent in Israeli academia. As an immigration researcher, I was interested in the effects of their immigrant identity on their integration in a new sociocultural environment. After concluding this project, I sought to widen its scope and also study ultra-Orthodox (Haredi) Jewish women studying for higher education degrees. Although these women have not immigrated to Israel in the physical sense, they do experience cross-cultural migration upon their encounter with Israeli higher education. Whereas recent years have seen a growing number of Haredi higher education students, mainly in dedicated Haredi institutes of higher education, in 2021, Haredi doctoral students (both men and women) in general universities in Israel numbered less than 100 (Cahaner, 2020). Thus, while the discourse on diversity and multiculturalism in public systems, including higher education, is high on the Israeli agenda, and while the higher education system enables social mobility and integration, it remains out of reach for certain groups due to various barriers (Hendin 2011). This is perhaps more so for Haredi women owing to the intersection of their gender and religious identities.
Many studies have addressed the integration of ultra-Orthodox Jews (Haredim), and particularly women, in Israeli academia, but have hitherto focused mainly on BA and MA studies. In Haredi society, these studies are seen as a means of improving pay and job market prospects (Caplan 2007; Kalagy 2012; Layosh 2014; Rubin and Novis-Deutsch 2017; Tal and Yinon 1998). Doctoral studies do not necessarily serve these purposes and provide little visible gain for the community. Moreover, they involve a long-term career commitment that includes studying and teaching in non-Haredi institutes, critical thought and writing, and an encounter with a normative world completely different—if not diametrically opposed—to the core values of the Haredi community. Thus, the innovation in this study lies in addressing the population of Haredi women doctoral students and graduates, and particularly in analyzing the identity transformation they undergo as they maneuver between the two sociocultural groups to which they belong.
Theoretical Background
Social, Collective, and Intersectional Identity
Social identity refers to an individual’s belonging to certain groups based on national, religious, gender and other categories. Cognitively, identity helps people define themselves and the boundaries separating them from other identities, and also includes psychological traits, social norms, and attitudes (Ashforth and Meal 1989; Brewerand and Gardner 1996). It is also related to behavioral and emotional aspects, and characteristics shared with other group members (Brewer and Gardner 1996; Tajfel and Turner 2004). According to Weber (1988), social identity is influenced by the individual’s social, historical, and political environment, and is also related to culture conveyed by social institutions. These in turn create arrangements whereby individuals shape their identity within personal and social spaces and play roles as expected of them (Fearon and Laitin 2000).
Collective identity is derived from individuals’ relations with the collective in terms of lifestyles, norms, memories, beliefs, etc. A collective framework provides individuals with the tools to judge and evaluate their reality, and harmonize their identity (Cohen, 2006). Collective identity develops, and is characterized by, internal diversity, including contradictions with other identities adopted by the individual in some cases (Hall 1990; Sagy 2006; Shoval 2010). Studies on the identity construction of traditional groups in higher education have shown how the two worlds may be reconciled to produce a stable identity (Erikson 1968). Nevertheless, the encounter with different identities may also be conflictual (Hole 1990).
Intersectional theory views the individual as being made up of multiple characteristics and belonging to multiple groups, and examines the interrelatedness of various identity elements. According to the intersectionality approach, an individual’s belonging to a gender and an ethnic and class group, for example, represents social and institutional systems that shape these identities. Identity elements are intertwined without the ability of distinguishing clearly between them, forming the main bases for social differentiation and stratification. For example, a study on women from non-dominant ethnic groups in higher education showed that they have complex identity characteristics, and that they create a kind of multifocal lens with which to perceive the world (Andersen 2005; Andersen and Collins 2007).
In the Israeli context, studies from the intersectionality approach have addressed identity tensions and contradictions experienced by Mizrahi Jewish professors (of Asian and African origin) (Shohat 2001) and clashes between ethnic and national identity (Yonah and Shenhav 2005). Others have focused on class (Gutwein, 2001) and gender (Toran 2009). All agree, however, that there is no single identity factor that defines the individual, but that the individual is made up of a combination of identities that do not always harmonize. The present study examines the combination of two major identity elements: the collective Haredi identity, associated with a devout, conservative, and segregative society, and the collective secular, rational, and critical academic identity.
Berry’s Model of Acculturation
The literature classifies different ways in which minority groups such as migrants deal with the majority group. One of the best-known typologies divides them into several prototypical models of acculturation. According to Berry (1997), the first model is assimilation (the melting pot). This model tends to cancel out or minimize all differences between groups and emphasize their common denominators. In this model, individuals are called upon to give up unique characteristics of their culture in favor of adopting the core values of a new one. When individuals such as migrants assimilate, they do not seek to retain their cultural identity; they prefer a new culture, adopting it wholeheartedly by acquiring the language, socializing with the locals, and adopting their lifestyle, and are often encouraged to do so by formal institutions. This model has dominated the acculturation of migrants in the USA (Matton and Maurine 1992). In the separation model, which characterizes most of Haredi society in Israel, it is important for the individuals to retain their culture of origin, and they seek to minimize the interaction with and influence of the new culture. In the integration model, which characterizes some sections in Haredi society, individuals are interested both in retaining their culture of origin and in interacting with members of other groups. Finally, marginalization occurs when both the heritage and the receiving cultures are rejected (Berry 1997). In the case of the population under study, the Haredi women are clearly required to integrate two different worlds. This study examines how they do it.
The Setting: Haredi Women and Academia in Israel
Haredi Society—General Background
Haredi society is unique in its culture and lifestyles, both in Israel and worldwide. It is growing rapidly—at a rate of about 4% annually—thanks to high fertility rates. In 2020, it represented some 6% of Israel’s population, and is expected to reach 16% by 2030 (Cahaner and Malach 2020). The Haredi population is among Israel’s poorest, although the recent decade has seen some increase in its labor market participation and income (Agmon 2019). The great majority of Haredim vote for sectorial parties who see to the budgeting of religious and educational institutes and maintaining the segregation of Haredi society from the rest of Israel (Malach 2019). Thanks to consistently high voter turnout and the nature of the Israeli parliamentary system, Haredi parties enjoy disproportional political power, contributing to the difficulty of promoting reforms in the Haredi community, including in the academization area.
The Academization of Haredi Society
The encounter of Haredi Judaism with academia has a long history. As early as nineteenth-century Europe, the encounter with secular (in those days, non-Jewish) academia challenged traditional Jews who were devoted exclusively to religious learning. The university environment was considered by many as taboo, as it would expose the Jewish student to promiscuity at best and heresy at worst and create the potential for intermarriage. Nevertheless, many Jewish women gradually started to work outside the home and encounter secular, liberal ideas. In 1917, an educational movement for girls called “Beit Yaakov” was founded, which became an extensive school network throughout Eastern Europe.
Upon the establishment of the State of Israel in 1948, Haredi schools’ autonomy was recognized by the government, and state support was provided. Upon graduating, boys would continue with their religious studies in the yeshiva, and girls would go on to study in teacher seminars, formal quasi-academic institutes. Today, Haredi men remain exclusively devoted to religious studies, whereas women study in programs mainly designed to promote their labor market integration and improve their pay as the main breadwinners in the Haredi family. They study mainly education, social work, administration, psychology, and law (Rubin and Novis-Deutsch 2017). In doing so, they inevitably experience a clash of identities (Tal and Yinon 1998). In that respect, it experiences a constant tension between the desire to preserve its traditions and lifestyle and accept certain aspects of modernity, with various sections within it holding different views and approaches toward that conflict (Layosh 2014).
Haredi Integration in Higher Education: Encountering Modern Values
The main characteristics of Haredi Judaism are neo-traditionalism, commitment to religious learning, strict compliance with religious commandments, and obedience to rabbinical authorities and, among certain currents, anti-Zionism (Friedman 1991). In recent years, a certain Haredi sector has grown in prominence. Called “modern Haredim,” they do not have distinct institutions or leadership, but are characterized mainly by a more modern lifestyle in terms of employment, higher education, residential habits, and leisure culture. Nevertheless, for them, and obviously even more so for more traditionalist groups in the community, higher education and doctoral studies in particular represent an intensive clash of two worlds (Cahaner 2020).
In Israel, there are 9 research and teaching universities, and 49 academic colleges engaged in teaching and theoretical and clinical research. All are subject to academic supervision by the Council for Higher Education (n.d.-a). In 2018–19, some 12,000 Haredim studied in all colleges and universities in Israel, representing 8.3% of the student population. Most (84%) of them studied for a bachelor’s degree, of whom 70% were women. Out of the remainder, only 90 studied for a doctorate. Integrating the Haredim in higher education is a key social interest. It is challenging, however, owing to the community’s fear that the encounter with academia may affect its unique cultural identity, among other things. Moreover, given the current nature of Haredi high schools, graduates find it relatively difficult to be admitted to colleges and universities. The Haredi demand for gender segregation also poses difficulties for institutes of higher education. In recent decades, the Council for Higher Education (n.d.-b) has supported 14 dedicated academic settings for Haredim of both genders, which are sensitive to their cultural needs.
Most of the Haredi students studied in academic colleges (55%) and in teacher colleges (28%), and only 17% studied in universities, as opposed to the general student population, of whom a third studied in universities. The age of male Haredi students is higher than average, mainly owing to the fact that they require a prolonged preparatory process in dedicated colleges because of the unique characteristics of their separate education system with its heavy stress on religious studies. Female Haredi students, on the other hand, are younger than non-Haredi students during BA studies, with 45% being younger than 21 years of age. The reason for that is their exemption from military services as well as the tendency to study in academic institutes instead of seminars for girls. In terms of disciplines, more Haredi students study education and teaching (35%) when compared with the entire student population (8%), as well as paramedical professions (12% as opposed to 6%). Few study social sciences, engineering, and the humanities. Somewhat surprisingly, relatively many study natural sciences (11% compared with 12% in the general population (Cahaner and Malach 2019, 31–33).
In early 2022, about half of Haredi men were employed (according to their statements), a rate lower than the target of a government resolution of August 2021, 65%. Conversely, about 80% of Haredi women are employed, in accordance with government targets. The gaps between Haredi and non-Haredi men and women are particularly high in the leading high-tech industry (3% and 14% percent of Haredi and non-Haredi men, as opposed to 5% and 7% percent of Haredi and non-Haredi women, respectively) (Cahaner and Malach 2021; Peleg-Gabbay 2022).
Recent years have seen a growing phenomenon of “new” Haredim—a group engaged in political activism, with strong presence in digital media and academia, without abandoning their commitment to the community. They are active in both conservative and liberal groups and are increasingly open to integration in general society on their own terms (Haredi Research Group 2022).
Over the years, in order to encourage Haredi higher education, the state provided support for dedicated Haredi tracks and programs in existing colleges and universities. These programs are designed to help Haredi students reduce their knowledge, literacy, and academic skill gaps created because of the unique characteristics of their education system (Ehrenfeld, 2017; Horowitz 2016).
Haredi Women’s Integration in Higher Education: Intersectionality and Coping
Haredi women belong to two minority groups: Haredi society as a cultural minority within Jewish society, and women in Israel. Their intersectional status as a minority within a minority means that they are subject to double marginalization (Neria-Ben Shahar 2008). In the spirit of the psalm quoted above, the “gloriousness within” of the Haredi woman is emphasized in multiple ways: in modesty lessons in school, in women-only religious lessons, and in texts in the Haredi press. Commanded to maintain an extremely strict standard of modesty, upon marriage they are tasked with both raising the children and maintaining the household and providing for the family, as the men are (ideally) devoted to religious studies. Thus, they have to support families with many children—68% percent of Haredi women work outside their home, compared with 79% of all Jewish women (Central Bureau of Statistics 2015).
In the past, most Haredi women worked in education. Teaching has a triple advantage for them. First, it is a “natural” ideological continuation of their role as child carers. Second, in performing that role they can be easily supervised, both by the school management and by the parents. Third, the working hours and vacation schedule are suitable for mothers and wives (Friedman 1999). The growing tendency of Haredi women to integrate in the general labor market is part of a broader phenomenon, which includes academic as well as political leadership (Miletzky 2017). According to Schwartz (2008), it was the saturation in the Haredi teaching market that led to the new trend, with relatively fewer Haredi women now being employed in education. Many postsecondary institutions for Haredi young women currently provide technological training in addition to religious studies. Since 2006, several institutes devoted to promoting Haredi employment in the general job market have been opened with government support. To conclude, a quiet revolution is underway. Women’s status in the Haredi family has improved, since they are the ones who hold instrumental knowledge. This is also evident in their growing academic ambitions.
Following Elor (1992), who titled her book Educated and ignorant: From the world of ultra-Orthodox Women, Haredi women’s higher education studies have been examined as a way of searching for and shaping a new identity. She argued that Haredi women had an essential role to play in maintaining the Haredi male learners’ society. Therefore, even when they became educated, their traditional status in their segregated society remained identical, and they did not act as change agents but rather as conservation agents. As a result, the conservative elements remained in place. Higher education was there, but it lacked the key aspects of enlightenment: equality, critical thought, modernity, and rationality. These values are inherently contradictory to those of the community. Accordingly, the academization process has been mainly “used” to serve instrumental needs, without addressing essential questions regarding academia and its social role. This study addresses this lacuna by focusing on advanced studies where devotion to research rather than a salary raise is the main career motivation.
Elor’s claim has been criticized, with others claiming that Haredi women’s higher education integration represents a turning point in Haredi society’s attitude toward general education and professional training (Weissman 1995). Among these women, academic studies raise questions regarding their priorities as wives and mothers as opposed to values such as self-realization, career promotion, dedication to work, and professional satisfaction—all associated with secular modern society—and hence many feel ambivalent (Layosh 2014).
A major study of 469 Haredi women studying for their bachelor’s and master’s degrees found that the economic motive was dominant in their choice of pursuing higher education, with ideological motives playing a lesser role. Indeed, most chose to study in religious institutions. One of the main fears they reported was the fear of spiritual religious “decline” (Rubin and Novis-Deutsch 2017).
The Current Study
A growing literature has recently been analyzing these trends of greater participation in the job market and higher education, and trailblazing women in media, politics, and society. Nevertheless, studies on Haredi women studying for or with a doctorate have not been conducted, despite the fact that this is an important and growing group as part of these trends.
This study examines Haredi women who have chosen to pursue doctoral studies. Such studies are not available in Haredi colleges, and these women have therefore entered public universities where the culture and atmosphere are modern and secular, there is no gender segregation, and critical thinking is valued. The study examines their experience in this intercultural encounter and their strategies for coping with the tensions it involves. The research has been approved by our institutional ethical board.
Method
Participants and Interviews
Eight Haredi women studying for or with a doctorate were interviewed for this study. All studied in Haredi institutions until the age of 18, and all are mothers (of between one and five children). Three of them are divorced. All studied in Israeli universities; two completed part of their studies in the USA. In all cases, their research areas were related to Haredi society.
Recruiting the interviewees was challenging since, beyond the overall cross-cultural gap, the interviewees have already become integrated in the Israeli academic world, and were not necessarily interested in describing their personal experiences. Initially, the interviewees were selected on the basis of personal acquaintance (convenience sampling), and subsequently using the snowball technique (Cohen and Arieli, 2011). During the interviews, however, we realized that we must also contact specific actors capable of shedding broader light on the issue, including the way academic institutions were preparing to assimilate Haredi students, and we therefore also interviewed a Haredi affairs advisor of a public Israeli university.
Semi-structured in-depth interviews of 1–2 hours were conducted online (by Zoom, during COVID-19) in 2020. The subject of our research was explained to the interviewees and their informed consent was obtained. We also assured their anonymity. The participants were asked to describe their academic experience, the barriers facing them, and the facilitators that helped them cope and succeed. They were also asked what, in their opinion, could help other Haredi women integrate in higher education in the future.
As data on the participants’ characteristics may enable readers to identify them, we have avoided specifics such as their belonging to Haredi subgroups or background in the USA. Most did not mention the specific Haredi community they belonged to and were not asked to. For the purpose of this study, the important thing was to include women educated in Haredi institutions prior to higher education and to exclude women born to non-religious families, and all have met these categories. Note further that whereas it is commonly assumed that the Haredi community in the USA provides more non-religious studies and therefore facilitates greater integration in general society, a study that compared Haredi education in the USA and Israel failed to show any significant differences, at least between Lithuanian–Haredi boys in the two countries (Malach and Ettinger 2021).
Interview Analysis
The interview transcripts were analyzed in the qualitative phenomenological paradigm. Accordingly, we gathered information about the participants’ personal experiences and sought to understand their meaning for them (Creswell 2007). This information was analyzed on the basis of Moustakas’ (1994) approach. In the first reading, we identified “significant statements” that could improve our understanding of the participants’ perspectives. These were grouped into clusters of meaning or themes, for which representative quotes were provided. The themes centered on the Haredi and academic identity experiences and their interphase. For example, statements such as “In the academic world, I realize I had to arrive in full armor” or “It’s hard for me to hear criticism of the society I come from in the university” were included under the theme “The Clash of Haredi and Academic Identities” (Ayalon and Sabar Ben-Yehoshua 2010; Creswell 2007).
Results
“I Was An Outsider among Them”: Haredi Identity
Family and Community Response
In characterizing their Haredi identity, the participants described its complexity and considerable inner diversity, sometimes even within their own families. They described different identities and alienation between different groups in Haredi societies, and their encounters with them. For example, one participant described how surprised she was upon meeting young women studying in a Haredi campus:I entered a Haredi campus because it was my only way to study […]. I experienced a culture shock upon encountering the Haredi women […] completely in tune with modernity […]. I felt we had no common language, that I was an outsider among them.
Diversity is also evident in their own family background: “My grandfather was a member of [a] Hassidic court, the other grandfather was a […] rabbi […]. My parents met in a co-ed youth movement […] there were encounters with the external culture.” This diversity notwithstanding, the difficulty of realizing academic aspirations within the closed Haredi society was there. A Haredi entrepreneur identified that difficulty and created a dedicated academic setting for Haredi women, currently in the process of expansion to include doctoral studies as well. She said: “Haredi women find it difficult to acquire higher education. I entered this vacuum and paved them the way to develop and actualize themselves academically, without compromising on their faith and identity” (Goldfinger, 2019). Havi Ernfeld—the “Mother of Haredi women students”—tailored a solution for Haredi women students while maintaining a high academic level.
Ernfeld’s activity is important in overcoming barriers to women’s higher education in the Haredi world, since their entry into colleges and universities involves clashes with the nuclear and extended family and the Haredi community as a whole. The participants described this clash in terms of intercultural gaps they needed to negotiate. One way of doing so was to highlight and conceal certain personal details, according to the circumstances of the interaction. Keren (39), for example, confided: “In my children’s PTA meeting, I would conceal the fact that I had a PhD. For them this means I have crossed over the line, it’s a secular word in my community.”
Given their understanding of the way (women’s) higher education is perceived in Haredi society, the participants defined their identity as straddling the two worlds, with the values of each being known to them. As also described in the literature (e.g., Friedman 1999; Robin and Novis 2017), three participants highlighted the perception of higher education in Haredi society as purely instrumental:Education is a way to improve one’s income […] first as well as second degree. A third degree is already for those who seek knowledge, personal enrichment, and is seen as a waste of time.
The Haredi perception is that the purpose of academic studies is to improve one’s job market status.
The first question I’m always asked is what my doctoral degree adds to my salary.
Some participants described their families as supporting their choice, and some described opposition by their husbands and parents, to the point of a complete change in the entire family’s attitude toward them. “My parents encourage education, but they are bound by the dictates of Haredi society”; “My parents used to be against it, but today they’re proud.”
As mentioned, all participants have children. Among the married ones, the husbands are devoted to religious studies, leaving them as the main breadwinners. The divorced participants, on the other hand, describe how this marital status made it easier for them to enter higher education: “Having divorced my husband, I have crossed a red line in Haredi society anyway, and then it is easier to cross other lines, such as getting a PhD”; “As a single mother, I’m less committed to [community] dictates, since I’m already out of the loop.”
Despite these difficulties, all participants expressed appreciation of the changes and growing openness of Haredi society, and enduring attachment to and appreciation of it. Sometimes, the encounter with anti-Haredi criticism in academia also made their position clearer, deepening their pride and faith in the values of Haredi society. Three participants expressed this sentiment:I’m not looking to change my way of life. I speak with the Almighty every day. I thank Him for what I have and feel good and comfortable with teaching in a Haredi academic institutes in everyday life.
The Haredim are a willingly segregated minority—nobody forced me into being Haredi.
I believe in the Haredi community—it’s a community that benefits its members.
In a press report titled “Dreams Are Meant to Come True,” the aforementioned entrepreneur describes her activism as promoting the integration of Haredi women into academia while maintaining a Haredi lifestyle. One of the women she helped said: “I have a family, praise be to God, I’m working full time and I’ve made a dream come true. It wasn’t easy, but new horizons have opened up for me.” Another participant in the project said: “I thought this was out of my reach and I was preoccupied with multiple obligations, but I managed to integrate [all] that with my studies” (Ehrenfeld, 2017). The availability of greater opportunities to begin higher education is reported in a positive tone in the community press, suggesting early signs of Haredi recognition of the ability to combine the two identities in the focus of this study.
Haredi Identity in the Clash with the Academic World
As they progress in their studies, Haredi women often experience a tension between the two identities. They are torn between their aspiration to advance in the academic world, which requires adopting a different set of values, primarily critical thinking, posing a potential threat to their religious faith, and community conventions that include strict obedience to rabbinical authority. According to one participant, “The perception of Haredi society is that academia would cause a weakening of faith […].”
The participants continued coping with the challenge of critical thinking as lecturers as well, a role in which they encounter the Haredi fear of higher education in their contact with students: “I explain to the students, the research does not seek to change your worldviews”; “One of the students told me that I had shattered her innocence and that she was sorry that she studied with me.”
Nevertheless, as mentioned, despite now being trained in critical thought, the participants also maintained their commitment to their collective identity: “I have sharp criticism on the party I’m voting for, but there’s no way I’m not voting for it.” Relatedly, they are well aware of the threats to Haredi identity in the encounter with academia, and some are also opposed to wholesale integration of Haredim in higher education: “It’s not good for everyone. […] My daughter got 700 in her psychometric test [at the bottom of the 94th percentile] but she won’t go to the university.” They are also keenly aware of the Haredim’s criticism of academic studies by Haredi educational institutions out of economic considerations. After decades of development, the growing integration in secular higher education competes with the Haredi institutions: “If women will go study in the secular academic institutes, this will hurt the Haredi institutes”; “A Haredi campus is good for my community, because it does not channel the learners to an academic career.”
“God Brought Me and My Supervisors Together”: Academic Identity
The participants have gone a long way in defining their identity since entering higher education, acquiring academic tools as well as codes of behavior and conduct. They also experienced the gaps between them and other students in terms of academic skills, and the excitement of the initial encounter with the secular campus and the fears it involved: “I would walk by the university and it would seem like a pipe dream to me”; “I was afraid I would be exposed to heresy”; “In academia, you can always be young.”
Exciting as it was for them, the interviewees were also critical of academia’s attitudes toward Haredi society out of a feeling that campus life involved an attempt to impose the values of a different culture on them, rather than letting their unique collective find its proper place: “It’s true I had no academic orientation […]. But academia is also unable to speak other languages. The academic system is blind to how much it is uncritical of itself.” In a similar vein, another participant said: “Academia is not truly open to diversity. It’s clear to me that I’m a token Haredi woman for the university; the university is also a token university for me.” Often, they felt criticized themselves: “They invited me to a forum in the Council for Higher Education, to discuss Haredi education, and asked me whether my horizons were opened ever since I began studying with men. I find that patronizing!”
However, on the level of interpersonal encounters with supervisors and colleagues, the participants described both academic and personal empowerment, sometimes even to the point of close friendships and interpersonal openness. All described encounters with researchers who have challenged them intellectually, and provided them with meaningful emotional support for years:“My supervisors […] helped me all the way, and gave me not only academic support, but also friendship”; “My post is a miracle—God brought me and my supervisors together.”
“I Don’t Meet the Standards of Either World”: The Identity Conflict
The “Meeting” with Academia
The encounter with academia prompted the women to adopt a complex identity structure, reconciling the contradictions between their Haredi and academic identities. It was not always easy for them to explain the academic track of years-long study and research. The curiosity and ambition required to pursue it were incompatible with the expectations of Haredi women. The participants were well aware of this complexity:I don’t meet the standards of either world, but I’m in the core of both. My life is Haredi, but I have different voices within me. With my academic friends I speak differently about the same subjects […] I have to choose what and how I say in each place, and I must not fail. I move across the identities.
The community also treats their choices as unusual: “I’m on the fence, I love both sides. My mother says that because I don’t have a husband I fell in love with my studies.” Sometimes, the encounter with academia also made them better appreciate the complexity of their Haredi identity:Until I got to higher education, I was afraid of heresy, of sitting next to secular people, but I never came across things that contradicted Jewish Law. […] Following my studies, there are some things in my world that I disagree with, but this does not affect my Jewish and religious identity.
The Haredi community’s attitude to higher education is ambivalent, combining elements of appreciation on the one hand and aversion and rejection on the other:The encounter with academia and education is a thorn in the Haredi world’s side, since although the intellectual sphere is not alien to us, academia is perceived as secularization […].
The Haredim admire education, and therefore it compliments my family that I’m a doctoral student, but they’re also attendant to the imperatives of society, that fears the influence of academia.
The Conflict with the Community and Family
The choice of higher education also affected the participants’ relations with the community. Despite the obvious gaps, the participants responded to the conflict by creatively forming an integrative identity seeking to contain both identities without compromising on either, in a way that enables a multifocal view of themselves (Andersen 2005; Andersen and Collins 2007). Despite diverging from the conventional life course of a Haredi woman, they are not critical nor contentious, but rather tolerantly accept the community’s attitude: “When I say I’m a doctoral student, they immediately check the length of my skirt”; “When they ask my son what is more shameful, having a divorcee or a lawyer and doctoral student for a mother, he answers that one covers up for the other.”
At the same time, the participants also described having had to deal with years-long opposition to their studies among their family and community. One of them described a difficult relationship with her husband, who was opposed to her studies, but said she did not give up: “I told him he could leave […]. When I’m in academia I feel alive, I couldn’t give it up.”
Discussion
This study examined how Haredi women students in higher education shape their academic identity alongside their religious identity and the strategies they use to reconcile the clash between the two. These action strategies were indicative of their interpretation of and attitudes toward the reality of the encounter between the identities and facilitated their negotiation of their two different identities. The findings indicate that the two identities are contradictory in some senses, and complementary in others. Four coping strategies arose from the interview analysis: setting boundaries between the identities; creating an identity with a dual commitment and responsibility; living in a sense of alienation; and identity cohesion. The four strategies are not mutually exclusive, and the participants have used several concurrently and according to circumstances.
(1) Setting boundaries between the identities. Owing to the conflict between the identities, the participants set clear boundaries between their social worlds and criticize each only mildly. They see to it that there is no excessive intermeshing of the two worlds and maintain the external characteristics, conversation topics, and social codes typical of each (Ashforth and Meal 1989). Since occasional contradictions are bound to arise, they try to find solutions that would be compatible with both identities.
For example, one participant wanted to investigate issues related to the Haredi world. On second thought, she decided not to undertake a study that could involve her with criticizing her own society. All participants expressed the aspiration to shatter each world’s stigmas on the other. Conversely, one participant described how she enjoyed teaching in a Haredi college: “I feel like in my college I’m in the academic world, but I’m also coming back home.” Thus, the participants experience boundaries between their identities, but at the same time feel they belong to each (Berry, 1997).
(2) Creating an identity with a double commitment and responsibility. The participants adopting this strategy feel proud of their identities and are connected to both. In some cases, they underplay the “other” identity, but when they have a chance to talk about it freely, they feel that they represent the Haredi world in the academic world and vice versa. They also feel committed to opinion leaders in both worlds. For example, they consult with their rabbi regarding their personal decisions, while at the same time consider their doctoral supervisor a professional mentor and source of support, perhaps more than other students do. Moreover, given their “migrant” status (a source of motivation) and preselection (assuming that only the most talented would face the cross-cultural challenge and study for a doctorate), their success rates are higher than average.
Both types of authority are highly appreciated and even revered by the participants. Moreover, their academic experience has made the participants more open-minded about secular society, and they have formed friendships with secular students, which probably would not have been formed otherwise. On the whole, theirs is an identity of commitment combined with criticism—the ability to be committed to and also critical of both worlds enables the students to recognize their two identities (Fearon and Laitin 2000; Toran 2009).
(3) Living in a sense of alienation. All participants referred to their sense of being different within each of the two communities. One of them said, “I’m like a chameleon.” When meeting their Haredi friends, who are all preoccupied with their family and children and far removed from academic pursuits, they feel alien. The same sense of alienation is experienced when interacting with fellow students, many of whom come from academic homes with a higher standard of living and different lifestyle, different family and intimate relationships, and different conversation topics. Thus, in both worlds they constantly feel only partly at home, as there is always some aspect of their identity that is incompatible with their social environment. The resulting sense of alienation becomes inherent to them, and is used as a tool for coping with each world.
(4) Identity cohesion. Approaching their academic career as congruent with Haredi society’s emphasis on the importance of lifelong learning highlights the affinity between the two worlds and identities (although this involves ignoring the fact that the Haredi learner society is an exclusively male society). Inquisitiveness, abstract thinking, philosophical reflection and the constant thirst for knowledge are familiar to learners from the Haredi world. In both identities, there is thus the passion for greater knowledge, involving persistence, commitment, and excellence, which is highly appreciated and rewarded in both worlds. In that sense, the participant repurposed tools and values adopted from their Haredi education.
By definition, Haredi society seeks to remain separate from general society in all areas of life, to obey rabbinical authorities, to sustain independent institutions, and to rely on the collective and the family for strength (Stern et al. 2022). Nevertheless, Haredim are also interesting in opening up to secular society (Haredi Research Group 2022). This conflict challenges both Haredi and secular society, which needs to integrate within it people who are similar to it in some senses, while maintaining highly traditional ways of life.
Conclusions
Collective identity often contains diverse and even contradictory identity aspects (Hall, 1990; Mosovich and Liberman 2018; Sagy 2006; Shoval 2010). The findings show that the participants’ identity does contain contradictions, but that they have found ways to reconcile them. The participants are ambitious and opinionated women who have sought to develop personally and intellectually through higher education studies. Nevertheless, this is culturally incompatible in Haredi society, and to cope with that discrepancy, they have adopted an alternative narrative, which, although diverging from what is expected of them as Haredi women and which may be even seen as subversive, they employ in a way that also enables them to go on living with their Haredi identity as well.
These women pose a challenge to Haredi society, since adult learning is preserved for men and for religious studies in it. The findings suggest that they have mastered an alternative type of learning, enabling them to develop themselves intellectually and pursue an academic career without breaking the boundaries of their traditional world. The result is a reconciliation of identities, while at the same time creating what may be seen as a radical new world of women who excel in secular studies just as the men excel in religious ones. Anchored strongly in both worlds, the women manage themselves in Haredi society and secular academia with a dual identity that is nevertheless fully coherent. This requires flexibility and constant negotiation, while rewarding the participants with a richer, fascinating, and harmonious life.
The gaps between the Haredi and secular worlds are huge—some would say unbridgeable—but recent decades have seen trends of growing Haredi participation in the job market, and academic studies provide growing opportunities for encounters between the two groups in various contexts. These encounters force the Haredim to find solutions for moving across two different cultural worlds. The State of Israel is interested in such integration, and the interest seems to be mutual. More broadly, the encounter between the communities may reduce prejudices and stereotypes held by the secular community about the Haredi community, and vice versa (Jobani and Peretz 2014).
The present study provides a persuasive example for integration by Haredi women without abandoning their core religious values. They have found a way to open up to the modern, secular academic world, which involves quite a few challenges for them, not only in terms of the curriculum, but also in terms of the need to use the Internet (which is taboo, at least in certain Haredi sectors), traveling to conferences, and international research collaborations. To do so, they have made use of the four unique strategies described above. Their coping processes should be seen as part of broader trends in Haredi society, which, while producing new challenges, suggest that it is possible for Israeli Haredim to retain their identity while integrating in general society.
Limitations and Future Directions
First, relatively few participants were interviewed, even though their number is high relative to the small population of around 100 students. Second, despite the fact that the interview was anonymous, the researcher, who is not Haredi herself, may have failed to detect some of the nuances of the participants’ complex identities. In a future study, it would be advisable to include a Haredi co-researcher or assistant. Third, some of the participants are currently in the process of developing their academic career, and may not have a broad enough perspective to identify the effects of their advanced studies on their identities. It is therefore recommended to follow up on this study by interviewing them again in several years’ time.
In a future study, I recommend examining the prices paid by Haredi women students owing to the constant need to reconcile their twin identities, including loneliness. On a more positive note, it would be interesting to examine the inspiration they provide to their children as women straddling the two worlds. Finally, the study focuses on one side of the intercultural encounter. It is recommended to also interview non-Haredi students who have studied with the participants, as well as lecturers, particularly their supervisors.
Acknowledgements
I thank Ami Asher for his meticulous editing and valuable comments on this article that have contributed significantly to its improvement.
Funding
The research was carried out in the research lab of the Center for the Advancement of a Shared Society at the Beit Berl College (Kfar Saba, Israel), with the funding of the Feldman Foundation.
Declarations
Conflict of interest
The authors have not disclosed any competing interests.
1 This project received funding from the European Union's Horizon 2020 research and innovation program under grant agreement # 787177.
The research was carried out in the research lab of the Center for the Advancement of a Shared Society at Beit Berl College with the funding of the Feldman Foundation TX.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
Agmon, Tamir. 2019. Demographic and economic trends in the Haredi population, 2009–2019. Knesset Center of Research and Information (Hebrew).
Andersen Margaret Thinking about women: a quarter century's view Gender and Society 2005 19 437 455 10.1177/0891243205276756
Andersen, Margaret and Patricia H. Collins. 2007. Race, class and gender: an anthology (6th ed.) Wadsworth.
Ashforth BA Meal FA Social identity theory and organization The Academy of Management Review 1989 14 1 20 39 10.2307/258189
Ayalon, Yael, and Naama Sabar Ben-Yehoshua. 2010. The process of content analysis according to grounded theory. In L. Kacen and M. Krumer-Nevo (Eds.), Qualitative data analysis (pp. 359–382). Ben-Gurion University Press (Hebrew).
Ben Shahar Neriya, Rivka. 2008. Haredi women and mass media in Israel: exposure patterns and reading modes (Unpublished doctoral dissertation). Hebrew University of Jerusalem (Hebrew).
Berry JW Immigration, acculturation and adaption Applied Psychology: An International Review 1997 46 1 5 68
Braun-Lewensohn, Orna and Tehila Kalagy. 2019. Between the inside and the outside world: Coping of ultra-Orthodox individuals with their work environment after academic studies. Community Mental Health Journal 55: 894–905. 10.1007/s10597-019-00392-x.
Brewer M Gardner M Who is thus ‘we’? Levels of collective identity and self-representations Journal of Personality and Social Psychology 1996 71 2 83 93 10.1037/0022-3514.71.1.83
Cahaner, L. 2020. Haredi society along the axis from conservativism and modernity. Center for Nation, Religion, and State (Hebrew).
Cahaner, Lee and Gilad Malach. 2019. Statistical report on ultra-Orthodox (Haredi) society in Israel, 2018: Abstract. The Israel Democracy Institute. https://en.idi.org.il/media/12041/the-yearbook-of-ultra-orthodox-society-in-israel-2018.pdf
Cahaner, L. and Malach, G. 2021. The yearbook of ultra-Orthodox society in Israel, 2021. Israel Democracy Institute (Hebrew). https://www.idi.org.il/media/17298/statistical-report-on-ultra-orthodox-society-in-israel-2021-new.pdf
Caplan, Kimmy. 2007. Internal popular discourse in Israeli Haredi society. Zalman Shazar Center for Jewish History (Hebrew).
Central Bureau of Statistics (2015) – Jerusalem, Israel. https://www.cbs.gov.il/he/pages
Central Bureau of Statistics. 2018. Press release: Religion and self-definition of extent of religiosity—Selected data from the Society in Israel Report No. 10. (Hebrew). https://www.cbs.gov.il/he/mediarelease/DocLib/2018/195/32_18_195b.pdf
Change. 2020. The project. https://www.change-h2020.eu/the_project.php
Cohen, Asher. 2006. Jews or non Jews; Jewish and Israeli Identity, The Challenge of Spreading Jewish Nationality), Bar Ilan University and Ketter, Jerusalem.
Council for Higher Education. n.d.-b. Increasing access to higher education for ultra-Orthodox Jews. https://che.org.il/en/increasing-access-higher-education-ultra-orthodox-jews/
Council for Higher Education (n.d.-b). Programs for the Haredi population (Hebrew).
Creswell, John W. 2007. Qualitative inquiry and research design: Choosing among five traditions (2nd ed.). Sage.
Ehrenfeld, Havi. 2017. Dreams are meant to come true. Channel 7 (Hebrew). https://www.inn.co.il/news/358696
Elor, Tamar. 1992. Educated and ignorant: from the world of ultra-Orthodox women. Am Oved (Hebrew)
Erikson, Erik. 1968. Identity: youth and crisis. Norton.
Fearon DJ Laitin D Violence and the social construction of ethnic identity International Organization 2000 54 4 845 877 10.1162/002081800551398
Feder, Ayala. 2020. Hidden heretics: Jewish doubt in the digital age. Princeton University Press.
Friedman, Menachem. 1991. The Haredi (ultra-Orthodox) society: sources, trends and processes. The Jerusalem Institute for Israel Studies (Hebrew).
Friedman, Menachem. 1999. The king's daughter is all glorious without: the Haredi woman. In D. Ariel-Joel, M. Leibovic, and Y. Mazor (Eds.), Blessed who made me a woman?. Yediot Ahronot-Hemed (Hebrew). (pp. 189–206)
Goldfinger, Rivki. 2019. On a personal note: This time with Havi Ehrenfeld. Channel 7 (Hebrew). https://www.inn.co.il/news/394707
Gutwein, Danny. 2001. Identity versus status: Multiculturalism as a neoliberal theory. Theory and Critique 19: 241–257. (Hebrew)
Hall S Rutherford J Cultural identity and diaspora Identity, community and cultural difference 1990 Lawrence and Wishart 222 237
Haredi Research Group. 2022. https://www.harediresearchgroup.org/
Hendin, Ayala. 2011. Integration of Arab students in the Israeli higher education system. Jerusalem: Hebrew University. (Hebrew)
Horowitz, Neri. 2016. Haredi society—Situation report 2016. The Haredi Institute for Public Affairs (Hebrew).
Jobani Y Perez N Toleration and illiberal groups in context: Israel’s ultra-orthodox “society of learners” Journal of Political Ideologies 2014 19 1 78 98 10.1080/13569317.2013.869454
Kalagy, Tehila. 2012. Between the wig and the veil: a comparative analysis of academization processes among women in conservative societies: the Haredi and Bedouin societies in Israel (Unpublished doctoral dissertation). Bar-Ilan Universith (Hebrew).
Katzir, Shai and Lotem Peri-Hazan. 2021. Discourse, legitimacy, and bottom-up change: An ultra-Orthodox activity to promote secular education in New York Hasidic schools. Journal of Education Policy. 10.1080/02680939.2021.1958261.
Kingsbury, Ian. 2020. Haredi education in Israel: Fiscal solutions and practical challenges. British Journal of Religious Education 42 (2): 193-201.
Layosh, Bella. 2014. The frontier women: Haredi women facing modern change. Resling (Hebrew).
Malach, G., and Ettinger, Y. 2021. Torah and knowledge: Haredi education for boys in the US. Israel Democracy Institute Policy Paper No. 162 (Hebrew). https://www.idi.org.il/media/17416/torah-learning-and-secular-studies-ultra-orthodox-boys-education-in-the-united-states_9-2-22.pdf
Malach, Gilad. 2019. The Haredi parties in Israel: what does the future hold? Israel Democratic Institute (Hebrew). https://www.idi.org.il/articles/25743
Malach, Gilad. 2021. Employment of ultra-Orthodox women – The next revolution? Israel Democracy Institute. https://en.idi.org.il/articles/34689
Matton, Donald I. and Maurine Richardson. 1992. Salad bowl—All together, all unique, all special, ERIC.
Miletzky, Yehudit. 2017. Straddling the fence: Haredi Women in heterogeneous work environments. The Haredi Institute for Public Affairs (Hebrew)
Moskovich Y Liberman I Group identity and social closeness: secular and Ultra-Orthodox Jews in Israel International Journal of Sociology and Social Policy 2018 38 3/4 259 279 10.1108/IJSSP-06-2017-0085
Moustakas C Phenomenological research methods 1994 Sage
Peleg-Gabbay, M. 2022. Haredi employment—An overview. Knesset Research & Information Center (Hebrew). https://fs.knesset.gov.il/globaldocs/MMM/08a5e3e2-3cd1-ec11-8147-005056aac6c3/2_08a5e3e2-3cd1-ec11-8147-005056aac6c3_11_19536.pdf
Perelman, Yossi, Meir Yaish, and Benjamin Bental. 2019. The price of religiosity: Enticing young Haredi men into secular academic studies. Rationality & Society 31 (2): 152–181. 10.1177/1043463119831271
Perry-Hazan, Lotem, Ilana Pinefter-Rosenbluh, and Elizabeth Muzikovskaya. 2021. Secular teachers in Jewish religious schools: Transition, resistance, and freedom of religion. Oxford Review of Education 47 (4): 495–512. 10.1080/03054985.2020.1862780
Rubin Osnat Novis-Deutsch N Haredi women in academia: tradition, modernity and regulative concerns Megamot 2017 2 109 140
Sagy, Avi. 2006. Jewish journey: questions on culture and identity. Shalom Hartman Institute (Hebrew).
Schwartz, Eliezer. 2008. Incentivizing employment of Haredi women. Knesset Research and Information Center (Hebrew).
Shohat, Ela. 2001. Forbidden memories. Bimat Kedem LeSifrut (Hebrew).
Shoval, Dorit. 2010. Us and them: internal dialogue between identity components. (Unpublished doctoral dissertation). Haifa University (Hebrew).
Sicherman, Haim and Lee Cahaner. 2012. Modern ultra-Orthodoxy: The Haredi middle class in Israel. Israel Democracy Institute (Hebrew).
Stern, Y., Yaffa, N., Malach, G., and Malchi, A. 2022. Jewish, Haredi and democratic: the State of Israel and democracy in Haredi eyes. Israel Democracy Institute (Hebrew).
Tajfel, Henri and John C. Turner. 2004. The social identity theory of intergroup behavior. In J. T. Jost & J. Sidanius (Eds.), Political psychology: Key readings. Psychology Press. 10.4324/9780203505984-1 (pp. 276–293)
Tal C Yinon Y The relationship between the personal value-attitude-behavior system and the degree of religiosity among Israeli teachers Studies in Education 1998 3 25 59
Toran, Nina. 2009. Ethnicity, gender and class in academia. Carmel (Hebrew).
Yonah Y Shenhav Y What is multiculturalism? 2005 Bavel On politics of differentiation in Israel
Weissman DR Bais Ya’akov as an innovation in Jewish women’s education: a contribution to the study of education and social change Studies in Jewish Education 1995 7 278 299
| 36467257 | PMC9703422 | NO-CC CODE | 2022-11-29 23:21:08 | no | Contemp Jew. 2022 Nov 28;:1-18 | utf-8 | Contemp Jew | 2,022 | 10.1007/s12397-022-09466-7 | oa_other |
==== Front
Pure Appl Geophys
Pure Appl Geophys
Pure and Applied Geophysics
0033-4553
1420-9136
Springer International Publishing Cham
36466132
3192
10.1007/s00024-022-03192-9
Article
Geophysical Studies of Geodynamics and Natural Hazards in the Northwestern Pacific Region: Introduction
Soloviev Alexander A. 1
Kossobokov Vladimir G. [email protected]
12
Eichelberger John C. [email protected]
3
1 grid.4886.2 0000 0001 2192 9124 Institute of Earthquake Prediction Theory and Mathematical Geophysics, Russian Academy of Sciences, Moscow, Russia
2 International Seismic Safety Organization (ISSO), Arsita, Italy
3 grid.175455.7 0000 0001 2206 1080 University of Alaska, Fairbanks, USA
28 11 2022
2022
179 11 38953902
10 11 2022
17 11 2022
17 11 2022
© The Author(s), under exclusive licence to Springer Nature Switzerland AG 2022
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
issue-copyright-statement© Springer Nature Switzerland AG 2022
==== Body
pmcThe Topical Issue of Pure and Applied Geophysics “Geophysical Studies of Geodynamics and Natural Hazards in the Northwestern Pacific Region” aims to communicate multidisciplinary research focused on one of the most dangerous territories of the world, where inhabitants and infrastructure are exposed to extreme natural hazards associated with unique tectonics and geodynamics of the region. Most of these hazards relate to rapid plate convergence together with tearing and long-distance under-thrusting of the Pacific plate. These have produced many great earthquakes and one of the largest, if not the largest, volcanic eruption of the last millennium. The intersection of the Aleutian and Kamchatka subduction zones forms a sharp cusp in the NW Pacific characterized with one of the highest levels of seismic activity. The Northwestern Pacific Region extends southwards to Philippine and Mariana Islands and eastward to Alaska. The tectonics of the Region is very complex with deep subduction and pull-apart basins. Formation of the gigantic linked dextral pull-apart basin system in the NW Pacific is due to NNE- to ENE-ward motion of east Eurasia in response to the Indo–Asia collision, which started about 50 Ma ago. The large amount of motion of the eastern Eurasia region contradicts any traditional rigid plate tectonic reconstruction, but agrees with the more recent concepts of non-rigidity of both continental and oceanic lithosphere over geological times. Often overlooked is that damage and deaths from earthquakes result more from induced landslides than they do from building collapse from the shaking itself. Other mass movements arise from general crustal instability and deglaciation.
This anthology of current research results from the original inspiration and efforts of Alexander Anatolievich Soloviev. Tragically, he died suddenly on 23rd of September, 2021 while walking with his granddaughter in Moscow. A new corresponding guest editor was assigned in November 2021 to proceed with the work started by Prof. Alexander A. Soloviev. The guest editors dedicate this Topical Issue to his memory. Contained herein are articles on the models, methods, and case studies related to the Northwestern Pacific Region and adjacent territories. In particular -
In Models, Kaban et al. (2022) present “A new Moho map for North-Eastern Eurasia based on the analysis of various geophysical data”. The key fields considered are residual gravity, topography and vertical gravity gradients from the Gravity Field and Steady-State Ocean Circulation Explorer (GOCE). A joint step-by-step analysis of these three fields enables improvement of the final results and better separation of the Moho signal. The new Moho map demonstrates several principal features that were not resolved in the previous studies and correspond well to observed tectonic fragmentation of the study area. New crustal patterns of different kinds have been found under the Verkhoyansk Range, in the continental part of the Laptev rift system, in the East Siberian Sea, in the offshore part of the Chukotka microcontinent, and in the Anadyr-Koryak fold system. The new Moho map is a significant improvement of the previous Moho maps of north-eastern Eurasia.
Zabarinskaya et al. (2022) review “Deep Mariana Island Arc: Highlights of the tectonosphere” in regard to its unique variety of geological features and events, including earthquakes, volcanoes, hydrothermal vents, cold seeps and the largest mud volcanoes on Earth. The Mariana Island Arc is a classic young island arc in the western Pacific Ocean. The authors consider its (i) Tectonic setting wedged between the Eurasian, Pacific and Australian Plates; (ii) Heat flow (up to 2000 mW m−2), which points to both high tectonic activity and an important role played by the convective component in the thermal regime; (iii) Gravity field, which varies substantially in all reductions, reflecting the structural and density inhomogeneities of the tectonosphere; (iv) Seismicity in the northwestward subduction of the Pacific Plate beneath the Philippine Sea Plate, which controls seismic activity within the Mariana Island Arc from aseismically subducting beneath the Philippine Sea Plate without causing large under-thrusting earthquakes at the plate interface due to the weak coupling between the down-going and overlying plates; (v) Volcanism, which are mostly seamounts (underwater volcanoes) with only nine peaks tall enough to form islands and enormous mud volcanoes in the area between two tectonic plates that have been colliding for over 50 Ma and are generating the most extreme fluid composition recorded in the oceans; (vi) Deep structure, which includes the active Mariana Trough, and also a typical active back-arc basin, and two crescent-shaped island arcs, the remnant West Mariana Ridge to the west and the active Mariana Arc to the east.
Bykov et al. (2022) in “Stress transfer and migration of earthquakes from the Western Pacific subduction zone toward the Asian continent” explore the impact of subduction on the geodynamics of the Asian continent by analyzing migration of slow strain and earthquakes from the Nankai, Japan and Kuril-Kamchatka segments deep into mainland. The estimations on the profiles, crossing the Kuril Islands, the Japanese Archipelago and Sakhalin Island toward the Asian continent, have revealed the transverse migration of earthquakes from the Japan–Kuril-Kamchatka subduction zone. The velocities of hypocenter migration of M ≥ 4.5 earthquakes from the Kuril-Kamchatka Trench via northern and central Sakhalin vary from 6 to 17 km/year at different depths. The profiles crossing the islands of Hokkaido and Sakhalin show the M ≥ 4.earthquake migration from the Kuril and Japan trenches at velocities of 8–27 km/year.
Kim et al. (2022) in “Neotectonics at the SE continental margin of the Korean Peninsula: Implications for the back-arc region behind the SW Japan Arc” discuss the geological evolution of the SE continental margin of the Korean Peninsula resulting from crustal extension with back-arc rifting to spreading followed by crustal shortening with back-arc closing. In particular, the authors investigate the geological structure of the area near the largest event ever recorded here instrumentally, the Mw 5.0 earthquake that occurred in 2016. Seismic reflection profiles reveal abundant faults in the epicentral area that make up strike-slip fault systems. They suggest that the Mw 5.0 earthquake occurred due to the reactivation of an extensional fault created during back-arc rifting, which currently induces dextral slip under the ENE–WSW-oriented compressional stress field in and around the Korean Peninsula. The maximum magnitude of earthquakes expected at the margin is estimated as no higher than Mw 6.0. Restoration of seismic profiles indicates that the current stress field was established after 5.5 Ma. The S-wave velocity structure of the uppermost mantle shows asthenospheric upwelling elongated along the continental margin, which may be considered an important regional source of the current stress field by inducing convection in the uppermost mantle toward the Korean Peninsula lithosphere.
Didenko et al. (2022) in “A gravity-derived Moho model for the Sikhote Alin orogenic belt” have performed a two-dimensional power spectrum analysis of the Bouguer gravity field to calculate the crustal thickness (Moho depth) of the study region and adjust the regional model. The new model compared with the available structural and geological data shows that it is highly correlated with the development of Cretaceous–Early Eocene orogenic and post-orogenic granitoid massifs and Cretaceous–Pliocene extrusive igneous rocks. The former geographically coincide with two linear zones in the Moho relief with depths of more than 35 km, and the latter with the Mesozoic–Cenozoic sedimentary basins and the East Sikhote Alin volcano–plutonic belt.
Bergal-Kuvikas et al. (2022) in “Pleistocene-Holocene monogenetic volcanism at the Malko-Petropavlovsk zone of transverse dislocations on Kamchatka: Geochemical features and genesis” present new geochemical and isotopic results of monogenetic volcanism in the study area, based on whole rock and trace element geochemistry. Determinations of the pressure (9–11 kbar) and temperature (1160–1240 °C) conditions using a glass thermobarometer suggest that, prior to eruption, magma of monogenetic cinder cones resided near the Moho boundary. This observation correlates with the crustal discontinuity detected by seismic exploration and magnetotelluric sounding. Although eruptions have not been observed historically in the study area, continued quiescence in the future cannot be guaranteed. Taking into account the location of major population centers of Kamchatka (~ 250,000 people, i.e. ~ 80% of the entire population of the peninsula) the authors highlight an urgent need of continuous monitoring of the nearby volcanoes. More detailed studies regarding the age, volume, and pyroclastic vs extrusive of the monogenetic volcanoes will provide reliable assessment and reduction of potential risks for inhabitants and infrastructures.
Stepnova et al. (2022) in their “Predictive model of rainfall-induced landslides in high-density urban areas of the South Primorsky Region (Russia)” collected all available historical data about landslide incidents in the study area of Vladivostok City and surroundings and have derived a model of logistic regression analysis of antecedent, cumulative, and daily precipitation for forecasting eventual landslides. The advantage of the model is its simple mathematical expression. Despite rather crude reliable precipitation input data from a single meteorological station, accuracy of the model results shows a first approximation efficiency in assessing the probability of rainfall-induced landslides and suggests its further implementation for the purposes of early warning.
In Methods, Pisarenko and Pisarenko (2022) introduce “A modified k-nearest-neighbors method and its application to estimation of seismic intensity”, a new efficient nonlinear quantitative evaluation of the seismic activity based on locations of the observed earthquakes. It requires neither a preliminarily delineated area nor a normalization procedure. The method provides statistically justified output using an explicit form of the ‘‘uncertainty relation’’ between the spatial smoothing effect and random errors and a suggested objective procedure for choosing the nearest neighbors. The analysis of seismic activity in two regions surrounding the Kuril Islands and Japan, 1904–2011, identified “spots of increased seismic activity” along with their quantitative statistical characteristics. The proposed modified k-nearest-neighbors method might prove useful for seismic hazard and risk assessment applicable to locating critical infrastructures, hazardous waste repositories, etc.
Wang et al. (2022) in “Exploring magnitude estimation for earthquake early warning using available P-wave time windows based on Chinese strong-motion records” investigate directly the magnitude-scaling relationships within the available P-wave time window (APTW) defined as the window period starting from the trigger time of the P wave and ending at the arrival of the S wave. They apply this APTW method to explore real-time magnitude estimation. The results of an offline application demonstrate the good performance of the APTW method in terms of the stability of magnitude estimation for small to moderate earthquakes and improved estimation for large earthquakes. This new method can provide stable estimates of earthquake magnitude faster than the routine determinations available, providing an alternative choice for magnitude estimation in earthquake early warning systems.
Agayan et al. (2022) in “Fuzzy logic methods in the analysis of tsunami wave dynamics based on sea level data” present an algorithm for registering the arrival of tsunami waves based on the operational data of sea level measurements. The algorithm makes use of discrete mathematical analysis including a fuzzy logic approach and provides tools for expert assessment at the stage of adjustment and tuning. Its adaptive capabilities allow using real time input data preceding the arrival of the first tsunami wave. This could provide a universal tool for restructuring interpretation of different processes in real time.
Liu et al. (2022) report on “Probabilistic analysis of the landslide hazard in cold regions: Considering multiple triggering factors and their interdependence”. The authors analyze the performance the four machine learning technologies in assessing landslide probabilities under multiple triggering factors. Based on the Copula theory, they have developed the joint distribution of temperature difference and precipitation as triggering factors, determined the critical intensity and the intensity dependence curves of triggering factors under different return periods so as to illustrate their interdependence and influence on the landslide hazard. Making use of the Receiver Operating Characteristic curve (ROC), optimal national maps of landslide susceptibility and hazard were designed. The presented statistical test results allow concluding that, indeed, the landslide hazard probability and hazard intensity are coupled with each other. Therefore, neither the intensity of the hazards nor the precipitating factors can be ignored in a reliable risk assessment for population and infrastructure.
Tao et al. (2022) in their “Test of a PSHA map of China with fortification benefit evaluation” present an innovative methodology for testing probabilistic seismic hazard assessment (PSHA) maps by estimating the two indices of fortification, namely, economic benefit and safety benefit. The authors emphasize the importance of the effect of fortification intensity on the vulnerability of the infrastructure. The estimates of expected losses and casualties are calculated based on the total areas in each of the five damage states with local parameters. The case study of the 1990 PSHA map of China exemplifies its valuable role in earthquake disaster mitigation by the economic benefit of RMB 17.5, as well as by safety benefit of 20,838 fewer deaths and 77,801 fewer serious injuries.
Zhang et al. (2022a) in “Pattern Informatics (PI) of seismicity considering earthquake magnitude? An experiment in the central China North–South seismic belt” investigate the performance of intermediate-term earthquake forecasting of the central China north–south seismic belt for the last quarter century making use of a modified pattern informatics (PI) method. The modification uses the mean magnitude of earthquakes as an index instead of the number of earthquakes above the cutoff magnitude of the original PI analysis. The authors justify the alternative index, named PIm, by physics-based consideration that the mean magnitude of earthquakes is related to the maximum-likelihood estimation of b-values, a characteristic of seismicity associated with the regional level of tectonic stress. It is worth noting that PIm (i) assigns more weight to larger events, (ii) does not outscore the forecast performance of the original PI, (iii) while the superposition of PI × PIm, i.e. PIJ, significantly reduces the number of ‘hotspots’ and thus can reduce the number of false alarms.
Zhang et al. (2022b) in “Time-dependent seismic hazard assessment based on the annual consultation: A case from the China Seismic Experimental Site (CSES)” propose an interdisciplinary approach to time-dependent neo-deterministic seismic hazard assessment (T-NDSHA) for the China Seismic Experimental Site (CSES) at a one year time scale. T-NDSHA combines the neo-deterministic seismic hazard assessment (NDSHA) with the forecast defined by the Annual Consultation on the Likelihood of Earthquakes, organized by China Earthquake Administration (CEA). Since 1972, this interdisciplinary practice of the ‘alert regions’ with increased probabilities of strong earthquakes featured real forward forecasting. The authors take the year 2014 as a showcase example to illustrate how T-NDSHA may be conducted and evaluated. The results of evaluation of the T-NDSHA performance using confusion matrix and the Molchan’s error diagram suggests ready-to-use mapping of expected macroseismic intensities that outperforms random guessing. The combination with NDSHA provides substantial improvement to Annual Consultation. The T-NDSHA approach is applicable to other regions where intermediate-term, middle-range earthquake forecasts are available and where the needs of emergency preparation are duly considered.
Stark (2022) urges in “Pay no attention to the model behind the curtain” that many widely used models are rarely examined carefully to validate a reliable connection to real-world phenomena. Common steps in modeling to support policy decisions may conflict with reality but are convenient, customary, or familiar even in situations where the phenomena obviously violate the assumptions of the models. Impressive computer outputs and quantitative statements about probability, risk, health and/or economic consequences, etc., are often driven by a model behind the curtain—a model to which we are discouraged from paying attention. Not all costs and benefits can be put on the same scale, not all uncertainties can be expressed as probabilities, and not all model parameters measure what they purport to measure. These fundamental ideas are exemplified by considering widespread Probabilistic Seismic Hazard Assessment (PSHA), collisions of birds with wind turbines, statistical analysis of clinical trials, gender bias in academia, soccer penalty cards, climate models, and forecasting impacts of climate change. Multidisciplinary examples illustrate in a clear and understandable way both how not to model data and not to misinterpret statistical tests. In closing, the author re-emphasizes the key principles to help ensure that models serve society: i.e. (i) assess uncertainty and sensitivity; (ii) complexity can be the enemy of relevance; (iii) match purpose and context; (iv) quantification can backfire; and (v) acknowledge ignorance.
In Case studies, Safonov (2022) in the article “The earthquake of February 13, 2020, M = 7.0 and seismotectonic conditions at intermediate depths of the Southern Kuril Islands” performed the inversion of earthquake focal mechanisms of intermediate depth earthquakes for the southern part of the Kuril Islands. He found, in the upper layer, compression prevails along the slab subparallel to the Pacific Plate motion in the mantle, while the extension along the slab prevails in the lower layer and is mainly rotated by 20° clockwise from the direction of its dip. Such a pattern of earthquake focal mechanisms could indicate some additional mantle resistance to subduction. The most complex pattern of the stress is found near the hypocenter of the recent February 13, 2020, Mw 7.0 at about 95 km ENE of Kuril’sk (Russia), where the focal mechanisms in the upper layer more resemble those from the lower layer. The author reports “seismic quiescence” in the lower layer near the focus of this earthquake that might be useful for operational earthquake forecasting in the region.
Klausner et al. (2022) in “Ahead-of-tsunami magnetic disturbance detection using intrinsic mode functions: Tohoku-Oki earthquake case study” document magnetic disturbances that occurred during the Tohoku-Oki tsunami of 11 March 2011 using empirical mode decomposition (EMD) in a dataset derived from a network of ground-based magnetometers (INTERMAGNET and GIS). These disturbances, obtained by filtering the magnetic field data using the first intrinsic mode function (IMF1) of EMD, propagate ahead of the tsunami at a speed in the range of 600 to 1.6 km/s. Thus they appear at the magnetic observatories ahead of the tsunami arrival from about 3 min in the near-field to 2 h or more in the far-field. The authors named these disturbances as “ahead-of-tsunami magnetic disturbances” (ATMDs) and note that ATMDs commonly arrive about 10 min after the arrival of seismic Rayleigh waves. Monitoring of ATMDs in combination with seismic record analysis might become extremely useful for a reliable early warning of disastrous tsunami.
Soloviev et al. (2022) report “On the frequency distribution of geomagnetic K indices in the Northwestern Pacific Region over the 19–24 Solar Cycles” based on a huge collection of analog records that are the oldest measure of geomagnetic activity (K index) from observatories located in the Northwestern Pacific region. The records were digitized by a joint effort of the Russian and Japanese scientific teams, making it possible to study long-term evolution of geomagnetic activity in this region over 1954–2020. The authors reveal appropriate distribution laws, analyze the correlation between time-varying distribution features and sunspot numbers over the 19–24 solar cycles, and establish that the probability of K = 8 or larger events being detected simultaneously at all observatories in the region is less than one hundredth of a percent. The findings are of special importance for regional forecasting and early warning of geomagnetic storms.
Rodkin (2022) in “The variability of earthquake parameters with the depth: Evidences of difference of mechanisms of generation of the shallow, intermediate-depth, and the deep earthquakes” discusses variation of earthquake source characteristics with depth. He suggests that typical values of earthquake source parameters, such as normalized duration of seismic process, apparent stress, etc., vary substantially over depth. These variations are consistent with the presence of deep fluid that decreases the effective friction in rocks and/or by metamorphic processes occurring in down-going slabs including dehydration embrittlement and solid-state (phase) transformations.
Boginskaya and Kostylev (2022) in the article entitled “Change in the level of microseismic noise during the COVID-19 pandemic in the Russian Far East” have noticed a significant sharp decrease up to 30–50% in the daily background seismic noise during the period of the self-isolation. The pandemic’s restrictions on the operation of public institutions and the mobility of the population in large cities of the Russian Far East resulted in the improved quality of seismological observations. This is evident from analyzing seismograms in the period from March 23, 2020 to April 12, 2020 recorded at the seismic stations of Khabarovsk and Vladivostok located in busy parts of the cities, as well as those of the Yuzhno-Sakhalinsk seismic station located in a relatively quiet part of the city. Power spectra and temporal variations of microseismic noise levels based on the broadband seismometers records confirm a strong anthropogenic impact on seismic monitoring and earthquake hypocenter determination. Moreover, these analyzes allowed for identification of the main sources of the induced microseismic noise in or near the cities of Khabarovsk, Vladivostok, and Yuzhno-Sakhalinsk.
Kostylev et al. (2022) consider “Seismic activity in the focus of the Uglegorsk earthquakes, Sakhalin Island, related to intensive development of coal deposits”. The authors studied earthquakes that occur near the Solntsevskoye brown coal field, the most promising deposit on the Sakhalin Island. Active mining is on-going with blasting operations performed on a large scale. Earthquake recurrence graphs in 2000–2010 and 2011–2020 are evidently different and quite significant. Analyzing the spatiotemporal distribution of seismic event epicenters reveals an increase in seismic activity in the region during the past few years and a change in its character from natural to mixed natural and technogenic. In particular, the focal mechanisms of seismic events in 2020 were classified as strike-slip faulting, which is not characteristic of most earthquakes in the Sakhalin Island region. An attempt is made to determine some regularity in the parameters of the produced blasts and earthquakes, using dynamic parameters of the seismic event frequency content including corner frequency of the focal velocity spectrum. The study may be important for practical safety decisions related to the procedures and scale of blasting operations.
Konovalov et al. (2022) report on “Possible connection between recent seismicity and fluid injection in the offshore oil and gas field area of Sakhalin Island, Russia”. They observed that seismic activity on the northeast coast of Sakhalin in 2013–2014 coincides with the start of pilot operations at injection wells in nearby oil and gas fields. Hypothesizing that the recorded seismicity is entirely induced by the fluid disposal, and considering injection rates of 105 m3/year, Konovalov et al. conclude that at least one induced earthquake with a magnitude M 5.5 is highly likely before 2041. The results obtained are of practical interest for developing seismic risk management strategies in the region.
Akiyama et al. (2022) describe “Spatiotemporal changes in fault displacements associated with seismovolcanic events in and around Miyakejima and Kozushima in 2000 inferred from GNSS Data”. This series of seismovolcanic events in central Japan began at the end of June 2000. Detailed analysis, inversion of the observed GNSS data, and location of microearthquakes identified three faults: Fault 1 to the northwest near Niijima; Fault 2 extending northwestward from Miyakejima volcano to Kozushima; and Fault 3 located beneath Miyakejima volcano. The data permitted considering several subfaults and performing more detailed inversion analyses to estimate the spatiotemporal dislocation field on the three faults. The inversion results show that (i) Fault 1 slipped significantly right-laterally by 18.9 m; (ii) the volume change is about 0.40 km3 and 0.74 km3 by opening on Faults 1 and 2, respectively; (iii) Fault 3 deflated by 0.57 km3. Therefore, since the total increase in volume on Faults 1 and 2 is greater than the total deflation on Fault 3 by a factor of 2, a magma supply from greater depths to Faults 1 and 2 is required for the volume overall balance.
Many authors who contributed to this Topical Issue continue to be inspired by Alexander Anatolievich Soloviev (21.10.1947–23.09.2021), a well-known Soviet-Russian geophysicist, former Director of Institute of Earthquake Prediction Theory and Mathematical Geophysics in Moscow (1998–2017), Corresponding Member of Russian Academy of Sciences (2000) and, in particular, by his integrity and a warm friendly attitude to people. Prof. Soloviev dedicated much of his attention to young researchers including participants of a series of Advanced schools on Nonlinear Dynamics and Earthquake Prediction, which he guided at the Abdus Salam Center for Theoretical Physics (Trieste, Italy, 1988–2011). Alexander Anatolievich (Sasha) made fundamental contributions to the development of methods to forecast locations of possible future strong earthquakes and to model dynamics of lithospheric blocks and faults. He developed approaches to universal description of situations preceding extreme events in complex systems of various natures including socioeconomic systems. Among his other scientific results are the discovery of the possibility of generating a magnetic field by the Couette-Poiseuille flow of a conducting fluid and the development of methods for calculating the movement of artificial satellites and other celestial bodies, taking into account perturbations of gravitational field, atmosphere and other factors. His scientific achievements were recognized by the global geophysical community and beyond.
Sasha left deep impression on everyone who met him as a kind gentleman as well as an accomplished scientist. We are devastated to have lost a wonderful man, talented scientist, wise mentor, colleague, and our friend…
Alexander A. Soloviev: 21.10.1947–23.09.2021.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
Agayan SM Bogoutdinov ShR Krasnoperov RI Efremova OV Kamaev DA Fuzzy logic methods in the analysis of tsunami wave dynamics based on sea level data Pure and Applied Geophysics 2022 10.1007/s00024-022-03104-x
Akiyama T Kawabata H Yoshioka S Spatiotemporal Changes in fault displacements associated with seismovolcanic events in and around Miyakejima and Kozushima in 2000 inferred from GNSS data Pure and Applied Geophysics 2022 10.1007/s00024-022-03084-y
Bergal-Kuvikas O Bindeman I Chugaev A Larionova Y Perepelov A Khubaeva O Pleistocene-Holocene monogenetic volcanism at the Malko-Petropavlovsk zone of transverse dislocations on Kamchatka: Geochemical features and genesis Pure and Applied Geophysics 2022 10.1007/s00024-022-02956-7
Boginskaya NV Kostylev DV Change in the level of microseismic noise during the COVID-19 pandemic in the Russian Far East Pure and Applied Geophysics 2022 10.1007/s00024-022-03019-7
Bykov VG Merkulova TV Andreeva MY Stress transfer and migration of earthquakes from the Western Pacific subduction zone toward the Asian continent Pure and Applied Geophysics 2022 10.1007/s00024-021-02924-7
Didenko AN Nosyrev MY Gilmanova GZ A gravity-derived Moho model for the Sikhote Alin orogenic belt Pure and Applied Geophysics 2022 10.1007/s00024-021-02842-8
Kaban MK Sidorov RV Soloviev AA Gvishiani AD Petrunin AG Petrov OV Kashubin SN Androsov EA Milshtein ED A new Moho map for North-Eastern Eurasia based on the analysis of various geophysical data Pure and Applied Geophysics 2022 10.1007/s00024-021-02925-6
Kim H-J Moon S Kim C Kim K-H Seo W Cho K-H Moon H-J Lee GH Neotectonics at the SE continental margin of the Korean Peninsula: Implications for the back-arc region behind the SW Japan Arc Pure and Applied Geophysics 2022 10.1007/s00024-022-02979-0
Klausner V Macedo HG Prestes A Ahead-of-tsunami magnetic disturbance detection using intrinsic mode functions: Tohoku-Oki earthquake case study Pure and Applied Geophysics 2022 10.1007/s00024-021-02919-4
Konovalov AV Stepnov AA Turuntaev SB Possible connection between recent seismicity and fluid injection in the offshore oil and gas field area of Sakhalin Island, Russia Pure and Applied Geophysics 2022 10.1007/s00024-022-03006-y
Kostylev DV Boginskaya NV Zakupin AS Seismic activity in the focus of the Uglegorsk earthquakes, Sakhalin Island, related to intensive development of coal deposits Pure and Applied Geophysics 2022 10.1007/s00024-021-02933-6
Liu Q Tang A Tao X Probabilistic analysis of the landslide hazard in cold regions: Considering multiple triggering factors and their interdependence Pure and Applied Geophysics 2022 10.1007/s00024-022-03152-3
Pisarenko VF Pisarenko DV A modified k-nearest-neighbors method and its application to estimation of seismic intensity Pure and Applied Geophysics 2022 10.1007/s00024-021-02717-y
Rodkin MV The variability of earthquake parameters with the depth: Evidences of difference of mechanisms of generation of the shallow, intermediate-depth, and the deep earthquakes Pure and Applied Geophysics 2022 10.1007/s00024-021-02927-4
Safonov DA The earthquake of February 13, 2020, M = 7.0 and seismotectonic conditions at intermediate depths of the Southern Kuril Islands Pure and Applied Geophysics 2022 10.1007/s00024-021-02926-5
Soloviev A Bogoutdinov S Nisilevich M On the frequency distribution of geomagnetic K indices in the Northwestern Pacific Region over the 19–24 Solar Cycles Pure and Applied Geophysics 2022 10.1007/s00024-021-02862-4
Stark PB Pay no attention to the model behind the curtain Pure and Applied Geophysics 2022 10.1007/s00024-022-03137-2
Stepnova YA Stepnov AA Konovalov AV Gensiorovskiy YuV Lobkina VA Muzychenko LE Muzychenko AA Orekhov AA Predictive model of rainfall-induced landslides in high-density urban areas of the South Primorsky Region (Russia) Pure and Applied Geophysics 2022 10.1007/s00024-021-02822-y
Tao Z Tao X Chen W Tao Z Test of a PSHA map of China with fortification benefit evaluation Pure and Applied Geophysics 2022 10.1007/s00024-022-03103-y
Wang Y Li S Song J Exploring magnitude estimation for earthquake early warning using available P-wave time windows based on Chinese strong-motion records Pure and Applied Geophysics 2022 10.1007/s00024-022-03062-4
Zabarinskaya L Rashidov V Sergeyeva N Deep mariana island arc: Highlights of the tectonosphere Pure and Applied Geophysics 2022 10.1007/s00024-022-02960-x
Zhang S Zhongliang W Zhang Y Pattern Informatics (PI) of seismicity considering earthquake magnitude? An experiment in the Central China North-South seismic belt Pure and Applied Geophysics 2022 10.1007/s00024-022-03079-9
Zhang Y Zhongliang W Romanelli F Vaccari F Peresan A Zhang S Jiang C Panza GF Time-dependent seismic hazard assessment based on the annual consultation: A case from the China Seismic Experimental Site (CSES) Pure and Applied Geophysics 2022 10.1007/s00024-022-03056-2
| 36466132 | PMC9703423 | NO-CC CODE | 2022-12-16 23:17:49 | no | Pure Appl Geophys. 2022 Nov 28; 179(11):3895-3902 | utf-8 | Pure Appl Geophys | 2,022 | 10.1007/s00024-022-03192-9 | oa_other |
==== Front
Theor Appl Climatol
Theor Appl Climatol
Theoretical and Applied Climatology
0177-798X
1434-4483
Springer Vienna Vienna
4289
10.1007/s00704-022-04289-w
Original Paper
Economic evaluation of the climate changes on food security in Iran: application of CGE model
http://orcid.org/0000-0002-1651-0473
Javadi Akram [email protected]
1
http://orcid.org/0000-0003-0993-9244
Ghahremanzadeh Mohammad [email protected]
1
Sassi Maria 2
Javanbakht Ozra 3
Hayati Boballah 1
1 grid.412831.d 0000 0001 1172 3536 Department of Agricultural Economics, Faculty of Agriculture, University of Tabriz, Tabriz, Iran
2 grid.8982.b 0000 0004 1762 5736 Department of Economics and Management, University of Pavia, Pavia, Italy
3 grid.412763.5 0000 0004 0442 8645 Department of Agricultural Economics, Faculty of Agriculture, Urmia University, Urmia, Iran
28 11 2022
119
26 1 2022
8 11 2022
© The Author(s), under exclusive licence to Springer-Verlag GmbH Austria, part of Springer Nature 2022, Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
The present study aims to examine the economic impact of changing climate variables on two components of food security in Iran: availability and access to food. Wheat and rice, the two most important foods in the country, were considered representatives of food security. A CGE model was developed to achieve the research goals. In this context, a stochastic model based on Monte Carlo simulation was used to provide three scenarios (best, average, and worst) indicating probable changes in climate variables. It is important to model the problem of changing climatic variables for irrigated crops, as groundwater resource depletion and restrictions on extraction from Iranian aquifers reduce planted areas and yields. Therefore, this study applies this model to both rain-fed and irrigated crops, whereas studies in the literature only evaluate rain-fed crops. Food security will face serious challenges as food supplies, and consumption of goods and services are declining in average and worst scenarios, according to findings. Consequently, the negative impact of climate change on food security and people’s livelihoods requires a review of the policies implemented within the country. Effective solutions include research and development to introduce drought-tolerant varieties and adopt appropriate strategies to adapt to climate change. Improving the incomes of farmers is one solution to mitigating the impacts of climate change.
==== Body
pmcIntroduction
Climate change and its continuation have been accepted as a serious issue around the world (Mendelsohn 2009). The most important consequence of this phenomenon is a change in the amount and pattern of precipitation (Sassi and Cardaci 2013). Although different economic sectors are affected by climate change, agriculture is largely dependent on climate as the main determinant of the location and production inputs of agricultural activities (Lizumi and Ramankutty 2015). Since water and temperature are the two main factors in the functioning of the physiological systems and in plant growth (Raza et al. 2019; Hatfield and Prueger 2015), changes in temperature and precipitation patterns directly affect agricultural activities. The agriculture sector is known to be the main source of food supply and food security (Pawlak and Kolodziejczak 2020; Hatfield and Prueger 2015). Therefore, climatic variables affect the amount of production (food availability) and also the human and physical capital determining access to food. These are of great importance in arid and semi-arid countries, such as Iran, which is located at mid-latitude with a particular climatic situation. More than 80 percent of the country is located in arid and semi-arid regions. The average annual rainfall in Iran is about 250 mm (mm), less than a third of the global average annual rainfall (Ministry of Agriculture-Jahad of Iran 2020). According to forecasts by the Intergovernmental Panel on Climate Change (IPCC), temperatures will increase by 1.5 to 4 degrees by 2100 and rainfall will decrease by 10 to 40% depending on different regions in Iran. This has been identified by the IPCC as a serious challenge for the production of strategic crops such as cereals that are expected to decline in yield (IPCC 2007).
According to the data provided by the Ministry of Agriculture-Jahad of Iran 2020, the highest level of cultivation among the country’s crops belongs to the cereal group (71.2%), of which the most important crops include wheat with about 69% of the cultivated area. Rice is another grain that makes up an important part of the Iranian grain diet (14% of grain consumption in 2020). These two products are considered in food security because of their importance in the Iranian food diet and also because of the existence of their separate accounts in the input–output table published by the Statistical Center of Iran, which is the basis for the design of the Social Accounting Matrix (SAM) for this study.
Food security was defined at the 1974 World Food Summit as follows: availability of adequate world food supplies of basic foodstuffs to sustain a steady expansion of food consumption and to offset fluctuations in production and prices at all times. In 1983, FAO expanded its concept to include ensuring that vulnerable people have access to available supplies and ensuring that all people at all times have both physical and economic access to the basic food that they need. In 1986, the World Bank’s influential report “Poverty and Hunger” focused on the temporal dynamics of food insecurity. This concept of food security is further elaborated in terms of access to all people at all times to enough food for an active, healthy life. In the mid-1990s, food security was recognized as a significant concern, spanning a spectrum from the individual to the global level. The UNDP Human Development Report 1994 promoted the construct of human security, which encompasses several dimensions, of which food security is only one. This concept is closely related to the human rights perspective on development that, in turn, has influenced discussions of food security. The 1996 World Food Summit adopted an even more complex definition: Food security at the individual, household, national, regional, and global levels is achieved when all everyone at all times has physical and economic access to sufficient, safe, and nutritious food to meet their dietary needs and food preferences for an active and healthy life. This definition is again refined in the State of Food Insecurity 2001, in which food security is a situation that exists when all people, at all times, have physical, social, and economic access to sufficient, safe, and nutritious food that meets their dietary needs and food preferences for an active and healthy life. This new concept emphasizes consumption, the demand side, and the issues of access by vulnerable people to food and focuses on the entitlements of individuals and households. This definition deals with various elements and dimensions, the most important of which are the four elements of “food availability,” “access to food,” “utilization,” and “sustainability.” Changes in climate variables affect both elements of food security (i.e., food availability and access to food) by affecting agricultural production and prices as well as the incomes of people who are dependent on agricultural activities. These two elements of food security at the national level are considered in this study because of their direct impact on the agricultural sector.
According to a report by the Food and Agriculture Organization of the United Nations (FAO) in 2017, Iran is one of the countries with moderate food security conditions. Given the country’s climatic conditions, there is no guarantee that these average conditions will remain stable and not change toward the critical state, although food insecurity has an adverse effect on the health, learning, and economic development of society. Therefore, this study aims to investigate the economic impacts of climate change (effect of precipitation and temperature) on the two elements of food security – food availability and access to food – in Iran.
In recent decades, studies have been multiplied to assess the economic impact of climate change on the agricultural sector and its consequences on food security using different methods, including econometric models, mathematical programming models (partial equilibrium), and computable general equilibrium (CGE) models. However, most of the studies conducted in Iran have investigated the effects of climate change on crop’s yield and production, for example, the study by Vaseghi and Esmaili (2008), which uses the Ricardian method to measure the economic impact of climate change on wheat in Iran. The results show that an increase in temperature and a decrease in rainfall over the next 100 years will lead to a 41% decrease in the yield of wheat in the country. Hosseini et al. (2014) studied the impact of climate change on the agricultural sector in the ZayandehRood watershed using a partial equilibrium method. The results of the positive mathematical planning model show that as a consequence of these changes, the gross profit of agricultural products in this basin will decrease by 18% over the next 30 years. It seems that the assessment of the economic impact of climate change and weather variables on food security and specifically on two dimensions of availability and access to food has received less attention from domestic researchers. This study is the first to attempt to fill this gap in the literature comprehensively and in the form of a CGE model for assessing climate change for food security. However, many studies around the world have examined this phenomenon from the economic perspective of food security, such as Alcamo et al. (2007), Winston et al. (2010), Sassi and Cardaci (2013), Chakrabarty (2016), Yadav et al. (2018), and Kogo et al. (2020). These studies used different approaches such as the Ricardian method, the partial equilibrium model, and the general equilibrium model (CGE) to study climate and food security issues. But the CGE assesses the problem across the economy to the model linkage between climate, agricultural, and economic variables, as well as links between different sectors of the economy versus other models. Accordingly, this paper uses the general equilibrium model (CGE) developed by Lofgren et al. (2002) at the International Food Policy Research Institute (IFPRI) in this issue based on the 2011 SAM for Iran (based on the latest input–output tables published by the Statistical Center of Iran). In addition, it uses a stochastic component predicted by a Monte Carlo simulation to simulate the impact of possible change on climate variables (precipitation and temperature). This method is contrary to the approaches used in the literature that have obtained precipitation models from general circulation models. Thus, the stochastic method predicts the probable mean of the climate variables and their extreme values, with a probability of 90%. Indeed, the stochastic approach considers the probability of the impact of a change in the climate variables. In fact, this paper used a CGE model with stochastic components, as studied by Harris and Robinson (2001) and Sassi and Cardaci (2013). However, with regard to these analyses, this paper considers this approach for both rain-fed and irrigated crops for further and more precise analysis because underground water resources have been discharged due to over-exploitation and there are restrictions on the irrigation of the products. Accordingly, as a result, in this study, the effects of climate variables on the production of irrigated areas, part of the wheat and all rice crops and the modeling of the yield response functions of these products, have also been considered, while they only assess rain-fed crops in their studies. Another novelty of this paper is that it utilizes the economy-wide general equilibrium analysis developed for Iran, going beyond the traditional methods which have determined the effects of physical variables (yield and production) cause of climate change (Vaseghi and Esmaili, 2008; Hosseini et al. 2014; Jafari et al. 2014). In addition to the issue of climate change, other recent issues such as the COVID-19 pandemic and the war between Russia and Ukraine have also led to the greater importance of food security, which shows that the availability and access to food needs more attention. This study attempted to investigate the economic variables related to these two dimensions of food security. The analysis and results of this study may contribute to inform policymakers and planners in Iran to change the political strategy, tackle climate change, and make the right decisions and actions to improve food security in Iran and furthermore to coordinate with international agreements on climate change programs.
The paper is organized as follows. The “Materials and methods” section describes the methodology by determination of productivity functions of crops in response to climate variables, stochastic model (based on Monte Carlo simulation), and CGE model. It is followed by the “Results” section discussing the achievement of the paper based on methodology. The last section concludes and recommends some policies.
Materials and methods
In this study, the methodology is generally presented in three sections: determining the cropping pattern, stochastic model, and economic pattern. In this way, in the first step, the functions used to simulate the effects of climate change variables on the yield of selected crops, namely, wheat and rice, are expressed as representative of food security in different agroecological areas. Then, the method of using these functions to determine the weather forecasting scenarios is presented through a stochastic model based on Monte Carlo simulation. Finally, a CGE model is developed to examine the impact of climate change variables on food security.
Cropping pattern (simulating the yield of rain-fed and irrigated crops)
In the literature, the most common way to simulate the yield of rain-fed crops is to use statistical models (Hayse 2000; Sassi and Cardaci 2013). The basis of statistical models is to use various forms of regression functions to empirically establish a relationship between crop yield and climate parameters. The functional form used to estimate the yield function of rain-fed crops was the quadratic form because it is more adapted to the nonlinear nature of the relationship between climate parameters and crop yields (Harris and Robinson 2001; Estrada et al. 2006; Eboh et al. 2012; Sassi and Cardaci 2013).
In this study, to consider the differences between different provinces in terms of both climate factors and agricultural production status, the United Nations Food and Agriculture Organization (AEZ) agricultural zoning system was used, dividing the country into 10 regions. The map of this division is specified in Fig. 1.Fig. 1 Map of agroecological zoning of Iran
The climate response functions of rain-fed crops yield are specified in Eq. (1):1 yz.t=α+β1RFt+β2RF2t+β3SDTemt+δTt
where yz.t indicates the yield of rain-fed wheat (rice is cultivated only in the irrigated way) in the agroecological zones z at time t; its unit is kg per hectare. RF, RF2, and SDTem are climate variables representing the monthly cumulative precipitation, the square root of the monthly cumulative precipitation, and the standard deviation of the mean temperature during the growth season, respectively. α, β1, β2, β3, and δ are the parameters that must be estimated for each region. T is a time trend variable.
The yield of irrigated crops depends on the farmers’ decision to irrigate in addition to the amount of rainfall during the growth season. In such cases, the use of the function of yield response to water proposed by the Food and Agriculture Organization of the United Nations (FAO) is effective in simulating the relationship between climate variables and the yield of irrigated crops (Evans et al. 2003; Rodrigues et al. 2009). In this function that was developed by Doorenbos and Kassam (1979), the yield response of each crop is estimated using the relationship between relative yield (Ya/Ym) and relative evapotranspiration (ETa/ETm). The mathematical form of the function of yield response to water of Doorenbos and Kassam (1979) is shown in Eq. (2).2 YaYm=1-ky+ky(ETaETm)
where Ya is the actual crop yield value (kg per hectare) and Ym is the potential crop yield value or maximum crop yield value (kg per hectare) which is determined based on field studies and expert opinions with accepting degrees of error. ky shows the coefficients of the crop yield response to water which are published in FAO Irrigation and Drainage paper No. 33 for different crops. ETa and ETm are the actual values of evapotranspiration (i.e., the sum of effective rainfall plus irrigation water) and the potential evapotranspiration, respectively. Given the potential yield values for each crop, ky and ETm coefficients were obtained from Cropwat software; for each actual annual yield value, a value is obtained for ETa by the mathematical equation mentioned in Eq. (2).
Since the variable ETm is directly related to the growth season temperature, Eq. (2) is rewritten as the relationship between the annual ETm and the average growth season temperature (TEM) for each agroecological zone to explain the relationship between climate variables and crop yield using linear regression method, as follows (Hosseini and Nazari 2015).3 Ya=Ym1-ky+ky(ETaa+b.TEM)
where a and b are the parameters that must be estimated.
Stochastic model based on Monte Carlo simulation methodology
After estimating the parameters of the yield climatic response functions, for each explanatory variable, a probability density function (PDF) is specified. For the amount of monthly cumulative rainfall and standard deviation of the mean temperature, references are made to the historical data of the period under study, which is 37 covering the years 1983–2019 as the changes in climate variables. It is defined by introducing a hypothesis on the lower and upper bound. According to Sassi and Cardaci (2013), this study assumed the lower bound equal to zero because the precipitation cannot be negative and the upper bound is limited but unknown to include extreme weather events. Then, based on three statistical tests measuring the similarity of historical data of variables with different PDFs, the type of PDF of the relevant variable is determined. These tests include chi-square (C-S), Kolmogorov–Smirnov (K-S), and Anderson–Darling (A-D). In this study, we defined three climate scenarios including worst, average, and best scenarios through specific changes in the rainfall (RF) and standard deviation of mean temperature (SDTem), as shown in Table 1.Table 1 Definition of three climate scenarios
Scenario Precipitation (mm) Temperature (°C)
Worst scenario 0 ≤ RF < 119 0 ≤ SDTem < 8.7
Average scenario 119 ≤ RF ≤ 280 8.7 ≤ SDTem ≤ 11.30
Best scenario RF > 280 SDTem > 11.30
Source: research calculations
After selecting the PDF for each explanatory variable, these functions replace the explanatory variables in the climatic response function of the respective yield; then, the stochastic model is calculated assuming a repetition of 1000 times for each crop. Finally, a cumulative density function is obtained for the crop yield, which is the output of each stochastic model. This function indicates the probability that Qc (crop yield of c) is less than or equal to the specified value of qc (i.e., f(Qc) = Prob (Qc ≤ qc)). Accordingly, three numbers are obtained to predict the mean, upper delimiter, and lower delimiter values of the yield variable. The difference between these numbers and the corresponding number in the SAM of 2011 gives the average, best, and worst scenarios, respectively, to simulate with the CGE model.
Finally, the numbers extracted from the stochastic model for each crop are added to one to obtain a parameter as a shock entering the CGE model (Sassi and Cardaci 2013; Harris and Robinson 2001). The mathematical form is presented in Eq. (4).4 rfcs=1+whcs;c=WheatandRices=Average,BestandWorstscenarios
rfcs is a parameter causing climate shock under three scenarios s to produce selected crops c in a CGE model. whcs s is equal to the results obtained from the Monte Carlo stochastic model.
Computable general equilibrium model
In this study, the main framework of the CGE model was made using the CGE model developed in the IFPRI by Lofgren et al. (2002) and Sassi and Cardaci (2013), according to the economic characteristics of Iran. The model data source is based upon the SAM compiled by this study for 2011, which was obtained using the latest input–output tables published by the Statistical Center of Iran. This matrix includes eight productive activities in five economic sectors including agriculture, food industry, other industries and mines, energy, transportation, and services producing a total of ten items of commodities and services. Table 2 shows the accounts of this matrix.Table 2 Accounts in the SAM of adjusted by the present study
Activities Commodities Factors Institutions Taxes Other accounts
Crop activity (CROP-A) Wheat commodity (WHT-C) Labor (LAB) Urban households (UHH) Indirect taxes (IN-TAX) Saving and investment (S-I)
Horticulture activity (HORT-A) Rice commodity (RIC-C) Capital (CAP) Rural households (RHH) Taxes (TAX) Rest of the world (ROW)
Other agriculture activity (OAGRI-A) Other crop commodity (OCROP-C) Land (LAND) Government (GOV) Subsidy (SUB)
Food industry activity (FODING-A) Horticulture commodity (HORT-C) Governmental enterprises (G-INTR)
Other industries and mines activity (OINDMIN-A) Other agriculture commodity (OAGRI-C) Non-governmental enterprises (NG-INTR)
Energy activity (EN-A) Food industry commodity (FODING-C)
Transportation activity (TR-A) Other industries and mines commodity (OINDMIN-C)
Services activity (SERV-A) Energy commodity (EN-C)
Transportation commodity (TR-C)
Services commodity (SERV-C)
Various data sources have been used to calculate the figures in the social accounting matrix, the most important of which include the input–output tables that have been used to calculate accounts of activities and commodities; the statistical yearbook, the balance sheet and economic reports published by the Iran central bank which is used in extracting the accounts of government, enterprises, and taxes; the results of the labor force survey that has been used in obtaining the account of factors as well as the results of the urban and rural household income and expenditure survey which has been used in separating the accounts of urban and rural households. In this way, the average annual income of an urban household and a rural household has been determined according to the types of income, which can be used to separate the income of factors between households. For this purpose, the annual income of each urban and rural household is multiplied by their population in 2011 to determine the annual income of all urban and rural households. In this way, the total share of urban and rural households from all types of income is calculated.
According to Fig. 2, the component of climate change with two characteristics of rainfall and temperature affects the value-added function of the agriculture sector (Harris and Robinson 2001; Karaky 2002; Sassi and Cardaci 2013). Therefore, the value-added equation is modified. A change in the value-added of activities leads to a change in the level of activity and then a change in the quantity of commodity offered in the market, indicating the availability of food in the domestic market. The change in climate variables affects crop yields. This affects food production and availability as well as prices, leading to scarcity and food insecurity (Rademacher-Schulz et al. 2012). In fact, the reduction in the supply of food is combined with the increase in their prices; it has a negative effect on the consumer group, while about 50% of the country’s arable land is in the rain-fed sector (Ministry of Agriculture-Jahad of Iran 2020), and the employment, income, and livelihood of the population working in this sector are also affected which means reduced economic access to food.Fig. 2 The effects of climate variables on food security
In other words, the following structure fits the SAM and represents that production activities are the supply side of the model and the flow of market goods is the demand side of the model. Food availability is determined by the disposable amount of composite commodities that, in combination with market prices and household income, brings about economic access to food, which is represented by consumption.
Within the framework represented by Fig. 2, the component of climate variable affects the function of value-added. Thus, the value-added equation is modified. The basic form of the function of value-added is shaped as Eq. (5).5 QVAa=αava(∑fδfava∙QFfa-ρava)-1ρava
where QVAa is the value-added, αava is the parameter related to efficiency, δfava represents the parameter of the share of factor f in activity a, QFfa is the value demanded of factor f from activity a, and ρava is the exponent of the value-added function obtained from the elasticity of the substitution of the factors of primary production (capital and labor).
The shock parameter introduced as rfcs represents the weather shock on the various activities of the producers of activity a. It was entered into Eq. (5), affecting the other economic parameters and variables (Harris and Robinson 2001; Sassi & Cardaci 2013).6 QVAa=rfcs∙αava(∑fδfava∙QFfa-ρava)-1ρava
All equations indicating the dimension of food availability and access to food are presented in the Appendix.
The equations are coded related to the decisions of economic agents and the equations associated with the constraints related to the economic system in GAMS software, and the initial values of variables and elasticities are determined from the SAM matrix database. The elasticity values are available in Table 9 and Table 10 in the appendix.
At this stage, the desired shock, i.e., the change of variables or parameters caused by climate change, is entered into the model, and the model is solved again to check the situation before and after the shock.
In this study, two dimensions of food security can be examined by investigating the changes in the variables of composite commodities, domestic supply, imports, and domestic demand price to show the dimension of food availability and the dimension of access to food under three climate scenarios by examining the changes in commodity price variables, the quantity of consumed commodities by household.
The data required includes a wide range of meteorological data at the level of synoptic stations including time series in the period 37 years (from 1983 to 2019) of monthly rainfall; minimum, maximum, and average monthly temperature; average wind speed: hours of sunshine per day and average monthly humidity; agricultural and cropping data in the provinces of the country include the yield of wheat (rain-fed and irrigated) and rice and production costs of some selected products to separate value-added between factors. The statistical references for collecting the mentioned data are the Meteorological Organization of Iran (2020) and the database of the Ministry of Agriculture – Jahad of Iran (2020).
Results
The simulation of the change of weather variables on the yield of crops has been done separately for 31 provinces of the country. But considering that the study units are agroecological zones of the country which include several provinces, they were aggregated into agroecological zones by averaging. In Table 3, the climatic characteristics of each of these zones in the period under review are reported.Table 3 Climatic characteristics of agroecological zones of the country
Agroecological zones Average rainfall (mm) Standard deviation of rainfall Average temperature (°C) Standard deviation of temperature Yield of rain-fed wheat (Kg/ha) Yield of irrigated wheat (Kg/ha) Yield of rice (Kg/ha)
AEZ.1
Northwestern zone
307.6 83.7 11.6 0.97 2671 2884 -
AEZ.2
Caspian coastal plain
926.3 166.7 16.8 0.61 2250 2721 3954
AEZ.3
Central Zagros zone
450.7 107.8 15.1 0.93 2798 2945 -
AEZ.4
Central zone
232.5 67.7 16.5 0.75 3256 3176 -
AEZ.5
Khorasan zone
222.20 68.1 14.8 0.79 2482 2630 -
AEZ.6
Arid central zone
91.7 37.3 18.2 0.75 1544 2184 4818
AEZ.7
Khozestan zone
226.4 79.6 25.8 0.75 2464 2551 3263
AEZ.8
Southern Zagrose zone
363.1 127.7 17.8 0.71 2923 3037 4264
AEZ.9
Arid southern zone
105.8 46.2 17.9 0.81 - 2378 -
AEZ.10
Southern coastal plain
207.4 113.9 26.1 0.48 - 2436 -
Source: research calculations based on data from the Meteorological Organization of Iran (2018) and the Ministry of Agriculture – Jahad of Iran, 2018
According to Table 3, AEZ.2 is the most abundant rainfall agroecological zone and AEZ.6 and AEZ.9 are the least rainfall agroecological zones in the country. AEZ.1 is the coldest zone, and AEZ.7 and AEZ.10 are the warmest agroecological zone in the country. Naturally, changes in climate variables will affect the activities of the agriculture sector. The decrease in rainfall and the change in its pattern in the rain areas will reduce the yield and potential production of rain-fed crops and the economic instability of production in these zones. In irrigated areas, the direct effect of the decrease in rainfall and the increase in temperature, especially for crops such as cereals that growth period coincides with the rainy season, yield reduction will occur.
Cropping pattern results (simulation of crop yield to climate variables change)
The results of estimating the climatic response functions of rain-fed wheat yield in each agroecological zone have been reported in Table 4. In AEZ.2, it is not possible to estimate the yield function of rain-fed wheat for this area due to heavy rainfall and excessive water storage in the soil. Also, according to the statistics of the Ministry of Agriculture – Jahad of Iran, in AEZ.9 and 10, the cultivation of rain-fed crops is very small. Therefore, no function was estimated for these zones.Table 4 The estimated coefficients of climatic variables in the yield function of rain-fed wheat in agroecological zones
Agroecological zones Monthly cumulative rainfall (RF) Square of RF (RF2) Standard deviation of temperature (SDTEM) Time trend (T)
AEZ.1 3.3459 (0/000) − 0.0031 (0.003) − 70.0183 (0.000) 10.9344 (0.000)
AEZ.3 2.7027 (0.000) − 0.0015 (0.002) − 96.1583 (0.001) 10.4559 (0.000)
AEZ.4 3.1315 (0.013) − 0.0031 (0.037) − 76.4611 (0.021) 11.4862 (0.000)
AEZ.5 1.9351 (0.050) − 0.0014 (0.008) − 91.6821 (0.000) 11.6381 (0.015)
AEZ.6 3.9665 (0.088) − 0.0155 (0.052) − 73.7211 (0.078) 11.6611 (0.000)
AEZ.7 8.2061 (0.009) − 0.0264 (0.068) − 156.4512 (0.005) 12.0132 (0.020)
AEZ.8 2.7951 (0.072) − 0.0030 (0.069) - 5.1631 (0.078)
Source: research findings
In general, the results show that the yield of rain-fed wheat is affected by rainfall and temperature. In cold regions (AEZ.1, AEZ.3, and AEZ.5), the most important rainfall includes spring rainfall, especially in May and June, and then autumn rainfall, especially in November. The reduction in these rainfalls has a negative effect on the yield and production of this crop. For the southern regions of the country with tropical climates (AEZ.7, AEZ.8, AEZ.9, AEZ.10), the rainfall of the winter months is more important. Also, the temperature of the warm months and the late growing season of this crop are also important for these regions, and their increase has a negative effect on the yield of wheat.
Table 5 presents the potential yield values (Ym), yield sensitivity coefficient to evapotranspiration (ky), and values of coefficients a and b for irrigated wheat based on the Doorenbos-Kassam equation in each of the agroecological zones of the country. The values obtained for coefficient b indicate an increase in the water requirement of the wheat crop per unit increase of temperature.Table 5 The values of Ym and Ky and values of coefficients a and b for irrigated wheat crops in each of the agroecological zones
Agroecological zones Potential yield (ym) (Kg/h) ky a b
AEZ.1 5000 1 282.91 24.99
AEZ.2 4300 1 − 73.04 36.63
AEZ.3 5500 1 332.28 39.09
AEZ.4 5100 1 231.10 28.76
AEZ.5 4000 1 645.59 20.40
AEZ.6 4900 1 − 73.57 79.01
AEZ.7 5700 1 − 652.75 77.04
AEZ.8 5000 1 621.24 17.63
AEZ.9 3300 1 403.66 64.63
AEZ.10 3600 1 1839.27 28.25
Source: research findings
Table 6 shows the values of these coefficients for the rice crop. The provinces producing more than 90% of the country’s rice include four zones (AEZ.2, AEZ.6, AEZ.7, and AEZ.8). Rice production in other areas was less than 10% of the total production of the country, for which the estimation of the Doorenbos-Kassam equation was avoided owing to the lack of permanent production in these areas during the study season.Table 6 The values of Ym and Ky and values of coefficients a and b for rice crops in each of the four agroecological zones
Agroecological zones Potential yield (ym) (Kg/h) ky a b
AEZ.2 4700 0.99 − 939.77 69.73
AEZ.6 5700 0.99 − 1019.41 77.69
AEZ.7 4300 0.99 276.83 13.46
AEZ.8 5300 0.99 634.47 18.29
Source: research findings
According to the results of Table 6, AEZ.6 has a higher value than the other zones for b, which indicates an increase in the water requirement of the rice crop per unit increase of temperature. This is probably due to the hot weather in summer in this region significantly increasing the evapotranspiration of the plant.
Results of stochastic model based on Monte Carlo simulation
In this section, to aggregate the yield values of wheat and rice in the agroecological zones, the average production of these crops during the period under review was used as the weight for averaging. On the other hand, since in the SAM, wheat is an account and is not separated into rain-fed and irrigated, rain-fed and irrigated wheat are also combined to obtain scenarios based on wheat yield. Table 7 shows the forecast of mean, upper, and lower yield values for wheat and rice in the whole country. As mentioned, the difference between these numbers and the corresponding actual yield of crops is the basis for calculating the shocks to the CGE model, defined in different scenarios for the model. The magnitude of these shocks is presented in Table 8.Table 7 The prediction of mean, upper, and lower values of wheat and rice yields in the whole country (kg/ha)
Product Mean value Upper delimiter Lower delimiter
Wheat 1697.67 2779.57 420.86
Rice 3900.71 6283.10 1640/08
Source: research calculations
Table 8 The simulated shock parameters entering the CGE model
Predicted change in yield based on predicted climate variables Shocks
Scenario Wheat Rice Wheat Rice
Average scenario − 0.0799 − 0.001 0.9201 0.999
Best scenario 0.5064 0.7591 1.5064 1.7591
Worst scenario − 0.7719 − 0.5408 0.2281 0.4592
Source: research calculations
Table 7 shows that in a normal climatic situation, the yield of wheat will be 1698 kg/ha. This figure will reach 2780 kg/ha in the best climatic conditions and 421 kg/ha in the worst climatic conditions. These predictions for the yield of rice show 3901 kg/ha for normal climatic conditions, 6283 kg/ha for best climatic conditions, and 1640 kg/ha for worst climatic conditions.
The figures inserted in the first and second columns of Table 8 show the predicted changes in wheat and rice yields compared to 2011 based on the predicted changes in climate variables. The numbers − 0.0799 and − 0.001 obtained for the wheat and rice crop in the average scenario indicate that in the normal climatic situation, the wheat yield will decrease by about 8% and the rice yield will have a decrease of 0.1%, which indicate that the prevailing weather conditions are not very favorable for these products. This situation is more prominent in the worst scenario. The results are similar to the studies by Vaseghi and Esmaili (2008), Parhizkari et al. (2014), and Hosseini and Nazari (2015) that reported a decrease in the yield of crops, especially cereals, due to the decrease in average rainfall and increase in average temperature in Iran.
The two columns on the right in Table 8 represent the shocks that will be applied to the basic CGE model. These numbers were obtained by inserting them in Eq. (4).
Results of the CGE model
Effects of climate variables change on the dimension of food availability
This section deals with examining the consequences of applying shocks calculated on economic components including the quantity of composite commodities, domestic supply, imports, and domestic demand prices of all ten groups of commodities and services introduced in the SAM. The effect of the simulated scenarios in this study on food availability with regard to the quantities of commodities accessible in the domestic market is presented in Fig. 3. In general, the results show the effect of changes in rainfall and temperature variables on wheat and rice, which is in line with the findings of Khalilian et al. (2014), Khaleghi et al. (2015), Ghaffari Esmaeli et al. (2018), and Eslami (2020) who confirm that there is a significant relationship between climate variables including temperature and precipitation with the amount of production in the agricultural sector especially decrease in the supply of wheat products due to this phenomenon in Iran. The commodities that are affected by the shocks directly, namely wheat, and rice, have undergone more changes. According to the figure, the amount of wheat commodity with a 90% probability will decrease between − 5.24% and − 46.12%, while this amount is between − 0.01% and − 28.78% for rice because of a 1% decrease in average rainfall. In the best scenario, it is expected that due to an increase of 1% in the average rainfall, the increase in the amount of wheat is 30.28% and about 42.36% for the amount of rice with a 90% probability. It is noteworthy that the reason for the lower reduction effect on rice than wheat is partly due to the greater dependence of wheat on rainfall because part of the wheat commodity is its rain-fed type. Also, as seen in Fig. 3, the commodity group of other cereals, horticultural crops, and other crops show relatively more changes compared to the other groups of commodities and services. This result was stressed by Khaleghi et al. (2015) which argue that although different economic sectors are affected by climate change, this effect is more on the sectors that are more interconnected with the agriculture sector. However, the path of changes of all groups of commodities and services in the scenarios is the same.Fig. 3 Change in the amount of composite commodities based on scenarios and commodities.
Source: research findings
The quantities of domestic supply of commodities and their import under three scenarios simulated in this study are reported in Figs. 4 and 5. Examining these two diagrams and comparing them, the amount of imports of commodities is not in balance with the decrease in domestic supply. It should be noted that according to the worst and average scenarios, the decrease in rainfall has led to a reduction in the region under cultivation and lower yields in the whole country, leading to a decrease in the average production of crops. The result is consistent with the studies by Goodbody et al. (2012), Sassi and Cardaci (2013), and Gouel and Laborde (2021), who agree that food availability is expected to deteriorate significantly, driven by inadequate rainfall. Therefore, the decrease in the quantity of commodities produced in the country has led to a reduction in their domestic supply. For example, with a 90% probability, the supply of wheat decreases between − 7.07% and − 66.86%, and the supply of rice decreases between − 0.79% and − 66.64% due to an increase of 1% in the average rainfall. The reverse is true in the best scenario because, as presented in Fig. 4, wheat and rice commodities will increase by 23.07% and 33.18%, respectively. In the case of other groups of commodities and services, the best scenario indicates an increase in the domestic supply of commodities, which may lead to a reduction in the domestic price of commodities and affect its imports.Fig. 4 Change in the domestic supply by scenarios and commodities.
Source: research findings
Fig. 5 Change in the amount of imported commodities based on scenarios and commodities.
Source: research findings
In determining the price of demand for domestic products, the quantities of domestic supply and imports, in combination with the amount of elasticity of their domestic demand substitution, are influential criteria. In the case of imports of strategic commodities such as wheat, other factors such as policies in the field of self-sufficiency in its production and the lack of import licenses can also be involved. As presented in Fig. 6, the quantity of rice imports with a 90% probability will increase between 2.08% and 38.61%. Rice is the only commodity group with increased imports due to the decrease in domestic production. Therefore, the domestic demand price of this product is expected to decrease, but according to Fig. 6, this has not happened. It seems that the reason for the increase in the price of its demand despite the increase in the amount of imports is related to the low elasticity of domestic demand substitution.1Fig. 6 Changes in domestic demand prices by scenarios and commodities.
Source: research findings
The change in the domestic demand price of commodities in the scenarios considered in this study is presented in Fig. 6. According to the graph, the price of rice is likely to increase significantly by 90% probability, which is between 24.68% and 108.07% (worst scenario and average scenario). Hence, an increase in its imports can be observed (Fig. 5). However, in the best scenario, owing to the increase in domestic supply, there is a possibility that the price of this product will decrease. The amount of this decrease will be 6.91%. Panahi et al. (2015) also concluded that climate change affects almost 30% of rice yield in Iran in the next 15 years and will eventually lead to an increase in its domestic demand price. Domestic demand price for wheat also shows an increase between 12.39% and 52.48% (average and worst scenario). In general, the decrease in the domestic supply of wheat and rice in average and worst scenarios increased the price of demand for these commodities in the country.
Effects of climate change variables on the dimension of access to food
Implementation of the proposed scenarios on the dimension of access to food is done by examining the status of the variables of composite commodities’ prices and the consumed commodities by a household for 10 groups of commodities and services. The data in Fig. 7 show the results of the simulation of the composite commodities price under the conditions of three scenarios. In the domestic market, the decrease in food supply under average and worst scenarios is combined with the increase in wheat and rice prices. With a 90% probability, it will increase the price of wheat between 6.94% and 34.76% and rice between 10.68% and 87.04%. The result is in accordance with the studies by Nelson et al. (2009) and Kogo et al. (2020), who express that climate change will cause price increases for major agricultural crops, such as rice, wheat, maize, and soybean. This situation has the most negative impact on consumers because their food security and living standards are endangered.Fig. 7 Change in the composite commodities price based on scenarios and commodities.
Source: research findings
Combining these conditions with the reduction of household income, especially in the rural group, a simulation of access to food is created, leading to a reduction in the consumption of all commodities for both urban and rural households. The reverse case happens in the best scenario. That is, if the scenario conditions are the best, with the improvement of production and yield of selected products, there will be a slight decrease in the price of composite commodities.
Following the decrease in the supply of commodities and, consequently, the increase in their prices, the consumption of different groups of commodities and services will also decrease. These results are in line with the findings of Pakravan et al. (2015) in Iran, who showed that in both urban and rural areas, the level of household food security index had a descending trend from 2005 to 2012. Meanwhile, in a study conducted by FAO (2015) to investigate the risks and responses of climate change on food security in several African countries, it was highlighted that both rainfall and temperature variability appear to exert a negative impact on household consumption and access to food.
According to Fig. 8, the demand for rice had the highest expected decrease for rural households. Hence, there is a decrease between − 8.10% and − 48.29% with a probability of 90%. Generally, this group of households has the most negative impact of simulated shocks for rice and wheat products compared to urban households (a reduction of 48.29% in rice consumption and 17.63% in wheat consumption in the worst scenario). However, for other commodities, the severity of shocks in both average and worst scenarios is imposed on the urban household. This seems to be due to the lower elasticity of demand for rice for urban households than for rural households. This is the opposite in the case of the best scenario, where there is a 90% probability for the increase of household demand from all groups of commodities and services. However, the severity of this increase is higher for wheat and rice than for other commodities. Moreover, the increase in demand for rice commodities for rural households in this scenario is higher than in urban households. Meanwhile, Raj et al. (2022) also concluded that the issue of climate change and its impact on access to food and food security is severe for rural households.Fig. 8 The diagrams of change in urban and rural household consumption by scenarios and commodities.
Source: research findings
According to the theory, the level of change in the consumption level of each commodity is a function of the change level in its price and the amount of demand elasticity. The greater the price elasticity of a product or the greater the level of change in its price level, the greater the percentage reduction in the amount of consumption of that product. Examining the diagrams in Fig. 8, it is found that the percentage reduction in wheat consumption is calculated as 35%. However, the percentage increase in its price in the worst scenario is much lower, at about 18% for rural households and 9% for urban households. This seems to be due to the low elasticity of demand for this product, which is also true for other products.
Conclusion
In this study, considering a computable general equilibrium (CGE) model, the economic effects of climate variables change were studied on two important food products of the country, wheat and rice, representing food security. Climate yield response functions for rain-fed and irrigated crops were used to simulate the climate scenarios. The estimated functions have affirmed a direct correlation of rainfall and indirect correlation of standard deviation of temperature with the yield of selected crops. Then, based on the Monte Carlo simulation, it was obtained climate predictive scenarios in three modes of worst, average, and best to simulate in the CGE model and compare with the baseline. These scenarios are illustrated based on a PDF (probability density function) of observed data from 1983 to 2019 with the sign of the mean and the delimiters values corresponding to a 90% probability, which it confirmed a reduction in the yield of the selected crops except in best scenario (optimistic condition).
By looking at the components relative to the dimension of food availability, it is concluded that its decline is due to a reduction in the amount of wheat and rice commodities available in the domestic market. In fact, the impact of climate change on food security through this dimension occurs due to changes in the productivity of selected products. About wheat products, this issue is significant due to the policy of reducing wheat imports. While the deterioration of access to food is due to households reducing consumption of commodities and services due to rising prices. The decrease in private consumption in worst and average scenarios could be a sign of economic inefficiency under these conditions.
This paper has clearly been worked to examine food security in Iran under climate change. Because the current policy approach about climate change needs to be reviewed and reformed. Results achieved suggest policymakers to work to maintain and increase the production of strategic crops such as wheat and rice. Among the effective solutions are more research and developments to introduce drought and heat-resistant cultivars. Maybe changing the planting date of such crops to prevent their growth period from adapting to moisture stresses as much as possible can be considered as a simple solution. Adopting appropriate strategies compatible with climate change is also recommended, including the use of modern irrigation systems, low-volume irrigation methods, and improving the pattern of cultivation, improving farmers’ incomes, and developing the food industry system. It is also recommended that the government should pay special attention to the phenomenon of climate change and its effects on food prices in the country’s macro-planning, especially food inflation targets. Finally, by more disaggregation of accounts in SAM, it will be more beneficial to investigate the impact of weather variables change on food security. However, due to the lack of necessary data in this study, the goods and services were separated into ten groups.
Appendix
The basic form of the function of value-added is shaped as Eq. (7).7 QVAa=αava(∑fδfava∙QFfa-ρava)-1ρava
where QVAa is the value-added, αava is the parameter related to efficiency, δfava represents the parameter of the share of factor f in activity a, QFfa is the value demanded of factor f from activity a, and ρava is the exponent of the value-added function obtained from the elasticity of the substitution of the factors of primary production (capital and labor).
Equation (8) denotes the activity production function with CES2 technology. This relationship relates the level of QAa activity with the CES function to the total value-added QVAa and the total QINTAa intermediate inputs.8 QAa=αaa(δaa∙QVAa-ρaa+1-δaaQINTAa-ρaa)-1ρaa
where QAa is the level of activity a, αaa is the parameter for efficiency, δaa shows the shared parameter of each factor of production, and ρaa denotes the exponent of the CES function, which is obtained from the substitution elasticity of the factors of production.
Equation (9) shows the total quantities of commodity c produced by all activities, which is a CES function of the commodities c produced by different activities.9 QXc=αcac(∑a∈Aδacac∙QAac-ρcac)-1ρcac-1
in which QXc is the quantities of commodity c produced by all activities, αaac is the transfer parameter of the function, δacac denotes the parameter of the share of variables in the function, and -ρcac shows the exponent parameter of the total output function derived from the substitution elasticity between the produced commodity c through different activities.
The domestic-produced commodities and services presented in Eq. (9) are either sold domestically or exported abroad. This assignment by the CET3-type function is as follows.10 QXc=αct(δct∙QEcρct+1-δctQDcρct)1ρct
where αct is the transfer parameter of the function, δct is the parameter of the share of commodities in the function, and ρct is the exponent parameter of the CET function obtained from the elasticity of substitution between domestic sales and exports. It indicates an incomplete substitution between two commodities. QE and QD also represent the amount of exports of commodities and the amount of domestic sales, respectively.
The above-mentioned domestic-produced commodities are combined with imports to produce a commodity called a composite commodity, which is, in fact, demanded in the market. Equation (11) presents the supply function of this composite commodity, known as the Armington function, that there is an incomplete substitution between domestic commodities and similar imported commodities, that there is an incomplete substitution between domestic commodities and similar imported commodities.11 QQc=αcq(δcq∙QMc-ρcq+1-δcqQDc-ρcq)-1ρcq
where QQc represents the quantity of composite commodity c; the other parameters used in this function have maps similar to the previous CET function. Equations (12) and (13) show the demand functions for domestic productions and imports.12 QMc=(αcq.δcq.PQcPMc)-11-ρcq.QQc
13 QDc=(αcq.(1-δcq).PQcPDDc)-11-ρcq.QQc
In this section, the domestic demand price and the price of composite commodities are presented. Equation (14) denotes the domestic demand price of commodities.14 PDDc=PDSc+icdc
PDDc indicates the demand price of commodity c which produces and sells domestically, PDSc represents the supply price of commodity c which produces and sells domestically, and icdc is the sales and shipping tax costs.
Equation (15) shows the price of PQc composite commodities as a weighted combination of the price of commodities sold domestically (PDDc) and the price of imports (PMc). The weights of this equation are the quantity of composite commodities (QQc), the amount of commodities sold domestically (QDc), and the amount of imported commodities (QMc). This price is at the demander level owing to the sales tax and shipping costs.15 PQc∙QQc=(PDDc∙QDc)+(PMc∙QMc)
Finally, assuming that each household maximizes the Aston-Gray utility function concerning its consumption expenditure, the result of the first-order condition is a linear expenditure system (LES) function, indicating that the expenditure is linear. Household consumption is related to the total income. The LES demand function of household consumption is shown in Eq. (16).16 QHch=γchm+βchm·EHh-∑c´∈CPQc´·γc´hmPQc
where QHch is the amount of commodity c consumption for household h, γchm denotes the amount of minimum commodity c subsistence consumption for household h, βchm is the consumption expenditures share of commodity c for household h, and EHh represents the disposable household income.
At this stage, the desired shock, i.e., the change of variables or parameters caused by climate change is entered into the model, and the model is solved again to check the situation before and after the shock. The shock parameter introduced as rfcs represents the weather shock on the various activities of the producers of activity a. It was entered into Eq. (7), affecting the other economic parameters and variables (Harris and Robinson 2001; Sassi & Cardaci 2013).17 QVAa=rfcs∙αava(∑fδfava∙QFfa-ρava)-1ρava
Elasticity
To calculate the elasticities, which are important information required in the CGE, other studies such as Reinert and Roland-Holst (1992), Salami (1998), and Javanbakht (2010) have been used.
Table 9 Required elasticities based on production activities
Activity Substitution elasticity of the factors Transfer elasticity between commodities of each activity
Crop activity 0.50 0.40
Horticulture activity 0.50 0.40
Other agriculture activity 0.50 0.40
Food industry activity 0.40 0.40
Other industries and mines activity 0.40 0.40
Energy activity 0.50 0.40
Transportation activity 0.75 0.40
Services activity 0.75 0.40
Source: Reinert and Roland-Holst (1992); Salami (1998); Javanbakht (2010)
Table 10 Required elasticities based on commodities and services
Commodities Armington elasticity of intermediate inputs Armington elasticity of capital commodities Armington elasticity of household consumer commodities Elasticity of export demand
Wheat commodity 1.05 0 0.003 − 5
Rice commodity 1.05 0 0.003 − 5
Other crop commodity 1.05 0 0.003 − 5
Horticulture commodity 1.05 1 0.003 − 5
Other agriculture commodity 1.5 0 0.003 − 5
Food industry commodity 0.5 0 0.5 − 5
Other industries and mines’ commodity 1.06 0.85 1.5 − 5
Energy commodity 1.27 1.27 1.27 − 5
Transportation commodity 0 0 0 − 5
Services commodity 0 0 1.5 − 5
Source: Reinert and Roland-Holst (1992); Salami (1998); Javanbakht (2010)
Data Availability
The image supporting Fig. 1 is publicly available in the Meteorological Organization of Iran, available at http://eamo.ir.
The data supporting Table 2 is obtained using the latest input–output tables published by the Statistical Center of Iran, available at https://www.amar.org.ir.
The data supporting Tables 3 and 4 are obtained using the Meteorological Organization of Iran and the database of the Ministry of Agriculture – Jahad of Iran. Available at https://www.maj.ir/Index and http://eamo.ir.
Code availability
The equations coded in Gomez software are related to this paper and calculated by the authors and are not available to the public.
Declarations
Ethics approval
The authors have no relevant financial or non-financial interests to disclose.
Consent to participate
All authors contributed to the study’s conception and design. Conceptualization: Akram Javadi, Mohammad Ghahremanzadeh, and Maria Sassi. Methodology: Akram Javadi, Mohammad Ghahremanzadeh, and Maria Sassi. Data collection and analysis: Akram Javadi, Ozra Javanbakht, and Boballah Hayati. The first draft of the manuscript was written by Akram Javadi, and review and editing was performed by Mohammad Gahremanzadeh. All authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.
Consent for publication
All authors agree to publish this article in your journal.
1 The exchange rate is fixed. 1 U.S. Dollar = 11,000 Iranian Rials (based on 2011 rate).
2 Constant Elasticity of Substitution.
3 Constant Elasticity of Transformation.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
Alcamo J Dronin N Endejan M Golubev G Kirilenko A A new assessment of climate change impacts on food production shortfalls and water availability in Russia Global Environ Change 2007 17 429 444 10.1016/j.gloenvcha.2006.12.006
Chakrabarty M (2016) Climate change and food security in India. Observer research foundation, Website: www.orfonline.org/2019
Doorenbos J, Kassam AH (1979) Yield response to water. FAO Irrigation and Drainage Paper 33. Rome, p 193
Eboh E, Oduh M, Ujah O (2012) Drivers and sustainability of agricultural growth in Nigeria. AIAE (African Institute for Applied Economics). Research paper 8, First Published, January 2012
Eslami A The impact of climate change on agricultural production and food security J Water Sustain Dev 2020 7 4 83 87
Estrada CGF Conde C Eakin H Villers L Potential impacts of climate change on agriculture: a case study of coffee production in Veracruz, Mexico Clim Change 2006 79 259 288 10.1007/s10584-006-9066-x
Evans EM, Lee DR, Boisvert RN, Arce B, Steenhuis TS, Prano M, Poats SV (2003) Achieving efficiency and equity in irrigation management: an optimization model of the El Angel Watershed, Carchi. Ecuador Agric Syst 77(1):1–22. 10.1016/S0308-521X(02)00052-5
Food and agriculture organization of united nations (2015) Climate Change and Food Security, risks and responses
GhaffariEsmaeli SM Akbari A KashiriKalaei F The impact of climate change on the economic growth of Iran’s agricultural sector: CGE model approach Iran J Agric Econ Dev 2018 32 4 333 342 10.22067/JEAD2.V32I4.69897
Goodbody S, Pound J, Bonifacio R (2012) FAO/WFP crop and food security assessment mission to South Sudan. Food and Agriculture Organisation (FAO) of the United Nations and World Food Programme (WFP), Rome
Gouel Ch, Laborde D (2021) The crucial role of domestic and international market-mediated adaptation to climate change. Journal of Environmental Economics and Management 106. 10.1016/j.jeem.2020.102408
Harris RL, Robinson S (2001) Economy-wide effects of El Nino/Southern Oscillation (ENSO) in Mexico and the role of improved forecasting and technological change. TDM Discussion Papers, 83. International Food Policy Research Institute (IFPRI), Washington, DC
Hatfield JL, Prueger JH (2015) Temperature extremes: effect on plant growth and development, weather and climate extremes 10: 4–10. 10.1016/j.wace.2015.08.001
Hayse J W (2000) Using Monte Carlo analysis in ecological risk assessments. Argonne National Laboratory, Supported by the U.S. Department of the Navy Under Contract W-31–109-ENG-38. http://web.ad.anl.gov/ecorisk/issue/pdf/montecarlo.pdf/2019
Hosseini SS Pakravan MR Etghayi M Effects of agriculture sector total support estimate on food security in Iran J Econ Res Agric Dev Iran 2014 44 4 533 544 10.22059/IJAEDR.2013.50957
Hosseini SS, Nazari MR (2015) Assessing the economic vulnerability of the country's agricultural sector to climate change. National climate change plan, third national climate change report, Environmental Protection Organizations in Iran. July 2015
Intergovernmental Panel on Climate Change (IPCC) (2007) Available on http://www.ipcc-data.org/2009
Jafari S Bakhshi Dastjerdi R MosaviMohseni R Studying the effects of non-oil exports on targeted economic growth in Iranian 5th development plan: a computable general equilibrium approach Iran J Econ Stud 2014 3 1 111 130 10.22099/ijes.2014.3114
Javanbakht O (2010) The effectiveness of the growth of the agricultural sector and other economic sectors of Iran from reducing the interest rate of facilities and increasing the supply of facilities: the approach of CGE model, Ph.D. Thesis, Agriculture and Natural Resources University of Tehran, Iran, 2010
Karaky R (2002) Climate variability and agricultural policy in Morocco. PhD Thesis. Department of Agricultural Economics, Purdue University, West Lafayette
Khaleghi S, Bazazan F, Madani Sh (2015) Effects of climate change on agricultural production and Iranian economy: social accounting matrix approach. Iranian Journal of Agricultural Economics Research 7(1):113–135
Khalilian S Shemshadi K Mortazavi SA Ahmadi M Investigating welfare effect of climate change on the wheat product in Iran Iran J Agric Econ Dev 2014 28 3 292 300 10.22067/JEAD2.V0I0.35472
Kogo BK Kumar L Koech R Climate change and variability in Kenya: a review of impacts on agriculturE AND FOOD SECURITY Environ Dev Sustain 2020 23 23 43 10.1007/s10668-020-00589-1
Lizumi T Ramankutty N How do weather and climate influence cropping area and intensity? J Global Food Secur 2015 4 46 50 10.1016/j.gfs.2014.11.003
Lofgren H, Harris R L, Robinson S (2002) A standard computable general equilibrium (CGE) model in GAMS. Microcomputers in Policy Research, 5. International Food Policy Research Institute (IFPRI), Washington, DC
Mendelsohn R The impact of climate change on agriculture in developing countries J Na Resour Policy Res 2009 1 5 19 10.1080/19390450802495882
Ministry of Roads and Urban (2020) Meteorological Organization of Iran, Weather and climate information section, http://eamo.ir
Ministry of Agriculture-Jahad of Iran (2020) Deputy of planning and economy. The Center of Information and Communication Technology, https://www.maj.ir/Index
Nelson GC, Rosegrant MW, Koo J, et al. (2009) Climate change, impact on agriculture and costs of adaptation. International Food Policy Research Institute (IFPRI), Washington, DC. 10.2499/0896295354
Pakravan MR Hoseini SS Salami H Yazdani S Identifying effective factors on food security of Iranian rural and urban household Iran J Agric Econ Dev Res 2015 46 3 408 395 10.22059/IJAEDR.2015.55514
Panahi M, Gitipajoh M, Ehsani A, Fathi A (2015) Investigating the impact of climate change on rice yield using the SWAP model. The fifth regional climate change conference, Tehran
Parhizkari A Mozaffari MM Hoseini M Economic analysis of climate change effects on wheat yield in Shahrood watershed J Agric Nat Resour 2014 18 88 100
Pawlak K Kołodziejczak M The role of agriculture in ensuring food security in developing countries: considerations in the context of the problem of sustainable food production Sustainability 2020 12 13 5488 10.3390/su12135488
Rademacher-Schulz C, Afifi T, Warner K, Rosenfeld T, Milan A, Etzold B, Sakdapolrak P (2012) Rainfall variability, food security and human mobility. An approach for generating empirical evidence. Intersections, 10. United Nations University, Institute for Environment and Human Security (UNU-EHS), Bonn
Raj S, Roodbar S, Brinkley C, Wolfe D W (2022) Food security and climate change: differences in impacts and adaptation strategies for rural communities in the global south and north, Review article, Front. Sustain. Food Syst, Sec. Climate-Smart Food Systems. 10.3389/fsufs.2021.691191
Raza A, Razzaq A, Mehmood S S, Zou X, Zhang X, Lv Y, Xu J (2019) Impact of climate change on crops adaptation and strategies to tackle its outcome: a review 8, 2,34. 10.3390/plants8020034
Reinert K Roland-Holst D Armington elasticities for United States manufacturing sectors J Policy Model 1992 14 5 631 639 10.1016/0161-8938(92)90033-9
Rodrigues GC Luis S Pereira LS Assessing economic impacts of deficit irrigation as related to water productivity and water costs Biosys Eng 2009 103 536 551 10.1016/j.biosystemseng.2009.05.002
Salami H Application of a general dynamic econometric model in studying agricultural production structure: a case study of crop production in Iran J Agric Sci Technol 1998 23 205 222
Sassi M Cardaci A Impact of rainfall pattern on cereal market and food security in Sudan: stochastic approach and CGE model Food Policy 2013 43 321 331 10.1016/j.foodpol.2013.06.002
Vaseghi A, Esmaili A (2008) Investigation of the economic impacts of climate change on Iran agriculture: a Ricardian approach (case study: wheat). Journal of Agricultural Science and Technology and Natural Resources, 12 (45): 685–696
Winston HY, Alam M, Hassan A, Khan A, Ruane AC, Rozenzweig A, Major DC, Thurlow J (2010) Climate change risk and food security in Bangladesh.WorldBank, 2010
Yadav Sh, Hegde V, Habibi A, Dia M, Verma S (2018) Climate change, agriculture and food security, Available on Wiley Online Library, 10.1002/9781119180661.ch1
| 36467860 | PMC9703425 | NO-CC CODE | 2022-11-29 23:21:09 | no | Theor Appl Climatol. 2022 Nov 28;:1-19 | utf-8 | Theor Appl Climatol | 2,022 | 10.1007/s00704-022-04289-w | oa_other |
==== Front
Comput Math Organ Theory
Comput Math Organ Theory
Computational and Mathematical Organization Theory
1381-298X
1572-9346
Springer US New York
9370
10.1007/s10588-022-09370-3
OriginalPaper
Vaccination trials on hold: malicious and low credibility content on Twitter during the AstraZeneca COVID-19 vaccine development
http://orcid.org/0000-0002-0327-3819
Horawalavithana Sameera [email protected]
1
De Silva Ravindu [email protected]
2
Weerasekara Nipuna [email protected]
2
Kin Wai N G [email protected]
1
Nabeel Mohamed [email protected]
3
Abayaratna Buddhini [email protected]
2
Elvitigala Charitha [email protected]
2
Wijesekera Primal [email protected]
4
Iamnitchi Adriana [email protected]
5
1 grid.170693.a 0000 0001 2353 285X CSE, University of South Florida, Tampa, USA
2 SCoRe Lab, Colombo, Sri Lanka
3 grid.452146.0 0000 0004 1789 3191 Qatar Computing Research Institute, Ar Rayyān, Qatar
4 grid.47840.3f 0000 0001 2181 7878 ICSI, University of California, Berkeley, Berkeley, USA
5 grid.5012.6 0000 0001 0481 6099 Institute of Data Science, Maastricht University, Maastricht, Netherlands
28 11 2022
122
© The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2022, Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
The development of COVID-19 vaccines during the global pandemic that started in 2020 was marked by uncertainty and misinformation reflected also on social media. This paper provides a quantitative evaluation of the Uniform Resource Locators (URLs) shared on Twitter around the clinical trials of the AstraZeneca vaccine and their temporary interruption in September 2020. We analyzed URLs cited in Twitter messages before and after the temporary interruption of the vaccine development on September 9, 2020 to investigate the presence of low credibility and malicious information. We show that the halt of the AstraZeneca clinical trials prompted tweets that cast doubt, fear and vaccine opposition. We discovered a strong presence of URLs from low credibility or malicious websites, as classified by independent fact-checking organizations or identified by web hosting infrastructure features. Moreover, we identified what appears to be coordinated operations to artificially promote some of these URLs hosted on malicious websites.
Keywords
AstraZeneca vaccine
Coordinated URL promotion
Misinformation
http://dx.doi.org/10.13039/100000185 Defense Advanced Research Projects Agency FA8650-18-C-7825 Iamnitchi Adriana
==== Body
pmcIntroduction
Throughout the COVID-19 pandemic that started in early 2020, social media platforms have played a critical role in widely spreading information, regardless of its veracity (Rosenberg et al. 2020). As COVID-19 vaccines moved through development, a significant surge in misinformation and anti-vaccine narratives became evident on Twitter (Smith et al. 2020; Bagherpour 2020). Twitter discussions about AstraZeneca, a vaccine developed at Oxford University, were no exception to this as the vaccine has been mired in controversy ever since its inception (Jemielniak and Krempovych 2021).
AstraZeneca’s vaccine development faced numerous challenges due to a temporary halt of its trials in September 2020 caused by an unexplained illness in one of the participants (Reuters 2021; Robbins et al. 2020). Details surrounding the event were not thoroughly disclosed, which lead scientists to question the transparency of the vaccine development efforts (Cyranoski and Mallapaty 2020). The temporary interruption of the AstraZeneca vaccine trials also sparked political debates as government officials, especially in the United States, had been trying to fast-track its approval and roll-outs (Forbes 2020). As AstraZeneca quickly resumed its clinical trials and entered its final phase of development, safety issues and concerns about its efficacy intensified (Cyranoski and Mallapaty 2020; PBS 2020).
This paper investigates the information shared on Twitter in a period that marked an important step in the development of the AstraZeneca vaccine. We ask several research questions in this study. To what extent is the presence of low credibility and malicious information about Astrazeneca vaccine related discussions in Twitter due to its global halt/suspension? What are the properties of the URLs that share low credibility and malicious information? Is there any coordinated behavior for promoting bad quality information, if so, how persistent is such behavior?
We focus on two time periods, before and after the temporary interruption of the vaccine development on September 9, 2020. While our focus in on AstraZeneca-related discussions, we also detected tweets that mentioned SputnikV, Moderna and Pfizer vaccines with stories that cast doubt and fear towards the safety and efficacy of the vaccines. A deeper analysis shows that vaccine-related stories are promoted by both low credibility and malware-hosted websites. In contrast to the low credibility websites that often promote vaccination myths and conspiracy stories, the malware hosted on other websites can be used to trace the web searching activities of the Twitter audience interested in COVID-19 vaccines. For example, the properties of the most popular malicious URLs in two time periods are quite similar, such as short-lived newly created domains, self-signed certificates, content management system and hosting geographical location.
Some user accounts use this information to target specific communities with stories promoting certain vaccines favorably, while questioning others. For example, Russian-sponsored media outlets took advantage of the halt of trials to push narratives discrediting the AstraZeneca vaccine and boost the promotion of SputnikV instead. We discovered persistent groups of users engaged in the rapid propagation and artificial inflation of a particular URL through simultaneous tweeting. This behavior suggests potential signs of coordination to promote malicious and low-credibility URLs on Twitter.
Related work
Along factual information, misinformation and disinformation related to COVID-19 regularly circulate in social media (Huang and Carley 2020). As vaccines became available, (mis)disinformation surrounding them started to increase and persist on Twitter (DeVerna et al. 2021). Combating the spread of (mis)disinformation and conspiracy theories regarding COVID-19 on social media has become a global concern as, even in the absence of any scientific evidence, it has increased the number of people hesitant to get vaccinated (Wilson and Wiysonge 2020; Chadwick et al. 2021).
Numerous research efforts aimed to assess the prevalence of misinformation related to COVID-19 in social media. Pranesh et al. (2021) quantified the magnitude of misinformation presence in COVID-19 related tweets across different languages. Karami et al. (2021) showed that vaccine discussions on Twitter are evolving with negative and non-negative attitudes on different issues. Their analysis revealed that tweets on topics related to vaccination sites, getting vaccines, and vaccine effectiveness exhibit positive sentiment, while tweets about government strategies related to vaccination, vaccination hesitancy, vaccine immunity, and social distancing often sparked negative discussions. Yang et al. (2020) and Sharma et al. (2022) investigate URLs embedded in tweets to evaluate the credibility of the content at domain level. These studies highlight that there is a comparable presence of both low-quality sources and sources from mainstream media in Twitter. In addition, they found that social bots are most likely to engage with the promotion and amplification of low credibility information.
Other studies analyzed the extent to which different users engage with misinformation. Teng et al. (2022) highlight that social bots’ contribution to misinformation is surprisingly low. The authors identified two user groups who engage with misinformation: the strong-adherent users, who aim at supplying or/and promoting misinformation for certain purposes, and the weak-adherent individuals, who are occasionally exposed to misinformation but are easily triggered to further spread misinformation. Echoing the same sentiment, Silva et al. (2020) showed that the significant majority of both factual and misinformation tweets were generated by real users (not bots) even though bot accounts tweeted more misinformation (8.5%) relative to factual information (5.2%). Contrary to these findings, Yang et al. (2020) observed a higher-than-normal participation rate of bot accounts in both posting and amplifying low-credibility content. Their results also show that bot-like tweeters attract more bot-like retweeters than human-like tweeters.
This paper builds on our previous analysis of Twitter discussions during a period of uncertainty in the AstraZeneca vaccine development (Horawalavithana et al. 2021). In contrast to many previous research, we characterize low quality content via both malicious and low credibility URL analysis to identify tweets which escalate community concerns with regards to vaccine efficacy. We also analyze the coordinated link-sharing behavior by revealing groups of users who share the same URL within unusually short time.
Data collection and processing
The mistrust in what was perceived as rushed vaccine development during the COVID-19 pandemic has been reported to undermine people’s intention to receive the COVID-19 vaccines (Vivek 2021). In addition, concerns related to the side effects reported in vaccine clinical trials raised questions about their safety (Nuzhath et al. 2020). One such major incident was reported in September 2020, when an unexplained illness appeared during the clinical trials for the AstraZeneca vaccine.
We collected tweets using the Twitter API from 1st of September 2020 to 15th of September 2020. The original keywords used to collect this dataset are AstraZeneca, “Astra Zeneca”, AZD1222, COVID, vaccine, immunity, “herd immunity”, Barrington, and “focused protection”. We used the same keyword list used to collect a similar dataset released as part of the 2021 Grand Challenge of the North American Social Network Conference (NASN) (NASN 2021). Our data collection cover many missing tweets in the NASN dataset after September 9th 2021 (Horawalavithana et al. 2021). We used the vaccine-related keywords in the regex condition: pfizer OR astrazeneca OR moderna OR (sputnik AND vaccine) to select tweets relevant to our study.
We grouped the Twitter messages into two disjoint subsets based on their inclusion of URLs. The first subset contains 3,212,586 Twitter messages (by 1,788,788 users) without URLs. The second subset consists of 3,958,864 messages (55% of all messages) that contain at least one URL. These messages are shared by 1,267,873 users and cite 3,247,946 distinct URLs from 167,186 distinct web domains. The number of such messages per day are presented in Fig. 1.
We pre-processed this dataset to identify hashtags, user mentions, and URL domains. We eliminated the URLs which link to other tweets. In addition, the external links (e.g., a tweet mentioning a YouTube video, or an external website domain) mentioned in messages are pre-processed as follows: The shortened URLs are expanded, and HTML parameters are removed from the URLs. The YouTube URLs are resolved to the base video URLs if they include a parameter referencing a specific time in the video. We represent the URLs by the parent domain when multiple child domains exist (e.g., fr.sputniknews.com, arabic.sputniknews.com, etc., are renamed as sputniknews.com). This pre-processing code of resolving URLs is publicly available.1Fig. 1 Number of Twitter messages over time. AstraZeneca vaccination trial halted date is highlighted in the green dashed vertical line
Figure 1 shows spikes in Twitter activity on September 3rd, 9th and 15th, 2020. On September 3rd, the most popular URLs point to a Twitter event on Dwayne Johnson, an American actor, and his family testing positive for COVID-19 and the URL is shared 1756 times.2 On September 9th, the most popular URLs point to mainstream news articles, with an article published on statnews.com receiving the highest number (1223) of shares.3 This article reports the halt in AstraZeneca vaccine trials in response to a potentially harmful reaction of a trial participant. We observe another spike on September 15th, where the majority of tweets (574) cite an article related to COVID re-infection.4 There are some other tweets (354) citing an article written about a whistle-blower named Dr. Yan who released a report suggesting coronavirus was ‘Lab Modified’.5
We consider two periods in our analysis, before (September 1–8) and after (September 9–15) the temporary interruption of the vaccine development on September 9, 2020. While the majority of the tweets mention AstraZeneca as expected (Fig. 2), we also detect tweets that mention Pfizer, SputnikV, and Moderna vaccines. When comparing the two periods, there is a significant increase in AstraZeneca mentioned tweets (363.7% increase) in the second period. Additionally, we observed that Pfizer mentioned tweets also increased significantly by 149.1%. We observed that SputnikV and Moderna vaccines mentioned tweets decreased by 50.2% and 57.5% respectively. There is also a significant increase (149.36%) of tweets mentioning both Pfizer and AstraZeneca vaccines.Fig. 2 Number of tweets with vaccine mentions
Table 1 Top-10 topics identified from Tweet texts ( September 1–8)
Topic label Top 10 words # Tweets
Vaccines Vaccin, antivax, polio, smallpox, hpv, pox, vacc, vax, chickenpox, typhoid 41,634
Covid Safety Aye, awhil, suscept, exagger, blindli, isnt, huh, arent, diplomaci, ahem 27,520
Politics Trump, djt, trumpster, drumpf, maga, trumpism, donald, trumper, libtard, honestli 25,641
Covid Deaths Deaths, death, die, mortem, dead, lethal, rip, tue, morbid, fatal 22,972
Covid Related News Amp, vox, rrb, rha, interf, wire, slash, tc, pedal, tele 22,896
Facemasks Mask, facemask, unmask, helmet, helm, wig, visor, bandana, hide, conceal 22,837
Donald Trump Vaccin, antivax, trump, djt, drumpf, trumpster, trumpism, barron, kushner, hpv 20,994
Child Safety School, schoolchildren, preschool, classroom, homeschool, teacher, uncov, kindergarten, isd, teach 13,173
Covid Tests Test, tester, retest, assay, exam, swab, trial, checkup, appt, dmv 12,914
Flu Virus Flu, influenza, flue, sars, sicker, swine, cdc, vaccin, ill, ebola 11,325
Table 2 Top-10 topics identified from Tweet texts (September 9–15)
Topic label Top 10 words # Tweets
Facemasks Mask, facemask, unmask, helmet, helm, safeguard, conceal, hide, visor, impun 29,824
Vaccines Vaccin, antivax, polio, smallpox, hpv, narrow, vax, pox, cdc, bcg 25,316
Covid Safety aye, awhil, arent, bhi, diplomaci, isnt, exagger, webpag, meh, blindli 22,205
Covid Related News Amp, vox, rha, interf, wire, circuit, ion, tele, pedal, slash 20,691
Donald Trump Trump, djt, trumpster, drumpf, maga, trumper, donald, trumpism, lyin, bluster 20,172
Covid Deaths Deaths, death, mortem, die, lethal, dead, fatal, morbid, rip, tue 19,759
Vaccines Vaccin, antivax, polio, smallpox, pox, hpv, chickenpox, vax, typhoid, cdc 18,073
Unemployment due to Covid Work, jobless, quit, iam, job, awhil, ive, dole, newsengin, workload 14,881
Covid Safety Aye, exagger, weve, plz, blindli, let, suscept, breweri, section, about 13,000
Politics Biden, vp, joe, djt, trumpster, trump, obama, msnbc, sotu, mccain 12,537
We used a pretrained embedding model universal-sentence-encoder6 provided in the Top2Vec package (Angelov 2020) to identify topics in the tweets dataset from the above mentioned two time frames. Using the default parameters, we obtained 2,527 and 2,225 topics before and after the vaccine development halt, respectively. We selected the top 10 topics from most occurring keywords from both time periods and performed a qualitative analysis (as shown in Tables 1 and 2). We noticed that face mask-related topics are prevalent in the discussions after the vaccine development halt. Twitter users have discussed about the usability and efficacy of the facemasks and the importance of wearing one (see Appendix Tables 7 and 8 for sample tweets). Other topics include COVID-19 related deaths, safety, and testing. In addition, we identified tweets discussing unemployment due to COVID-19 or due to shutdown of businesses (see Appendix Table 9 for sample tweets).
URL analysis
We analyze the URLs cited in the tweets to identify the extent of sharing poor quality information that were originated from outside of the platform. To this end, we group the URLs into low credibility URLs (Sect. 4.1) and malicious URLs (Sect. 4.2).
Low credibility information sources
We aim to investigate how Twitter users react to low credibility information sources across the two different periods. We grouped the web domains according to the classification made by two sources (i.e., Media Bias/Fact Check (MBFC 2020) and the Factual (Factual 2020). We considered as low credibility those web domains that were in at least one of the following categories: questionable sources, conspiracy-pseudoscience, or (very)low credibility rating. We identified 14,215 (2%) URLs from 377 low credibility information sources that are shared in 42,271 (2.7%) messages posted before the halt of trials, and 12,768 URLs from 363 low credibility sources shared in 36,906 (2.7%) messages posted after the halt of trials. Table 3 shows the top 10 most popular domains by number of tweets from each time frame of interest.
thegatewaypundit.com and rt.com are the most popular low credibility web domains by number of mentions (4326 and 3626) and number of engaged users (1947 and 1971) despite publishing fewer URLs/articles than other domains in the period corresponding to before the halt. These two web sources are known for spreading propaganda and promoting conspiracy theories in their articles. After the halting of the AstraZeneca trials, the most shared low credibility web domains were rt.com and zerohedge.com. Both domains are classified with low to very low credibility due to the promotion of pseudoscience misinformation. During both periods, we also observed that sputniknews.com made it to the top-10 of low credibility domains shared in Twitter. Similar to other Russian state-sponsored outlets in the list, sputniknews.com exploited the halt of the AstraZeneca trials to push narratives discrediting the vaccine and boost the promotion of SputnikV instead.
The list of most popular low credibility web domains did not changed drastically across the two distinct time periods, but their rankings did (as shown in Fig. 3a and b).Table 3 Twitter sharing characteristics for low credibility domains as identified by MBFC
Period Domain # Tweets # Users # URLs Overall rank
September 1–8 thegatewaypundit.com 4326 1947 358 27
rt.com 3626 1971 687 31
zerohedge.com 2948 1815 237 36
dailymail.co.uk 2548 1392 1122 45
foxnews.com 2281 1196 622 52
granma.cu 1536 486 192 86
westernjournal.com 1459 876 277 93
fr24news.com 1414 21 1407 96
sputniknews.com 1363 602 649 99
news18.com 1070 273 660 129
September 9–15 rt.com 2972 1478 646 36
zerohedge.com 2930 1692 237 39
dailymail.co.uk 2485 1411 1039 49
foxnews.com 2273 1390 616 56
granma.cu 1864 518 238 72
thegatewaypundit.com 1589 902 208 83
sputniknews.com 1509 648 698 91
fr24news.com 1188 15 1184 108
news18.com 879 270 541 137
westernjournal.com 727 493 177 169
We also include the overall ranking of the low credibility domains out of all domains in the dataset
Fig. 3 Twitter sharing characteristics of most popular domains. a and b Show the Top-10 domains by the number of distinct URLs in each time frame. The size of the markers in this plot are proportional to the number of URLs associated with the domain
We also noted some cases in which tweets citing the same URL often share the same article heading. These users promoted certain topics through massive repetition of messages via injecting URLs. For example, an article published in zerohedge.com was in the Top-10 most popular URLs on the day when the AstraZeneza vaccine development halted.7 However, this article tried to build an alternative frame highlighting a statement by the US House Speaker Nancy Pelosi about the issue instead of reporting the details of the main event.
Malicious URLs
We used VirusTotal (VT) (VirusTotal 2021) to extract the maliciousness of URLs. VT provides the state-of-the-art aggregated intelligence for domains and URLs, and relies on more than 70 third-party updated antivirus (AV) engines. For all distinct URLs in our collection, we extracted VT scan reports via querying the publicly available API. Each VT scan report contains of the verdict from every AV engine, information related to the URL such as first and last seen dates of the URL in the VT system, hosting IP address, final redirected URL (if applicable), content length, etc. Each AV engine in a VT report detects if the URL is malicious or not. In order to indicate the maliciousness of a URL, we looked at how many engines flag it as such.
In this study, we labelled a URL as malicious if at least one AV engine detects it as malicious. Such malicious URLs, in general, are either phishing websites that steal user credentials and/or personally identifiable information from victims or malware hosting websites that attempt to install malware on victims’ devices. Before suspending the trial, we observed that 35.9% of the malicious URLs utilize URL shortening services with top 4 services being bit.ly, tinyurl.com, ow.ly and goo.su whereas as only 20.2% of benign URLs utilize such services. We noticed rather different proportions after the trail. 25.9% of the malicious URLs utilize URL shortening services with top 4 services being bit.ly, ow.ly, hubs.ly, and rb.gy whereas only 7.2% of benign URLs utilize such services. This observation is consistent with the trend that malicious actors are increasingly using URL shortening services to camouflage malicious URLs to present non-suspicious-looking URLs to users (FAS 2020). We found that 30.80% and 40.66% of the domains related to malicious URLs are ranked below 100K by Alexa Amazon (2021) before and after the halt respectively (the lower the rank value, the higher the popularity). This indicates the alarming reality that malicious actors are able to reach a large user base reaping a high return on investment for their attacks.Table 4 Details of the URL hosting and lexical features
Feature Description Type
VT_Dur URL duration in VirusTotal (VT) Hosting
PDNS_Dur Domain duration in Passive Domain Name Resolution (PDNS) Hosting
#IPs # hosting IPs Hosting
#Queries # times the domain is accessed Hosting
#NSes # Name servers Hosting
Is_NS Do the apexes of the domain name and NS domain name match? Hosting
#SOAs # administrative domains Hosting
Is_SOA Do the apexes of the domain name and admin domain match? Hosting
#Domains # domains hosted on the IP Hosting
#Queries_IP # times the IP is accessed Hosting
ASN Autonomous System Number Hosting
Org Organization owning the Autonomous System Numbers (ASN) Hosting
Geo Geographic location of the ASN of particular IP Hosting
Server Web server used for hosting Hosting
Minus The number of dashes appear in the fully qualified domain name (FQDN) Lexical
Suspicious_TLD Does the domain name include a suspicious country code top-level domain (ccTLD) Lexical
Fake_TLD Does the domain name include a fake gTLD (com, edu, net, org, gov)? Lexical
Brand Does it impersonate a popular Alexa top 1000 brand? Lexical
Pop_Keywords Does the domain name include popular keywords Lexical
URL length The length of the URL Lexical
Is_IDN Is internationalised domain name? Lexical
NS_Domain Is Name Server Domain? Lexical
Fig. 4 Malicious URL clusters based on the lexical and hosting features. Each point is a URL, and it is colored according to the cluster it belongs
We further analyzed the malicious URLs to identify related malicious URLs. To this end, based on the lexical features in the literature Silva et al. (2021) and the hosting features mentioned in Table 4, we clustered the malicious URLs using PCA/OPTICS algorithm. While lexical features identify characteristics related to URLs themselves, hosting features, extracted from Farsight Passive DNS (PDNS) data (Farsight Security 2021), capture the characteristics of underlying hosting infrastructure. As shown in Fig. 4, these features collectively identify 4 distinct malicious URL clusters. We manually verified the accuracy of the top 2 clusters by checking the web page content, registration information and domain certificate information. The clusters observed in the two time frame are quite similar in their properties such as short-lived newly created domains, self-signed certificates, content management system (CMS) technology and hosting geographical location. We further analyzed the clusters based on the maliciousness of URLs. The maliciousness of a URL can loosely be measured by #VT, the number of VT positives. An interesting observation is that URLs belonging to different maliciousness levels share similar lexical and hosting features. We further analyzed these malicious URLs in terms of where they are hosted. Unexpectedly, we found that 80.04% of these malicious URLs are hosted in content delivery networks (CDNs) such as Cloudflare and Akamai. While CDNs provide fast delivery of content across the globe through their distributed computing infrastructure, we believe a key reason why malicious actors utilize such services is to improve attack agility and stay below the radar of malicious domain detection mechanisms in place. This observation is further reinforced with the increased utilization of public cloud computing infrastructure (33.5% of all malicious URLs) sharing hosting IPs with tens of thousands of unrelated domains, which are mostly benign. Such shared IPs are usually not blocked in practice due to the collateral damage.
Coordinated URL sharing behavior
Previous work on detecting coordination has focused on accounts who consistently amplify/boost sources of information (e.g., users who co-retweet the same tweets) in a social media platform (Keller et al. 2020; Pacheco et al. 2021; Weber and Neumann 2020). In this work, we are interested in those accounts who introduce new information into the platform, specifically in the form of URLs. One of our objectives is to characterize the URL sharing activities in the COVID-19 vaccine related discussions. To this end, we describe the presence of two groups of URLs (i.e., low-credibility and malicious URLs) in Twitter tweets. According to previous research (Pacheco et al. 2021; Kin Wai et al. 2021), low-credibility URLs are often promoted by coordinated groups of users in diverse contexts (e.g., U.S. elections, Hong Kong protests, and the Syrian civil war).
To detect coordinated link-sharing behavior, we employed the methodology proposed in Giglietto et al. (2020), where coordination is defined as “different users who repeatedly share the same URLs in an unusually short period of time.” The computation of this time threshold is based on the analysis of inter-arrival times between tweets of the same URL for each period of interest. Particularly, the timing behavior of the top quickest shared URLs, as identified based on the time differences between consecutive tweets, is further explored. The desired threshold is computed by calculating the median time that it takes these top URLs to reach a certain proportion of their total shares. As can be seen, the inference of this threshold relies on three parameters: (1) the percentage of top URLs for the analysis, (2) the number of consecutive tweets to consider in the inter-arrival analysis (e.g., time difference between first and second post, or first and last post), and (3) the proportion of tweets each URL needs to reach in order to compute the median time.
We experimented with different configurations of these parameters to identify a reasonable coordination time threshold, in which simultaneous postings of the same URL would be considered unusual compared to the activity patterns of the entire period. The time threshold is mostly sensitive to both the percentage of quickest URLs and the number of consecutive tweets considered in the time difference analysis. For example, when considering high percentage of URLs (e.g., 25%, 50% or 75%) or a large number of consecutive tweets, the coordination interval would be too long and not sufficiently strict. Stringent time thresholds were found when considering smaller values for these parameters. Specifically, we considered the median time that it takes the 10% quickest URLs, ranked by their time differences between the first and second posts, to reach 50% of their shares. The threshold was 5 s for both periods, before and after the halt of trials.
We compared the coordinated networks that spanned across the two different time periods: September 1–8 and September 9–15, which correspond to Twitter activities from before and after the halt of the AZ trials, respectively. We only considered URLs that were shared at least twice and by different user accounts. From September 1 to September 8, 143,782 URLs were posted in 705,917 tweets. From September 9 to September 10, 128,426 URLs were posted in 643,667 tweets.
For each time period, we constructed the network of timely-coordinated accounts by considering only those pairs of users who post the same URL within the corresponding time interval threshold. The edge weights between users correspond to the number of URLs posted simultaneously within the threshold. In an attempt to reduce the chances of false positives (i.e., simultaneous postings happening by chance), we remove from these networks connections with an edge weight of 1. We acknowledged that additional filtering strategies could reduce even further the amount of false positives in the network. For example, inspecting the time difference between the publication date of an article and the timestamps of simultaneous tweets could reveal instances of organic behavior (e.g., crowds sharing synchronously due to a share-button functionality on websites). Unfortunately, we failed to scrape accurate publication dates for many articles due to web pages being inactive or changes to the original date caused by updates.
Table 5 shows a comparison between the coordinated networks corresponding to each time period we investigate across various network properties. We found that there are no drastic differences in terms of the size, density, and number of unique URLs shared between the two networks across different periods. Most connected components in both periods consist of dyads and triads (89% in the before network and 90% in the after network).Table 5 Basic network properties for the two coordinated networks from before and after the halt of trials
Statistics Before (September 1–8) After (September 9–15)
Number of Nodes 2794 2528
Number of Edges 2998 2745
Number of URLs 8831 7420
Number of Domains 1136 1030
Connected Components 1039 919
Number of Dyads 774 699
Number of Triads 148 124
Fig. 5 Network of coordinated users before (left) and after (right) the halt of AstraZeneca trials. The edge weight represents the number of URLs posted simultaneously between two users. The red edges occur between users who co-shared at least one non-credibility web domain, the blue edges are between users who co-shared credibility web domains, and the gray edges are between users where the credibility of web domains is unknown. Pair of users with edge weight of 1 are removed
To investigate the level of trustworthiness of the news sources present in each coordinated network, we grouped the web domains according to their credibility ratings as mentioned in Sect. 4.1. We identified 385 URLs from 34 low credibility web domains, 2860 URLs from 77 credibility web domains, and 12,895 URLs from 1387 web domains with unknown credibility score. Figure 5a and b show the coordinated networks, from before and after the halt of trials, induced on only those network components that consist of at least one coordinated connection sharing credibility or low credibility domains. That is, we ignore those components in the coordinated network that exclusively shared domains with unknown credibility. The induced network for the period before the halt consists of 508 nodes and 820 edges (out of which 56 are to low credibility sources, which is 2% of the total edges in the original coordinated network). The induced network for the period after the halt consists of 444 nodes and 515 edges (out of which 58 are to low credibility sources, which also represents 2% of the total edges in the non-induced network).
Similarly, we also investigated the level of maliciousness of URLs in the coordinated networks by inducing on network components with at least one connection to a URL classified as malicious by VT. We identified 623 malicious URLs from 52 different web domains. The induced network for the period before the halt consists of 276 nodes and 632 edges (out of which 188 are to malicious sources, which is 6% of the total edges in the original coordinated network). The induced network for the period after the halt consists of 222 nodes and 273 edges (out of which 130 are to malicious sources, which represents 4.7% of the total edges in the non-induced network).
These observations suggest that the extent to which both low credibility and malicious sources are promoted by coordinated groups of users is relatively similar across the two time frames. The number of coordinated nodes that overlap between the two different periods is 314 for the induced networks by credibility and 147 for the induced networks by maliciousness, which is more than half of the users in both cases. This highlights that groups of coordinated users, who promote either low credibility or malicious URLs, are persistent across the two different periods we study.
Table 6 shows the top-10 most popular low credibility sources shared by coordinated users across the two periods of interest. We found that, in both time frames, the most popular sources were zerohedge.com, news18.com, and sputniknews.com. The first is a website in the conspiracy-pseudoscience category, and the last two are classified as questionable sources with a poor fact-checking record. We observed a high presence of Russian web sources among the low credibility domains in the coordinated networks. Some are well-known domains such as sputniknews.com and rt.com, and others less popular such as cnnn.ru and inosminews.ru, which we found are news aggregators that often re-publish articles from the first two. We noticed that web domains shared by coordinated groups remained relatively similar in terms of their popularity across the two different periods of interest. Finally, we found that only a small number of coordinately shared URLs (26) from two web domains are classified as both low credibility and malicious: rt.com and tmz.com.Table 6 Twitter sharing characteristics for low credibility URLs posted by users found in the coordinated networks from before and after the halt of AstraZeneca trials
Period Domain # Tweets # Users # URLs
September 1–8 zerohedge.com 152 12 47
news18.com 89 6 24
sputniknews.com 44 6 21
westernjournal.com 33 3 16
dailymail.co.uk 31 13 15
yc.news 28 4 7
theepochtimes.com 26 3 13
rt.com 21 8 8
breitbart.com 18 3 9
foxnews.com 17 4 7
September 9–15 zerohedge.com 198 11 19
news18.com 68 6 19
sputniknews.com 60 6 30
foxnews.com 37 6 14
theepochtimes.com 34 5 16
rt.com 28 9 9
westernjournal.com 23 5 11
dailymail.co.uk 22 9 11
breitbart.com 22 6 9
yc.news 20 4 5
Conclusions
In times of crisis, whether political or health-related, online disinformation is amplified by social media promotion of alternative media outlets (Horawalavithana et al. 2020; 2021). This study adds to the growing body of work (Ferrara et al. 2020) that investigates the misinformation activity during the COVID-19 crisis by analyzing a Twitter dataset collected between September 1 and 15, 2020. This period covers events related to the AstraZeneca vaccine development phase trials. Our contributions complement previous observations (Horawalavithana et al. 2021; Singh et al. 2020) in multiple ways.
First, we found a significant increase of AstraZeneca vaccine mentioned tweets in the period following the halt of AstraZeneca vaccine trials. These messages also contain topics related to the usability and efficacy of the facemasks, vaccine safety and COVID related deaths. We also detect tweets that mention SputnikV, Moderna and Pfizer vaccines with the stories that aimed at casting doubts and fear towards the safety and efficacy of COVID-19 vaccines. This vaccination trial event was also linked to multiple other COVID-19 stories. For example, politicized discussions concerning the decisions by government officials to roll-out prematurely the vaccines were also prominent. On the other hand, topics related to unemployment due to COVID-19 or due to shutdown of businesses were popular after the vaccination trial event.
Second, we discover a strong presence of malicious and low-credibility information sources shared on Twitter messages. Not only URLs from low-credibility sources, as classified by independent fact-checking organizations, were present in the dataset, but many of them pointed to pages with malicious code. We found that a significant portion of these low-credibility and malicious URLs (36%) used URL shortening services to a greater extent than non-malicious URLs (<1%). In addition, they were usually hosted on well-established and reputable content delivery networks in an attempt, we believe, to avoid detection.
Third, we found similar properties of the malicious URLs shared before and after the vaccine development halt event. For example, most popular malicious URLs in two time periods are quite similar in their properties such as short-lived newly created domains, self-signed certificates, content management system and hosting geographical location. Moreover, URLs belonging to different maliciousness levels share similar lexical and hosting features.
Finally, we discovered potential signs of coordination to promote malicious and low-credibility URLs on Twitter. Specifically, we identified groups of users who potentially engage in the rapid propagation and artificial inflation of a particular URL through simultaneous tweeting. Our observations suggest that the extent to which both low credibility and malicious sources are promoted by coordinated groups of users is relatively similar across the two time periods before and after the vaccine development halt event. The code and the dataset used in this work are publicly available (De Silva et al. 2022).
Our analysis is useful for multiple stakeholders ranging from individuals, educators, health professionals, journalists, researchers and governments. According to the Surgeon General of the United States, understanding the malign objectives around COVID-19 discussions would help to reduce confusion and mistrust around vaccines and promote public health efforts (Vivek 2021). In this work, we press the needs to extend the definition of COVID-19 related malign content into both low-credibility and malicious content types. This would open new directions for journalists and researchers to broaden health misinformation research done with social media datasets. In addition, they can identify high quality information sources to avoid amplifying malign content. As these malign content can change the perception of general public towards participation in future vaccine trials, we urge health professionals to better inform vulnerable communities. We also show how the malign content spreads across Twitter with potential signs of coordination. This might be helpful for individuals to identify potential coordination campaigns and avoid amplifying malign content unwittingly. Educators can use our findings to share common tactics used by bad actors and use them as evidence to improve educational programs.
Further work is needed to fully comprehend the dubious objectives of bad actors active in times of crisis. For example, bad actors might have chosen this event strategically to maximize the spread of low quality information. These actors can deploy the same strategy in future conversations, thus having content moderation techniques to limit what they can share is important. However, we can only speculate on the motivation behind the use of malware shared along with vaccine-related stories. Bad actors can use this strategy to target specific user communities with the new stories promoting certain vaccines favorably, while questioning others (FAS 2020). On the other hand, the low-credibility news sources might have reported this event opportunistically in an attempt to promote vaccine hesitancy. People might have engaged with these low quality sources to watch out the information space around COVID-19 vaccines. According to (Smith et al. 2020), there is a deficit of high quality information sources to seek vaccine information. Bad actors use this information deficit as an advantage to push low quality information. We believe this analysis can be extended in understanding the role of bad actors during similar emotionally-charged conversations in the future. Another direction is to analyze the change of public perception around COVID-19 vaccines before and after the issues reported in the vaccine trials or the exposure to the news of side effects.
Appendix A Example tweets in Twitter discussions
See Tables 7, 8 and 9.Table 7 An excerpt from tweets discussing face masks (September 1–8)
Tweets
Wear a mask, covid is https://t.co/fCKTDlvDNF
Think twice before wearing a face shield to protect against COVID-19 instead of a cloth face mask—here’s why https://t.co/4DYERlEbd9https://t.co/fJeOtgT7zA
@NBCNews Protect yourself From Covid 19 Use Mask & Facial Shield (see photo).Remember that on the street & in public Transportation People R wearing the mask Incorrectly. U can Get Coronavirus. It’s Ur Health to Protect Yourself. https://t.co/VS3noQKhq5 We R 501(C)(3) Donate Now https://t.co/WqfpP4q8vx
@NTVNewsNL @DonBradshawNTV So if someone gets sick with COVID-19, do they not have to self quarantine now because we are all wearing NON-MEDICAL masks???
Wearing home-made reusable face-cover/mask is essential to stay protected from COVID-19.
@TODAYshow @TodayParents Please share my face mask shop so we can all look good masked up & get back to living MadMaskr https://t.co/f8rz6kRtC3 via @Etsy #facemasks #MaskMandate #TuesdayTips #fashionstyle #fashionblogger #COVID #Corona #MaskUpNOLA #MaskUpAZ #MaskUpMN #Masks #MaskMyAss #TuesdayThoughts
BarackObama Protect yourself From Covid 19 Use Mask & Facial Shield (see photo).Remember that on the street & in public Transportation People R wearing the mask Incorrectly. U can Get Coronavirus. It’s Ur Health to Protect Yourself. https://t.co/VS3noQKhq5 We R 501(C)(3) https://t.co/5qWZZ46MJv
WHO #WearAMask challenge! By wearing a mask, you are sending a message of solidarity & protecting other people, especially those most vulnerable to COVID_19 Take a photo or a video of yourself wearing a mask, share it & nominate friends to do the same https://t.co/b0OunAFcpx 02
Your brain’s powers of facial recognition are going to need some time to get used to the face masks we’re wearing to keep each other healthy https://t.co/GAWbElS0Ec
This Face Mask Doesn’t Stop COVID After All https://t.co/kre8POd3fF
Table 8 An excerpt from tweets discussing facemasks (9–15 September)
Tweets
Do you want to take of your mask, meet your friends, hug your family, go where you want, and take your chances with the covid flu. If you think living in fear wearing a mask is no life for anybody. Support an open and uncensored debate. Sign And Share x https://t.co/kjCiuNvI1E https://t.co/DHtAg9nnDI
Guys remember to wear a mask. covid is https://t.co/obnxQUGnMN
@ewsunionmp @Indersinghsjp @JPNadda @OfficeOfDrNM @OfficeofSSC @BJP4MP @INCMP @narendramodi @OfficeOfKNath @vdsharmabjp @TCGEHLOT Im staying in tamilnadu near thiruporur & i went out for groceries i saw ppl gathering without masks spitting coughing smoking holding hands i feel horible n scary about community spreading of covid shopkeeprs r nt even wearing mask or gloves y dey r behaving like illiterates??
@ElijahSchaffer Please share my face mask shop so we can all look good masked up & get back to living MadMaskr https://t.co/f8rz6kRtC3 via @Etsy #facemasks #MaskMandate #TuesdayTips #fashionstyle #fashionblogger #COVID #Corona #MaskUpNOLA #MaskUpAZ #MaskUpMN #Masks #MaskMyAss #TuesdayThoughts
Wearing a mask could protect you from COVID-19 in more ways than you think https://t.co/JNMzggXx6n
@enews Protect yourself From Covid 19 Use Mask & Facial Shield (see photo).Remember that on the street & in public Transportation People R wearing the mask Incorrectly. U can Get Coronavirus. It’s Ur Health to Protect Yourself. https://t.co/VS3noQKhq5 We R 501(C)(3) Donate Now https://t.co/QKh6sN1zl8
@realDonaldTrump @POTUS Have you observed that just about all the people sitting behind Trump were wearing masks during his rally in Henderson, NV? I guess they are trying to protect him from possibly deadly Covid exhales. Notice a significant space between him and the front row. No masks.
@OfficialSidFC @sardesairajdeep Im staying in tamilnadu & i went out for groceries i saw ppl gathering without masks spitting on road coughing holding hands i feel horible n scary about community spreading of covid shopkeeprs r nt even wearing mask or gloves y dey r behaving like illiterates.r nt they unaware?
@JThakers Im staying in tamilnadu & i went out for groceries i saw ppl gathering without masks spitting on road coughing holding hands i feel horible n scary about community spreading of covid shopkeeprs r nt even wearing mask or gloves y dey r behaving like illiterates.r nt they aware?
@CNBC Test the Mask. The Covid-19 is captured in the masks. YouTube Link: https://t.co/9ju0Pc3gMF
Table 9 An excerpt from tweets discussing unemployment (9–15 September)
Tweets
@SenThomTillis Please do not leave me behind. I’m a 56 year old woman that lost my job because of COVID-19. I’ve worked since I was 19. This pandemic is not over! #SaveThe600 #ExtendPUA #DoYourJob #Extend600 #ExtendUI
@SenSchumer Hello Senator Schumer Im hoping you can help me.I filed for UI June 22 To this day I heard nothing.Over 100 calls and new stories every time.I spoke level 3 reps. Nothing. I worked in a public school,shut down due to COVID. I thought we would be taken care of but I was wrong HELP https://t.co/GHjba95vUv
@LoisWeiss So wait I lost my job during COVID 19 and I only make 200 dollars in UI...am I qualified for 300 dollars?
@halsey @KarriKuzma PLEASE PLEASE HELP. We’re really struggling right now. My uncle just got let go from his job due to covid. I’m trying really hard to better my health so I can get a job. Please anything helps CashApp: $Jenn052192
@CashpersCraving @TriciaHuff14 @piccmeeprizes @J4CKMULL I lost my job due to Covid,I had to sell most of our stuff just to be able to eat these last few months we can’t afford food..my family always comes first for me & if anyone can help us we would appreciate it more than you know. .Godbless and I’m sorry I even have to ask.ty
@pulte Please I would be sooooo thankful I’ve been unemployed since March from COVID and have been able to land a job bills are piling up so please $stefany12
@Trump_owo @LindseyGrahamSC @actblue I’ve donated A LOT and I am unemployed due to your abusive inaction re: COVID 19. And I will continue to donate. SOOOO STFU! https://t.co/g6vqcgcs8z
@tapairportugal Can you take care of your customers instead of screwing them?? Please and thank you. Covid has messed a lot up–be willing to work with your customers. Why are you refusing to extend my voucher when it was impossible to use it this summer and cannot choose dates for next year yet
@AlwayzInTrouble Hope you will check out my doc. If you like it please spread the word. Due to Covid, no work for me. So, I’m living on my marketing budget. So, word of mouth is now my marketing. https://t.co/o2AkKx8iNB
@pulte I’ve been laid off work because I have COVID symptoms i don’t know how I’m going to pay my rent and all other bills. £abbielengthorn
Acknowledgements
This work is partially supported by the DARPA SocialSim Program and the Air Force Research Laboratory under contract FA8650-18-C-7825.
Author Contributions
SW and YL conceived and designed the experiments. YZ, ZF, RW, RX, HG, BG, TS, and LZ performed the experiments. YZ, ZF, RW, MZ and HZ interpreted the data and prepared the figures. SW, YL, and YZ wrote and revised the manuscript.
Funding
This work is partly supported by the DARPA SocialSim Program and the Air Force Research Laboratory under contract FA8650-18-C-7825.
Data Availability
This work uses a publicly available dataset focused on the Twitter discussions around this event and released as part of the 2021 Grand Challenge of the North American Social Network Conference.
1 https://github.com/pnnl/socialsim
2 https://twitter.com/i/events/1301328910227963907.
3 https://www.statnews.com/2020/09/08/astrazeneca-covid-19-vaccine-study-put-on-hold-due-to-suspected-adverse-reaction-in-participant-in-the-u-k/.
4 https://dranganathans.blogspot.com/2020/09/covid-reinfection-sign-of-super.html.
5 https://thenationalpulse.com/2020/09/14/whistleblower-dr-yan-releases-report-suggesting-coronavirus-was-lab-modified.
6 https://tfhub.dev/google/universal-sentence-encoder/4.
7 https://www.zerohedge.com/markets/ft-confirms-astrazeneca-covid-19-vaccine-caused-serious-spinal-issues-test-patient
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
Amazon (2021) Alexa Top 1M. https://www.alexa.com/topsites. Accessed Dec 2021
Angelov D (2020) Top2vec: distributed representations of topics. CoRR abs/2008.09470. https://arxiv.org/abs/2008.09470
Bagherpour A (2020) Covid misinformation is killing people. https://www.scientificamerican.com/article/covid-misinformation-is-killing-people1/. Accessed Dec 2021
Chadwick A Kaiser J Vaccari C Online social endorsement and covid-19 vaccine hesitancy in the United Kingdom Soc Media + Soc 2021 10.1177/20563051211008817
Cyranoski D Mallapaty S Scientists relieved as coronavirus vaccine trial restarts but question lack of transparency Nature 2020 585 7825 331 332 10.1038/d41586-020-02633-6 32929259
De Silva R, Weerasekara N, Kin Wai NG et al (2022) COVID-19 vaccine disinformation. https://github.com/qcri/Vaccine_Disinfo
DeVerna M, Pierri F, Truong BT et al (2021) Covaxxy: a global collection of english Twitter posts about COVID-19 vaccines. CoRR abs/2101.07694. https://arxiv.org/abs/2101.07694
Factual T (2020) The best and worst sources for unbiased news. https://www.thefactual.com/blog/biased-factual-reliable-new-sources/. Accessed Dec 2021
Farsight Security (2021) DNS database. https://www.dnsdb.info/. Accessed Dec 2021
FAS (2020) Spanish-language vaccine news stories hosting malware disseminated via URL shorteners. https://fas.org/disinfoblog/spanish-language-vaccine-news-stories-hosting-malware-disseminated-via-url-shorteners/. Accessed Dec 2021
Ferrara E Cresci S Luceri L Misinformation, manipulation, and abuse on social media in the era of covid-19 J Comput Soc Sci 2020 3 2 271 277 10.1007/s42001-020-00094-5 33251373
Forbes (2020) Why suspension of the AstraZeneca vaccine trial matters. https://www.forbes.com/sites/williamhaseltine/2020/09/09/why-astrazenecas-move-to-pause-their-vaccine-trial-matters/. Accessed Dec 2021
Giglietto F Righetti N Rossi L It takes a village to manipulate the media: coordinated link sharing behavior during 2018 and 2019 Italian elections Inf Commun Soc 2020 23 6 867 891 10.1080/1369118X.2020.1739732
Horawalavithana S, Ng KW, Iamnitchi A (2020) Twitter is the megaphone of cross-platform messaging on the White Helmets. In: International conference on social computing, behavioral-cultural modeling and prediction and behavior representation in modeling and simulation. SBP-BRiMS 2020, Springer, pp 235–244. 10.1007/978-3-030-61255-9_23
Horawalavithana S, De Silva R, Nabeel M et al (2021) Malicious and low credibility URLs on Twitter during the AstraZeneca covid-19 vaccine development. In: International conference on social computing, behavioral-cultural modeling and prediction and behavior representation in modeling and simulation. SBP-BRiMS 2021, Springer, pp 3–12. 10.1007/978-3-030-80387-2_1
Huang B, Carley KM (2020) Disinformation and misinformation on Twitter during the novel coronavirus outbreak. CoRR abs/2006.04278. https://arxiv.org/abs/2006.04278
Jemielniak D Krempovych Y An analysis of AstraZeneca covid-19 vaccine misinformation and fear mongering on Twitter Public Health 2021 200 4 6 10.1016/j.puhe.2021.08.019 34628307
Karami A Zhu M Goldschmidt B Covid-19 vaccine and social media in the U.S.: exploring emotions and discussions on Twitter Vaccines 2021 10.3390/vaccines9101059
Keller FB Schoch D Stier S Political astroturfing on Twitter: how to coordinate a disinformation campaign Polit Commun 2020 37 2 256 280 10.1080/10584609.2019.1661888
Kin Wai NG, Horawalavithana S, Iamnitchi A (2021) Multi-platform information operations: Twitter, Facebook and YouTube against the White Helmets. In: Workshop proceedings of the 15th international AAAI conference on web and social media. 10.36190/2021.36
MBFC (2020) Media bias/fact check—search and learn the bias of news media. https://mediabiasfactcheck.com/. Accessed Dec 2021
NASN (2021) Nasn 2021: grand challenge. https://www.insna.org/nasn-2021-grand-challenge
Nuzhath T, Tasnim S, Sanjwal RK et al (2020) Covid-19 vaccination hesitancy, misinformation and conspiracy theories on social media: a content analysis of Twitter data. SocArXiv 10.31235/osf.io/vc9jb
Pacheco D, Hui PM, Torres-Lugo C et al (2021) Uncovering coordinated networks on social media: methods and case studies. In: Proceedings of the international AAAI conference on web and social media, vol 15, pp 455–466. https://ojs.aaai.org/index.php/ICWSM/article/view/18075
PBS (2020) FDA weighs whether to resume AstraZeneca vaccine trials after patient’s serious side effect. https://www.pbs.org/newshour/health/fda-weighs-whether-to-resume-astrazeneca-vaccine-trials-after-patients-serious-side-effect. Accessed Dec 2021
Pranesh R, Farokhenajd M, Shekhar A et al (2021) CMTA: COVID-19 misinformation multilingual analysis on Twitter. In: Proceedings of the 59th annual meeting of the association for computational linguistics and the 11th international joint conference on natural language processing: student research workshop. Association for Computational Linguistics, pp 270–283. 10.18653/v1/2021.acl-srw.28
Reuters (2021) Covid-19 vaccine AstraZeneca roller-coaster ride. https://www.reuters.com/business/healthcare-pharmaceuticals/covid-19-vaccine-astrazeneca-roller-coaster-ride-2021-03-24/. Accessed Dec 2021
Robbins R, Feuerstein A, Branswell H (2020) AstraZeneca covid-19 vaccine study is put on hold. https://www.statnews.com/2020/09/08/astrazeneca-covid-19-vaccine-study-put-on-hold-due-to-suspected-adverse-reaction-in-participant-in-the-u-k/
Rosenberg H Syed S Rezaie S The Twitter pandemic: the critical role of Twitter in the dissemination of medical information and misinformation during the covid-19 pandemic Can J Emerg Med 2020 22 4 418 421 10.1017/cem.2020.361
Sharma K, Zhang Y, Liu Y (2022) COVID-19 vaccines: characterizing misinformation campaigns and vaccine hesitancy on Twitter. In: Proceedings of the international AAAI conference on web and social media, vol 16, pp 920–931. https://ojs.aaai.org/index.php/ICWSM/article/view/19346
Silva M, Ceschin F, Shrestha P et al (2020) People still care about facts: Twitter users engage more with factual discourse than misinformation: a comparison between covid and general narratives on Twitter. CoRR abs/2012.02164. https://arxiv.org/abs/2012.02164
Silva RD, Nabeel M, Elvitigala C et al (2021) Compromised or attacker-owned: a large scale classification and study of hosting domains of malicious urls. In: 30th USENIX security symposium (USENIX Security 21). USENIX Association
Singh L Bode L Budak C Understanding high and low-quality url sharing on covid-19 Twitter streams J Comput Soc Sci 2020 3 2 343 366 10.1007/s42001-020-00093-6 33263092
Smith R, Cubbon S, Wardle C (2020) Under the surface: Covid-19 vaccine narratives, misinformation and data deficits on social media. https://firstdraftnews.org/long-form-article/under-the-surface-covid-19-vaccine-narratives-misinformation-and-data-deficits-on-social-media/
Teng X, Lin YR, Chung WT et al (2022) Characterizing user susceptibility to covid-19 misinformation on Twitter. In: Proceedings of the international AAAI conference on web and social media, vol 16, pp 1005–1016
VirusTotal (2021) VirusTotal free online virus, Malware and URL scanner. https://www.virustotal.com/. Accessed Dec 2021
Vivek M (2021) Confronting health misinformation: the U.S. surgeon general’s advisory on building a healthy information environment. https://www.ncbi.nlm.nih.gov/books/NBK572169/. Accessed Dec 2021
Weber D, Neumann F (2020) Who’s in the gang? Revealing coordinating communities in social media. In: 2020 IEEE/ACM international conference on advances in social networks analysis and mining (ASONAM), IEEE, pp 89–93
Wilson SL Wiysonge C Social media and vaccine hesitancy BMJ Glob Health 2020 10.1136/bmjgh-2020-004206
Yang KC, Torres-Lugo C, Menczer F (2020) Prevalence of low-credibility information on twitter during the COVID-19 outbreak. In: Proc. ICWSM Intl. Workshop on Cyber Social Threats (CySoc). 10.36190/2020.16
| 36466588 | PMC9703426 | NO-CC CODE | 2022-11-29 23:21:09 | no | Comput Math Organ Theory. 2022 Nov 28;:1-22 | utf-8 | Comput Math Organ Theory | 2,022 | 10.1007/s10588-022-09370-3 | oa_other |
==== Front
Appl Microbiol Biotechnol
Appl Microbiol Biotechnol
Applied Microbiology and Biotechnology
0175-7598
1432-0614
Springer Berlin Heidelberg Berlin/Heidelberg
36441207
12300
10.1007/s00253-022-12300-7
Applied Microbial and Cell Physiology
Extracellular vesicles of Candida albicans regulate its own growth through the l-arginine/nitric oxide pathway
Wei Yu 12
Wang Zheng 12
Liu Yaqi 1
Liao Binyou 1
Zong Yawen 12
Shi Yangyang 12
Liao Min 12
Wang Jiannan 1
Zhou Xuedong 12
Cheng Lei [email protected]
12
http://orcid.org/0000-0003-4215-2873
Ren Biao [email protected]
1
1 grid.13291.38 0000 0001 0807 1581 State Key Laboratory of Oral Diseases &, National Clinical Research Center for Oral Diseases, West China School of Stomatology, Sichuan University, Chengdu, 610000 Sichuan Province China
2 grid.13291.38 0000 0001 0807 1581 Department of Operative Dentistry and Endodontics, West China School of Stomatology, Sichuan University, Chengdu, 610000 Sichuan Province China
28 11 2022
2023
107 1 355367
13 9 2022
12 11 2022
15 11 2022
© The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2022, Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Abstract
Candida albicans is the main conditional pathogenic fungus among the human microbiome. Extracellular vesicles (EVs) secreted by C. albicans are important for its pathogenesis. However, the effects and mechanisms of EVs on C. albicans own growth are not clear. Here, we isolated EVs from C. albicans cells grown in four culture media, including RPMI 1640, DMEM, YPD, and YNB, and measured their effects on the own growth of C. albicans in these media. All the C. albicans EVs from the four media could promote the growth of C. albicans in RPMI 1640 and DMEM media, but had no effects in YPD and YNB media, indicating that the effects of EVs on C. albicans growth were dependent on some media contents. By comparing the media contents and transcriptome analysis, arginine was identified as the key factor for the growth promotion of C. albicans EVs. EVs activated the l-arginine/nitric oxide pathway to promote the growth of C. albicans through that EVs increased the NO levels and upregulated the expression of NO dioxygenase gene YHB1 to reduce the intracellular reactive oxygen species (ROS) and cell apoptosis. During the host cell infections, C. albicans EVs synergistically enhanced the destructive effects of C. albicans to host cells, including RAW264.7, HOK, TR146, and HGEC, suggesting that the growth promotion by EVs enhanced the pathogenesis of C. albicans. Our results demonstrated the important roles of EVs on C. albicans own growth for the first time and highlight its synergism with C. albicans to increase the pathogenesis.
Key points
• C. albicans extracellular vesicles (EVs) promoted its own growth.
• EVs activated the l-arginine/NO pathway to reduce ROS and apoptosis of C. albicans.
• EVs enhanced the damage to the host cell caused by C. albicans.
Supplementary Information
The online version contains supplementary material available at 10.1007/s00253-022-12300-7.
Keywords
Fungal infection
Extracellular vesicles
Pathogenesis
Intracellular ROS
Cell apoptosis
http://dx.doi.org/10.13039/501100001809 National Natural Science Foundation of China 82071106 82271033 81600858 81870778 81991500 81991501 Cheng Lei Ren Biao Key Research and Development Projects of Science and Technology Department of Sichuan Province2021YFQ0064 Ren Biao http://dx.doi.org/10.13039/100012551 Applied Basic Research Program of Sichuan Province 2020YJ0227 Ren Biao Technology Innovation R&D Project of Chengdu2022-YF05-01401-SN Ren Biao Research Funding from West China School/Hospital of Stomatology Sichuan UniversityRCDWJS2021-19 Cheng Lei issue-copyright-statement© Springer-Verlag GmbH Germany, part of Springer Nature 2023
==== Body
pmcIntroduction
Candida albicans is one of the most common opportunistic fungal pathogens (Lee et al. 2021). It can colonize in the oral cavity, vagina, and digestive tract of healthy people, and become pathogenic when the host is under low immune conditions (Pappas et al. 2016), particularly in immunocompromised individuals, such as patients with AIDS, patients undergoing chemotherapy, and individuals receiving immunosuppressant therapies (Lohse et al. 2018). C. albicans infections include superficial mucosal infection, dermal infections, and disseminated bloodstream infections with mortality rates above 40% (Calderone and Fonzi 2001; Pappas et al. 2004; Wenzel 1995). It is also one of the most common coinfected fungal species with SARS-CoV-2 in COVID-19 (Chen et al. 2020b).
Extracellular vesicles (EVs) are naturally released nano-scale particles from different cells (Oliveira et al. 2010; van Niel et al. 2018). They were once considered carriers of cell “garbage” or debris shed by cells, but now, EVs have proved to be important for the cell communications in unicellular and multicellular organisms (Dawson et al. 2020; Rybak and Robatzek 2019). C. albicans EVs are composed of proteins, lipids, nucleic acids, and carbohydrates (de Toledo Martins et al. 2019), including proteins with 1, 3-β-glucosidase activity (Gow and Hube 2012), enolase, 3-phosphate dehydrogenase (Gpdh), phosphoglycerate kinase (Pgk), and phosphoglycerate mutase (Karkowska-Kuleta et al. 2011), which were found to be highly related to the fungal cell attachment and the interactions with the host (Sandini et al. 2011). C. albicans EVs were capable to regulate its pathogenic process and drug resistance (Rodrigues et al. 2013; Roszkowiak et al. 2019). An endosomal sorting complex required for transport (ESCRT)-deficient mutant of C. albicans reduced EVs production and greatly increased sensitivity to the antifungal drug fluconazole, indicating the effects of EVs on the antifungal drug responses of C. albicans (Zarnowski et al. 2018). C. albicans EVs also stimulated macrophages to produce NO and cytokine IL-10, and dendritic cells to produce cytokines such as TGF-β and TNF-α (Zamith-Miranda et al. 2018) indicating the crosstalk between C. albicans EVs and host immune cells. Recently, EVs produced by C. albicans were found to inhibit its biofilm formation and yeast to hyphae transition. The sesquiterpenes, diterpenes, and fatty acids from C. albicans EVs stopped its filamentation and promoted the formation of pseudo hyphae (Bitencourt et al. 2022; Honorato et al. 2022). However, the effects and mechanisms of C. albicans EVs on its own growth are still unclear.
Arginine is one of the most versatile amino acids in cells (Zou et al. 2019). It is a precursor not only for protein synthesis but also for the synthesis of nitric oxide, urea, polyamines, proline, glutamate, creatine, and agmatine (Wu and Morris 1998). Arginine has a variety of functions, including immunomodulatory, antioxidant, anti-inflammatory, regulation of cell proliferation, anti-apoptosis, and regulation of lipid metabolism. It has been also widely used in clinical nutritional therapy (Gogoi et al. 2016; Khalaf et al. 2019; Szefel et al. 2019). Currently, there are four main arginine metabolic pathways: (1) converting to creatine under the action of arginine-glycine transferase (Barcelos et al. 2016); (2) biosynthesizing agmatine through arginine decarboxylase (ADC) decarboxylation (Hyvönen et al. 2020); (3) producing bioactive nitric oxide (NO) and citrulline through nitric oxide synthase (eNOS) (Wu et al. 2021); (4) decomposing into ornithine and urea by arginase (Longo et al. 2020). C. albicans can metabolize arginine and grow on the media in which arginine was the sole nitrogen source (Schaefer et al. 2020). Arginine can promote the hyphal growth and biofilm formation of C. albicans and also enhance the cross-kingdom interactions with bacteria (Xiong et al. 2022). Arginine is a precursor of NO as well, while the endogenous NO produced by C. albicans protects itself against azoles (Li et al. 2016). However, the role of arginine in the actions of C. albicans EVs remains unknown.
In our study, aiming to reveal the effects of EVs on the growth of C. albicans, EVs from the C. albicans cells in four media were isolated. We found that EVs were proved to promote the growth of C. albicans through the l-arginine/nitric oxide pathway to reduce intracellular reactive oxygen species (ROS) and cell apoptosis for the first time. Our results highlighted the important roles of EVs on C. albicans own growth and pathogenesis.
Materials and methods
Strain and culture conditions
C. albicans SC5314 (ATCC MYA − 2876) was grown in YPD plates (4 g yeast extract, 8 g anhydrous glucose, 8 g peptone, 8 g agar dissolved in 400 mL deionized water) at 37 °C overnight (Chen et al. 2020a; Zhou et al. 2021; Zhu et al. 2021). For the treatment with EVs, the colonies of C. albicans were picked out and placed into phosphate buffered saline (PBS), adjust the final suspensions to 1 × 106 colony-forming unit (CFU)/mL concentration in different medium, including YPD, YNB (Solarbio, Beijing, China), RPMI 1640 (Thermo Fisher Scientific, Waltham, MA, USA), and DMEM (Gibco, Grand Island, NY, USA), and incubated at 37 °C.
Isolation of EVs by overspeed centrifugation
C. albicans EVs were isolated by overspeed centrifugation as described previously (Martínez-López et al. 2022). Briefly, C. albicans was incubated in YPD liquid medium overnight at 37 °C. The precipitates were collected by centrifugation at 4000 g 4 °C and resuspended in PBS. After being counted under a microscope using cell counting plates, C. albicans was inoculated in YPD/YNB/RPMI 1640/DMEM medium to a final concentration of 1 × 106 CFU/mL and cultivated at 30 °C and 150 rpm for 72 h. C. albicans supernatant was then collected by centrifugation at 15,000 g for 15 min at 4 °C. The supernatant was filtered by a filter with a pore size of 0.35 μm. Then, a 100 KD ultrafiltration tube was used to concentrate the supernatant to 1/20 of the original volume. The supernatant was removed by centrifugation at 4 °C and 100,000 g for 2 h, and the precipitate was rinsed with frozen PBS. The precipitate was centrifuged again at 4 °C and 100,000 g for 2 h. After the supernatant was removed, the precipitate was resuspended with 1 mL frozen PBS. The final EVs’ suspension was collected and filtered with a 0.22 μm filter, stored at 4 °C until used.
Identification of EVs
Scanning electron microscope (SEM) observation
EVs were observed by SEM as described previously (Chen et al. 2017). Ten microliters of EVs collected above was evenly spread on round cell climbing sheets, air-dried at room temperature, and fixed in 2.5% glutaraldehyde solution overnight at 4 °C. The fixed EVs’ samples were dehydrated successively with different concentrations of ethanol (50%, 60%, 70%, 80%, 90%, 95%, and absolute ethanol), and each concentration was dehydrated for 15 min. After drying and spraying gold, the samples were observed by scanning electron microscopy Tecnai G2 F20 S-TWIN (FEI Company, Hillsboro, OR, USA).
Transmission electron microscopy (TEM) observation
EVs were observed by TEM (Karkowska-Kuleta et al. 2020). Briefly, 100 μL of EVs was aspirated and dropped onto the copper grid for 10 min. Fifty microliters of 1% phosphotungstic acid was aspirated and stained on the copper grid for 2 min. Samples were rinsed with deionized water for 2 times, air dried, and observed on the transmission electron microscopy (FEI Company, Hillsboro, OR, USA).
Particle size detection
Particle size distribution of EVs was detected by nano-size analyzer (Honorato et al. 2022). Briefly, the suspension of EVs was diluted to 2 mL and added to the quartz dish of nano-size analyzer Zetasizer Nano ZS (Malvern Panalytical, Malvern, UK). The range of EVs’ particle size was detected at 25 °C.
EVs’ protein concentration detection
The concentrations of EVs’ proteins were measured according to the instructions of BCA protein quantification kit (Beyotime, Chengdu, China). Briefly, EVs’ suspensions were absorbed into a 96-well plate, and 200 μL of reaction solution was added to each well. The samples were then incubated at 37 °C for 30 min. The absorbance value of each well sample at A562 nm was detected by microplate reader (Thermo Fisher Scientific, Waltham, MA, USA). The absorbance values of different concentrations of EVs’ protein were calculated according to the BCA standard curves.
Colony forming unit count
C. albicans were treated with 15 μg/mL EVs from YNB or PBS control in different medium at 37 °C for 24 h. The influence of different amino acids on promoting proliferative effects was tested in 0.2% glucose YNB and YPD media in addition of 0.02%, 0.05%, 0.1%, 0.2%, 0.3% l-arginine, 0.2% l-cysteine, 0.2% l-proline, 0.2% l-leucine, 0.2% l-isoleucine, 0.2% l-valine, 0.2% l-methionine, 0.2% l-glutamic acid, 0.2% l-ornithine, 0.2% l-citrulline, and 0.2% l-histidine, respectively (all the amino acids were from Solarbio, Beijing, China). The influence of different glucose concentrations on promoting proliferative effects was tested using media including in RPMI 1640 medium with the addition of 0.2%, 2% glucose, and YNB medium in addition of 0.2%, 0.4%, 0.8%, and 2% glucose, respectively. The influence of oxidants and antioxidants on promoting proliferative effects was tested using media including 0.2% glucose YNB with 0.25 mM, 0.5 mM H2O2 (Boster, Wuhan, China) or 0.125 mM, and 0.25 mM glutathione (GSH) (Solarbio, Beijing, China). After being mixed in 96-well plates, C. albicans cells were serially diluted with PBS; then, 150 μL of C. albicans dilutions was spread on YPD plates. After 12 h of culture at 37 °C, the CFUs were counted. All the experiments were repeated in triple.
Growth curve of C. albicans
The growth curves of C. albicans were detected as described previously (Wang et al. 2021). C. albicans was treated with 15 μg/mL EVs and PBS, then inoculated in YNB medium to a concentration of 1 × 104 CFU/mL into 96-well plates, covered with 100 μL mineral oil (Sigma-Aldrich, Saint Louis, MO, USA) to prevent evaporation. The plates were incubated in a Multiskan Spectrum (Chro Mate1, Awareness Technology, Palm City, FL, USA) at 37 °C. The absorbance at OD600 nm was measured every 30 min. Absorbance at different time points was plotted to generate the growth curve. Three duplicate wells were set for each group.
Determination of cellular reactive oxygen species
Reactive Oxygen Species Assay Kit (Beyotime, Chengdu, China) was used for ROS (reactive oxygen species) detection. C. albicans was inoculated in 96-well plates with three replicate wells in each group. After the treatment with 15 μg/mL EVs, DCFH-DA (2′,7′-dichlorodihydrofluorescein diacetate) was added with a final concentration of 10 μM in 3 h, 6 h, 9 h, 12 h, and 24 h, and transferred to incubation in 37 °C for 1 h. PBS was served as control. A microplate reader (Thermo Fisher Scientific, Waltham, MA, USA) was used to detect the fluorescence intensity with the wavelength of excitation light at 488 nm and emission light at 525 nm. After fluorescence intensity detection, CFU was counted for each well and the related ROS levels were normalized by CFU.
Fungal cell apoptosis analyzed by flow cytometry
Flow cytometry was performed using the Annexin V-FITC/PI Apoptosis Detection Kit (Yeason, Beijing, China). After the treatment of 15 μg/mL EVs and PBS control at 37 °C for 6 h, C. albicans was centrifuged at 300 g for 5 min at 4 °C. The precipitate was washed twice with pre-cooled PBS. After the centrifugation to collect the precipitate, PBS was discarded, and then resuspended with 100 μL binding buffer. Then, 5 μL Annexin V-FIFC and 10 μL PI were added and the samples were mixed well. After the reaction at room temperature for 10–15 min in the dark, 400 μL binding buffer was added. A flow cytometer (Beckman FC500, Carlsbad, CA, USA) was used for analysis, and the excitation and emission wavelengths were set to 488 and 525 nm, respectively. All experiments were performed in triplicate.
Transcriptomic analysis
Transcriptomic analysis was conducted as described previously (Yawen et al. 2022). Total RNA was extracted from the C. albicans cells treated with 15 μg/mL EVs or PBS control at 37 °C for 6 h by using TRIzol Reagent (Plant RNA Purification Reagent for plant tissue) according to the manufacturer’s instructions (Invitrogen, Carlsbad, CA, USA). Genomic DNA was removed using DNase I (Takara Inc., Chengdu, China). Then, RNA quality was determined by 2100 Bioanalyser (Agilent, Chengdu, China) and quantified using the ND-2000 (NanoDrop Technologies, Thermo Fisher Scientific, Waltham, MA, USA). Only high-quality RNA sample (OD 260/280 = 1.8 ~ 2.2, OD 260/230 ≥ 2.0, RIN ≥ 6.5, 28S:18S ≥ 1.0, > 1 μg) was used to construct a sequencing library (Hu et al. 2021). Then, sequencing is done using a sequencer Illumina Novaseq 6000 (Illumina, San Diego, CA, USA) in the Majorbio company (Shanghai, China).
Determination of nitric oxide content
Measurement of NO in C. albicans was conducted as described by Li et al. (2016). C. albicans was inoculated in 96-well plates with three replicate wells in each group. After the treatment with 15 μg/mL EVs and PBS control at 37 °C for 1 h, DAF-FM DA (3-amino,4-aminomethyl-2′,7′ -difluorescein, diacetate) (Beyotime, Chengdu, China) was added at a final concentration of 5 μM, and then incubated at 35 °C for 1 h in the absence of light. The excitation and emission light wavelengths were set as 495 and 515 nm, respectively, with microplate readers (Thermo Fisher Scientific, Waltham, MA, USA) to detect the fluorescence intensity. After fluorescence intensity detection, CFU was counted for each well and the related NO levels were normalized by CFU.
Lactate dehydrogenase (LDH) cytotoxicity assay
LDH cytotoxicity assay was conducted as described previously (Zhou et al. 2021). Macrophage RAW264.7, human oral keratinocytes (HOK), human squamous carcinoma cells (TR146), and human gingival epithelial cells (HGEC) were inoculated in 96-well plates at 1 × 105 cells/mL using DMEM medium without fatal bovine serum (FBS) and antibiotics. The cells were cultured overnight at 37 °C, 5% CO2. Then, PBS, C. albicans, 15 μg/mL EVs, and C. albicans + 15 μg/mL EVs were added and treated for 24 h. EVs were isolated from RPMI 1640 medium. Cytotoxicity LDH Assay Kit (Dojindo, Beijing, China) was used for LDH detection. After the treatment, 50 μL of Dye Mixture and Assay Buffer was added. After 30-min reaction, 25 μL stop solution was added. The absorbance of each well sample at A490 nm was detected by microplate reader (Thermo Fisher Scientific, Waltham, MA, USA). All experiments were performed in triplicate.
RNA extraction and qPCR
qPCR was conducted as described previously (Zhu et al. 2021). C. albicans treated with 15 μg/mL EVs and PBS control at 37 °C for 6 h. The samples were collected by centrifugation at 4000 g for 10 min at 4 °C after treatment with EVs and PBS control in C. albicans and resuspended with TRIzol reagent (Invitrogen, Carlsbad, CA, USA). The total RNA was extracted after the wall was broken with liquid nitrogen. Excess DNA in RNA was removed and reverse transcribed using the PrimeScriptTM RT Reagent Kit Reverse Transcription Kit (Takara Inc., Chengdu, China) (Hu et al. 2021; Kong et al. 2022). Real-time PCR was performed using reverse transcription cDNA as template. Gene amplification was performed following the SYBR® PremixEx Taq™ kit (Takara Inc., Chengdu, China) two-step strategy: (1) 95 °C for 30 s; (2) 40 PCR cycles (95 °C for 5 s, a gene-specific annealing temperature for 30 s). The primer sequence was 18S-F: TGGAAGCTGCTGGTATTGAC, 18S-R: TCCTTTTGCATACGTTCAGC. YHB1-F: ATCGATTTAGAAGCCGCAGA, YHB1-R: GACCACGTTCAGGTTTTGGT. The qPCRs were run on LightCycler 480 II (Roche, Basel, Switzerland). The formula for calculating the relative value of gene expression was 2−△△Ct (Zhou et al. 2018). All experiments were performed in triplicate.
Statistical analysis
Data are represented as the mean ± standard deviations (SD) from at least three biological replicates and three technical replicates. The level of significance was analyzed by unpaired t-test and one-way ANOVA. p < 0.05 was considered significant. The Jin value method formula was used to calculate the collaboration index: EA+B = EA + EB-EA EB, Q > 1.15 indicates a synergistic effect (Jin 2004). All statistical analyses were performed using GraphPad Prism 8 v8.3.1 (GraphPad software, Beijing, China).
Results
EVs promoted the growth of C. albicans
EVs were successfully isolated from C. albicans cells grown in different media by overspeed centrifugation. C. albicans EVs were nanoparticles with bilayer membranes (Fig. 1a, b) with the sizes arranging from 100 to 500 nm (Fig. 1c). The protein concentrations of EVs were 200–300 μg/mL. When C. albicans was treated with EVs (5, 10, and 15 μg/mL), EVs significantly promoted the growth of C. albicans at dose dependent manner (Fig. 1d, e) and 15 μg/mL was then selected for further evaluation. To determine whether different media would affect the C. albicans EVs’ promoting properties, EVs from C. albicans cells grown in four culture media including RPMI 1640, DMEM, YPD, and YNB media were isolated. Interestingly, all the isolated EVs significantly promoted the growth of C. albicans in RIMI 1640 and DMEM media but had no effects in YNB or YPD medium (Fig. 1f), indicating that the growth promotion of EVs was dependent on the contents of C. albicans growth media.Fig. 1 C. albicans EVs promoted its growth. a SEM of C. albicans EVs isolated from YNB medium. Scale bar, 1 μm. b TEM of C. albicans EVs isolated from YNB medium. Scale bar, 100 nm. c Range of size distribution of C. albicans EVs isolated from YNB medium measured by nanoparticle tracking analysis (NTA). d CFUs of C. albicans grown in RPMI 1640 medium treated with different concentrations of EVs and PBS were served as control. e Growth curves of C. albicans treated with 15 μg/mL EVs and PBS were served as control. f C. albicans treated with different EVs grew in different media. All of the experiments were performed in three distinct replicates, and the data are presented as the means ± SD, **p < 0.05, ****p < 0.0001, no significance (ns) p > 0.05
EVs regulated the arginine metabolism of C. albicans
To identify how the EVs promoted the growth of C. albicans, transcriptomic analysis of C. albicans treated by EVs in RPMI 1640 medium was performed. EVs significantly upregulated 150 genes and downregulated 315 genes compared to the control group (Supplemental Fig. S1a). GO enrichment analysis indicated that the cellular and metabolic processes of EVs treated C. albicans were significantly changed (Supplemental Fig. S1b). KEGG pathway enrichment analysis indicated that the arginine and proline metabolism pathway and arginine biosynthesis pathway were significantly enriched after the EVs’ treatment (Fig. 2a). The genes related to arginine biosynthesis pathway were significantly upregulated, while the genes associated with arginine degradation were significantly downregulated (Fig. 2b, Supplemental Fig. S2a, b, Supplemental Table S1). The transcriptome data suggested that the growth promotion capability of EVs on C. albicans own growth was highly associated with C. albicans arginine metabolism related pathways.Fig. 2 Transcriptomic analysis of C. albicans treated by EVs. a The enriched KEGG pathways of C. albicans treated with 15 μg/mL EVs compared to that from C. albicans treated with PBS. b Heat map of shifted specific genes from the arginine biosynthesis and arginine degradation pathways
C. albicans growth promotion by EVs was dependent on arginine
To confirm whether EVs’ growth promotion capability depended on arginine according to the transcriptomic analysis, the components of the four different media were firstly compared. The media differences between growth promotion media (RPMI 1640 and DMEM) and non-promotion media (YNB and YPD) were mainly glucose and amino acids (Supplemental Table S2). Then, glucose and amino acids were added to YNB and YPD media to verify whether they could affect the growth promotion capability of EVs. The addition of glucose did not affect the growth promotion characters of EVs on C. albicans in both promotion and non-promotion media, including YNB and RPMI 1640 media, respectively (Supplemental Fig. S3); however, when arginine was added to non-promotion media (YNB and YPD media), EVs significantly promoted the growth of C. albicans at dose dependent manner (Fig. 3a, b). Then the growth promotion was confirmed whether it was arginine specific by adding different amino acids into YNB media, and it turned out that EVs can only promote the growth of C. albicans under the addition of arginine (Fig. 3c) indicating that the promotion of C. albicans growth by EVs was dependent on arginine.Fig. 3 The growth promotion of EVs was dependent on arginine. a Effects of different arginine concentrations on the growth regulation induced by 15 μg/mL EVs. EVs were isolated from YNB medium and C. albicans grew in YNB medium. PBS was served as control. b Effects of the growth regulation induced by 15 μg/mL EVs in YPD medium with or without arginine. EVs were isolated from YNB medium. PBS was served as control. c Effect of media containing different amino acids on the growth regulation induced by 15 μg/mL EVs. EVs were isolated from YNB medium and C. albicans grew in YNB medium. All of the experiments were performed in three distinct replicates, and the data are presented as the means ± SD, *p < 0.05, ***p < 0.001, ****p < 0.0001, no significance (ns) p > 0.05
EVs activated l-arginine/nitric oxide pathway
Since the arginine biosynthesis pathway was upregulated but the degradation pathway was downregulated, we then tested whether the upregulated arginine biosynthesis activated the l-arginine/nitric oxide pathway. YHB1, a nitric oxide dioxygenase gene, plays essential roles in nitric oxide scavenging/detoxification in C. albicans (Ullmann et al. 2004). Therefore, the expression of YHB1 was firstly measured. EVs significantly upregulated the expression of YHB1 (Fig. 4a). The intracellular nitric oxide (NO) levels of C. albicans treated by EVs in different media were then measured. EVs significantly increased the intracellular NO levels of C. albicans grown in RPMI 1640 and DMEM media, as well as the YNB and YPD media with the addition of arginine, but had no effects in YNB and YPD media (Fig. 4b), in line with the arginine dependent growth promotion of EVs, and also indicating that EVs promote the growth of C. albicans through the activation of the l-arginine/nitric oxide pathway.Fig. 4 EVs activated the l-arginine/NO pathway. a YHB1 mRNA expression in C. albicans after 15 μg/mL EVs’ treatment. EVs were isolated from YNB medium and C. albicans grew in YNB medium with 0.2% arginine. b Intracellular NO content in C. albicans after treatment with 15 μg/mL EVs in different medium. EVs were isolated from YNB medium. All of the experiments were performed in three distinct replicates, and the data are presented as the means ± SD, **p < 0.01, ***p < 0.001, ****p < 0.0001, no significance (ns) p > 0.05
EVs reduced the ROS level of C. albicans
The accumulation of intracellular NO levels plays important roles in the regulation of oxidative stress in cells (Araujo and Welch 2006; Förstermann et al. 2017). Then, the reactive oxygen species (ROS) were measured from the C. albicans cells since the EVs significantly increased the intracellular NO levels of C. albicans in the presence of arginine. EVs significantly reduced the intracellular ROS contents in C. albicans to 65% compared to that from the PBS control group at 3 h and 35% at 24 h (Fig. 5a). The addition of oxidant H2O2 and antioxidant GSH was employed to further validate the effects of EVs on ROS production. The addition of H2O2 significantly reduced the growth promotion abilities of EVs on C. albicans, while the addition of GSH significantly enhanced the growth promotion of EVs (Fig. 5b, c), further indicating that EVs promoted the growth of C. albicans by the reduction of the ROS level. ROS is an important inducer for cell apoptosis. Therefore, the fungal cell apoptosis of C. albicans treated by EVs was then measured. EVs significantly reduced the ratio of early apoptosis, late apoptosis, and cell necrosis in C. albicans (Fig. 5d, e), indicating that EVs inhibited the ROS production through the activation of l-arginine/nitric oxide pathway to reduce the cell apoptosis and to promote the growth of C. albicans.Fig. 5 EVs decreased ROS accumulation of C. albicans to reduce fungal cell apoptosis. a Intracellular ROS content in C. albicans. The triangle-labeled group was C. albicans grew in YNB medium with 0.2% arginine, and the circle-labeled group was C. albicans grew in YNB medium without arginine. The control group was C. albicans treated with PBS, and the experimental group was C. albicans treated with 15 μg/mL EVs. The percentages shown in the figure are C. albicans treated with EVs vs C. albicans treated with PBS. b Effect of oxidant (H2O2) treatment on the growth regulation of C. albicans by 15 μg/mL EVs. c Effect of antioxidant (GSH) treatment on the growth regulation of C. albicans by 15 μg/mL EVs. d Flow cytometry is used to detect the proportion of early apoptosis, late apoptosis, and cell necrosis in C. albicans. The first three figures are PBS-treated groups, and the last three figures are 15 μg/mL EV-treated groups. e Statistical analysis of flow cytometry. All of the experiments above were used EVs isolated from YNB medium and C. albicans grew in YNB medium with 0.2% arginine. All the experiments were performed in three distinct replicates, and the data are presented as the means ± SD, *p < 0.05, ****p < 0.0001, no significance (ns) p > 0.05
EVs enhanced the damage to the host cell caused by C. albicans
The pathogenesis of C. albicans affected the growth promotion of EVs was then evaluated. EVs alone showed weak capabilities on different host cells, including macrophage RAW264.7, human oral keratinocytes (HOK), human squamous carcinoma cells (TR146), and human gingival epithelial cells (HGEC). However, EVs significantly increased the cell damages caused by C. albicans (Fig. 6a). The synergistic effect between EVs and C. albicans was evaluated using the Jin value method (Jin 2004). The Q values of RAW264.7, HOK, TR146, and HGEC were respectively 3.82, 1.92, 2.98, and 2.35, indicating that EVs and C. albicans had a synergistic effect on cell destruction. Meanwhile, the host cell damage caused by EVs and C. albicans combinations was also positively related to the C. albicans cell numbers (Fig. 6b).Fig. 6 EVs’ synergies with C. albicans to destroy host cells. a Percentage of cytotoxicity of different cells after 15 μg/mL EVs and PBS control treatment. EVs were isolated from RPMI 1640 medium and the cells grew in DMEM medium. b Cytotoxicity of different concentrations of C. albicans and EVs in RAW264.7 cells. All of the experiments were performed in three distinct replicates, and the data are presented as the means ± SD, ***p < 0.001, ****p < 0.0001
Discussion
Fungal EVs were first reported in 1972 (Gibson and Peberdy 1972) and further studies have proved the important roles of fungal EVs in drug resistance, fungal pathogenicity, and host immune response (Liebana-Jordan et al. 2021; Reales-Calderón et al. 2017; Yang et al. 2020). Recently, Zarnowski et al. (2021) investigated the effects of C. albicans EVs on the biofilm development and found that EVs could promote the formation of extracellular matrix in C. albicans to increase the antifungal drug resistance and the adhesion and spread of C. albicans. Honorato et al. (2022) found that EVs inhibited C. albicans hyphal development and promoted pseudomycelial formation with multiple budding sites, indicating that fungal EVs were messengers affecting biofilm formation, morphogenesis, and virulence of C. albicans. Bitencourt et al. (2022) proposed that EVs released from filamentous C. albicans promoted the development of the C. albicans mycelial state, while EVs released from yeast-like C. albicans promoted the proliferation of the C. albicans yeast state. Different media greatly influenced the morphology of C. albicans and the contents of EVs (Brown et al. 2014). In our study, EVs isolated from different media, including YPD, YNB, DMEM, and RPMI 1640, could significantly promote the proliferation of C. albicans grown in RPMI 1640 and DMEM media, but not YNB and YPD media, suggesting that the EVs’ growth promotion capability dependents on the medium in which C. albicans growth, rather than the medium in which the EVs were isolated. According to previous studies, 1202 proteins were identified in C. albicans EVs (Dawson et al. 2020). The protein species of C. albicans EVs in mycelial and yeast states are different, while the protein content of C. albicans EVs from the yeast state was 10–100 times higher than that from mycelium state (Zarnowski et al. 2018). Lipids also play an important role due to the similarity of lipid composition within EVs (Rodrigues et al. 2007). The major lipids found in EVs are phospholipids, ergosterols, and ceramides, which are major components of cell membranes (Vargas et al. 2015). Ceramide, known as a “virulence regulator,” is an important immunogenic molecule, and antibodies against ceramide can inhibit fungal growth (Nimrichter and Rodrigues 2011). The protein contents including the proteins with 1, 3-β-glucosidase activity (Gow and Hube 2012), β-1, 6-glucan, mannanan, 3-phosphate dehydrogenase (Gpdh), phosphoglycerate kinase (Pgk), and phosphoglycerate mutase (Karkowska-Kuleta et al. 2011) from EVs can directly affect fungal growth, cell attachment, and host recognition (Sandini et al. 2011), and these protein components may be the key components to promote the growth of C. albicans. Further investigations are needed to identify the key factors from EVs that promote the own growth of C. albicans.
Arginine has a variety of functions, including antioxidant, anti-inflammatory, anti-apoptosis, proliferation promotion, and lipid metabolism regulation (Bronte and Zanovello 2005; Luiking et al. 2005; Popovic et al. 2007; Stechmiller et al. 2005). The l-arginine/nitric oxide pathway is widely recognized as an important regulator of cellular function and communication (Gogoi et al. 2016). It has been broadly applied in the development of septic shock, hypertension, and atherosclerosis, as well as the antihypertensive effect of invertase inhibitors (Palmer 1993; Wu et al. 2021). In this pathway, arginine acts as a substrate to generate NO by endothelial nitric oxide synthase (eNOS) (Moncada and Higgs 1993). NO is a free radical gas that can interact with biological free radicals. It is a potent free radical scavenger/terminator and antioxidant (Boudko 2007). The role of l-arginine/nitric oxide pathway in fungi has not been well studied. Li et al. (2016) identified the presence of endogenous NO in C. albicans and confirmed its participation in the oxidative stress response of C. albicans, but they failed to identify the classic eNOS sequences from the C. albicans genome indicating that C. albicans has a new type of enzyme with NOS-like activity. YHB1 encodes a nitric oxide dioxygenase with the function of nitric oxide scavenging/detoxification. YHB1 can be rapidly activated by NO, while the high level of intracellular NO can also upregulate its expression to enhance the reduction of intracellular oxidative stress (Ullmann et al. 2004). In our study, we proved that EVs upregulated the expression of YHB1 and increased the intracellular NO levels of C. albicans under arginine condition, while EVs also decreased the ROS accumulation and related cell apoptosis of C. albicans. Combining the transcriptome analysis, our results indicated that at the presence of arginine, EVs upregulated the arginine biosynthesis and activated the l-arginine/nitric oxide pathway to increase the intracellular NO levels, then inhibited the ROS accumulation to reduce the cell apoptosis (Fig. 7). In this study, we also found that although EVs had weak host cell damage abilities, but EVs significantly enhanced the cell damage abilities of C. albicans for the first time, indicating that EVs promoted the pathogenesis of C. albicans through the growth promotion. This might provide us with a new way to reduce the pathogenesis of pathogenic fungi.Fig. 7 Schematic diagram of growth promotion of EVs’ pathway. EVs activated the l-arginine/nitric oxide pathway to increase the intracellular NO levels, then inhibited the ROS accumulation to reduce the cell apoptosis and increased its pathogenicity
In conclusion, our study demonstrated that C. albicans could regulate its own growth through secreted EVs. EVs promoted C. albicans growth by reducing intracellular ROS accumulation and decreased cell apoptosis through l-arginine/nitric oxide pathway. EVs also enhanced the abilities of C. albicans to damage host cells. Our study highlighted the effects and mechanism of EVs on C. albicans itself and provided new information for fungal infections and treatment in the future.
Supplementary Information
Below is the link to the electronic supplementary material.Supplementary file1 (PDF 182 KB)
Acknowledgements
We greatly thank Professor Min Hu from Second Affiliated Hospital of West China School for the technical support of flow cytometry.
Author contribution
BR, LC, and XZ conceived and designed research. YW, ZW, YL, JW, and YS conducted experiments. YZ, ML, and BL contributed new reagents or analytical tools. YW and YL analyzed data. YW wrote the manuscript. LC and BR critically revised the manuscript. All authors read and approved the manuscript.
Funding
This study was supported by the National Natural Science Foundation of China grants (81870778, 82071106, 82271033, 81600858, 81991500, 81991501), Key Research and Development Projects of Science and Technology Department of Sichuan Province (2021YFQ0064), Applied Basic Research Programs of Sichuan Province (2020YJ0227), Technology Innovation R&D Project of Chengdu (2022-YF05-01401-SN), and the Research Funding from West China School/Hospital of Stomatology Sichuan University (RCDWJS2021-19).
Data availability
All data generated or analyzed during this study are included in this published article (and its supplementary information included in this published article and its supplementary information files). The sequencing data from this study have been submitted to NCBI’s Sequence Read Archive under accession no. PRJNA877381.
Code availability
Not applicable.
Declarations
Ethical approval
This article does not contain any studies with human participants or animals performed by any of the authors.
Consent for publication
All authors consent to the publication of this manuscript.
Conflict of interest
The authors declare no competing interests.
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
Araujo M Welch WJ Oxidative stress and nitric oxide in kidney function Curr Opin Nephrol Hypertens 2006 15 1 72 77 10.1097/01.mnh.0000191912.65281.e9 16340670
Barcelos RP Stefanello ST Mauriz JL Gonzalez-Gallego J Soares FA Creatine and the liver: metabolism and possible interactions Mini Rev Med Chem 2016 16 1 12 18 10.2174/1389557515666150722102613 26202197
Bitencourt TA Hatanaka O Pessoni AM Freitas MS Trentin G Santos P Rossi A Martinez-Rossi NM Alves LL Casadevall A Rodrigues ML Almeida F Fungal extracellular vesicles are involved in intraspecies intracellular communication mBio 2022 13 1 e0327221 10.1128/mbio.03272-21 35012355
Boudko DY Bioanalytical profile of the L-arginine/nitric oxide pathway and its evaluation by capillary electrophoresis J Chromatogr B Analyt Technol Biomed Life Sci 2007 851 1–2 186 210 10.1016/j.jchromb.2007.02.011
Bronte V Zanovello P Regulation of immune responses by L-arginine metabolism Nat Rev Immunol 2005 5 8 641 654 10.1038/nri1668 16056256
Brown AJ Brown GD Netea MG Gow NA Metabolism impacts upon Candida immunogenicity and pathogenicity at multiple levels Trends Microbiol 2014 22 11 614 622 10.1016/j.tim.2014.07.001 25088819
Calderone RA Fonzi WA Virulence factors of Candida albicans Trends Microbiol 2001 9 7 327 335 10.1016/s0966-842x(01)02094-7 11435107
Chen H Zhou Y Zhou X Liao B Xu HHK Chu CH Cheng L Ren B Dimethylaminododecyl methacrylate inhibits Candida albicans and oropharyngeal candidiasis in a pH-dependent manner Appl Microbiol Biotechnol 2020 104 8 3585 3595 10.1007/s00253-020-10496-0 32125481
Chen X Liao B Cheng L Peng X Xu X Li Y Hu T Li J Zhou X Ren B The microbial coinfection in COVID-19 Appl Microbiol Biotechnol 2020 104 18 7777 7785 10.1007/s00253-020-10814-6 32780290
Chen H, Han Q, Zhou X, Zhang K, Wang S, Xu HHK, Weir MD, Feng M, Li M, Peng X, Ren B, Cheng L (2017) Heat-polymerized resin containing dimethylaminododecyl methacrylate inhibits Candida albicans biofilm. Mater (Basel) 10(4). 10.3390/ma10040431
Dawson CS Garcia-Ceron D Rajapaksha H Faou P Bleackley MR Anderson MA Protein markers for Candida albicans EVs include claudin-like Sur7 family proteins J Extracell Vesicles 2020 9 1 1750810 10.1080/20013078.2020.1750810 32363014
de Toledo Martins S Szwarc P Goldenberg S Alves LR Extracellular vesicles in fungi: composition and functions Curr Top Microbiol Immunol 2019 422 45 59 10.1007/82_2018_141 30242512
Förstermann U Xia N Li H Roles of vascular oxidative stress and nitric oxide in the pathogenesis of atherosclerosis Circ Res 2017 120 4 713 735 10.1161/circresaha.116.309326 28209797
Gibson RK Peberdy JF Fine structure of protoplasts of Aspergillus nidulans J Gen Microbiol 1972 72 3 529 538 10.1099/00221287-72-3-529 4564688
Gogoi M Datey A Wilson KT Chakravortty D Dual role of arginine metabolism in establishing pathogenesis Curr Opin Microbiol 2016 29 43 48 10.1016/j.mib.2015.10.005 26610300
Gow NA Hube B Importance of the Candida albicans cell wall during commensalism and infection Curr Opin Microbiol 2012 15 4 406 412 10.1016/j.mib.2012.04.005 22609181
Honorato L de Araujo JFD Ellis CC Piffer AC Pereira Y Frases S de Sousa Araújo GR Pontes B Mendes MT Pereira MD Guimarães AJ da Silva NM Vargas G Joffe L Del Poeta M Nosanchuk JD Zamith-Miranda D Dos Reis FCG de Oliveira HC Rodrigues ML de Toledo Martins S Alves LR Almeida IC Nimrichter L Extracellular vesicles regulate biofilm formation and yeast-to-hypha differentiation in Candida albicans mBio 2022 13 3 e0030122 10.1128/mbio.00301-22 35420476
Hu Y, Niu Y, Ye X, Zhu C, Tong T, Zhou Y, Zhou X, Cheng L, Ren B (2021) Staphylococcus aureus synergized with Candida albicans to increase the pathogenesis and drug resistance in cutaneous abscess and peritonitis murine models. Pathogens 10(8). 10.3390/pathogens10081036
Hyvönen MT, Keinänen TA, Nuraeva GK, Yanvarev DV, Khomutov M, Khurs EN, Kochetkov SN, Vepsäläinen J, Zhgun AA, Khomutov AR (2020) Hydroxylamine analogue of agmatine: magic bullet for arginine decarboxylase. Biomolecules 10(3). 10.3390/biom10030406
Jin Z About the evaluation of drug combination Acta Pharmacol Sin 2004 25 2 146 147 14769200
Karkowska-Kuleta J Kedracka-Krok S Rapala-Kozik M Kamysz W Bielinska S Karafova A Kozik A Molecular determinants of the interaction between human high molecular weight kininogen and Candida albicans cell wall: identification of kininogen-binding proteins on fungal cell wall and mapping the cell wall-binding regions on kininogen molecule Peptides 2011 32 12 2488 2496 10.1016/j.peptides.2011.10.021 22074954
Karkowska-Kuleta J, Kulig K, Karnas E, Zuba-Surma E, Woznicka O, Pyza E, Kuleta P, Osyczka A, Rapala-Kozik M, Kozik A (2020) Characteristics of extracellular vesicles released by the pathogenic yeast-like fungi Candida glabrata, Candida parapsilosis and Candida tropicalis. Cells 9(7). 10.3390/cells9071722
Khalaf D, Krüger M, Wehland M, Infanger M, Grimm D (2019) The effects of oral L-arginine and L-citrulline supplementation on blood pressure. Nutrients 11(7). 10.3390/nu11071679
Kong LX Wang Z Shou YK Zhou XD Zong YW Tong T Liao M Han Q Li Y Cheng L Ren B The FnBPA from methicillin-resistant Staphylococcus aureus promoted development of oral squamous cell carcinoma J Oral Microbiol 2022 14 1 2098644 10.1080/20002297.2022.2098644 35859766
Lee Y Puumala E Robbins N Cowen LE Antifungal drug resistance: molecular mechanisms in Candida albicans and beyond Chem Rev 2021 121 6 3390 3411 10.1021/acs.chemrev.0c00199 32441527
Li DD Yang CC Liu P Wang Y Sun Y Effect of nitric oxide on the antifungal activity of oxidative stress and azoles against Candida albicans Indian J Microbiol 2016 56 2 214 218 10.1007/s12088-016-0580-x 27570314
Liebana-Jordan M, Brotons B, Falcon-Perez JM, Gonzalez E (2021) Extracellular vesicles in the fungi kingdom. Int J Mol Sci 22(13). 10.3390/ijms22137221
Lohse MB Gulati M Johnson AD Nobile CJ Development and regulation of single- and multi-species Candida albicans biofilms Nat Rev Microbiol 2018 16 1 19 31 10.1038/nrmicro.2017.107 29062072
Longo LM Despotović D Weil-Ktorza O Walker MJ Jabłońska J Fridmann-Sirkis Y Varani G Metanis N Tawfik DS Primordial emergence of a nucleic acid-binding protein via phase separation and statistical ornithine-to-arginine conversion Proc Natl Acad Sci U S A 2020 117 27 15731 15739 10.1073/pnas.2001989117 32561643
Luiking YC Poeze M Ramsay G Deutz NE The role of arginine in infection and sepsis JPEN J Parenter Enteral Nutr 2005 29 1 Suppl S70 74 10.1177/01486071050290s1s70 15709548
Martínez-López R Hernáez ML Redondo E Calvo G Radau S Pardo M Gil C Monteoliva L Candida albicans hyphal extracellular vesicles are different from yeast ones, carrying an active proteasome complex and showing a different role in host immune response Microbiol Spectr 2022 10 3 e0069822 10.1128/spectrum.00698-22 35604172
Moncada S Higgs A The L-arginine-nitric oxide pathway N Engl J Med 1993 329 27 2002 2012 10.1056/nejm199312303292706 7504210
Nimrichter L Rodrigues ML Fungal glucosylceramides: from structural components to biologically active targets of new antimicrobials Front Microbiol 2011 2 212 10.3389/fmicb.2011.00212 22025918
Oliveira DL Nakayasu ES Joffe LS Guimarães AJ Sobreira TJ Nosanchuk JD Cordero RJ Frases S Casadevall A Almeida IC Nimrichter L Rodrigues ML Characterization of yeast extracellular vesicles: evidence for the participation of different pathways of cellular traffic in vesicle biogenesis PLoS One 2010 5 6 e11113 10.1371/journal.pone.0011113 20559436
Palmer RM The L-arginine: nitric oxide pathway Curr Opin Nephrol Hypertens 1993 2 1 122 128 10.1097/00041552-199301000-00018 7522910
Pappas PG Rex JH Sobel JD Filler SG Dismukes WE Walsh TJ Edwards JE Guidelines for treatment of candidiasis Clin Infect Dis 2004 38 2 161 189 10.1086/380796 14699449
Pappas PG Kauffman CA Andes DR Clancy CJ Marr KA Ostrosky-Zeichner L Reboli AC Schuster MG Vazquez JA Walsh TJ Zaoutis TE Sobel JD Clinical practice guideline for the management of candidiasis: 2016 update by the infectious diseases society of America Clin Infect Dis 2016 62 4 e1 50 10.1093/cid/civ933 26679628
Popovic PJ Zeh HJ 3rd Ochoa JB Arginine and immunity J Nutr 2007 137 6 Suppl 2 1681s 1686s 10.1093/jn/137.6.1681S 17513447
Reales-Calderón JA Vaz C Monteoliva L Molero G Gil C Candida albicans modifies the protein composition and size distribution of THP-1 macrophage-derived extracellular vesicles J Proteome Res 2017 16 1 87 105 10.1021/acs.jproteome.6b00605 27740763
Rodrigues ML Nimrichter L Oliveira DL Frases S Miranda K Zaragoza O Alvarez M Nakouzi A Feldmesser M Casadevall A Vesicular polysaccharide export in Cryptococcus neoformans is a eukaryotic solution to the problem of fungal trans-cell wall transport Eukaryot Cell 2007 6 1 48 59 10.1128/ec.00318-06 17114598
Rodrigues ML Franzen AJ Nimrichter L Miranda K Vesicular mechanisms of traffic of fungal molecules to the extracellular space Curr Opin Microbiol 2013 16 4 414 420 10.1016/j.mib.2013.04.002 23628115
Roszkowiak J, Jajor P, Guła G, Gubernator J, Żak A, Drulis-Kawa Z, Augustyniak D (2019) Interspecies outer membrane vesicles (OMVs) modulate the sensitivity of pathogenic bacteria and pathogenic yeasts to cationic peptides and serum complement. Int J Mol Sci 20(22). 10.3390/ijms20225577
Rybak K Robatzek S Functions of extracellular vesicles in immunity and virulence Plant Physiol 2019 179 4 1236 1247 10.1104/pp.18.01557 30705070
Sandini S Stringaro A Arancia S Colone M Mondello F Murtas S Girolamo A Mastrangelo N De Bernardis F The MP65 gene is required for cell wall integrity, adherence to epithelial cells and biofilm formation in Candida albicans BMC Microbiol 2011 11 106 10.1186/1471-2180-11-106 21575184
Schaefer K, Wagener J, Ames RM, Christou S, MacCallum DM, Bates S, Gow NAR (2020) Three related enzymes in Candida albicans achieve arginine- and agmatine-dependent metabolism that is essential for growth and fungal virulence. mBio 11(4). 10.1128/mBio.01845-20
Stechmiller JK Childress B Cowan L Arginine supplementation and wound healing Nutr Clin Pract 2005 20 1 52 61 10.1177/011542650502000152 16207646
Szefel J Danielak A Kruszewski WJ Metabolic pathways of L-arginine and therapeutic consequences in tumors Adv Med Sci 2019 64 1 104 110 10.1016/j.advms.2018.08.018 30605863
Ullmann BD Myers H Chiranand W Lazzell AL Zhao Q Vega LA Lopez-Ribot JL Gardner PR Gustin MC Inducible defense mechanism against nitric oxide in Candida albicans Eukaryot Cell 2004 3 3 715 723 10.1128/ec.3.3.715-723.2004 15189992
van Niel G D'Angelo G Raposo G Shedding light on the cell biology of extracellular vesicles Nat Rev Mol Cell Biol 2018 19 4 213 228 10.1038/nrm.2017.125 29339798
Vargas G Rocha JD Oliveira DL Albuquerque PC Frases S Santos SS Nosanchuk JD Gomes AM Medeiros LC Miranda K Sobreira TJ Nakayasu ES Arigi EA Casadevall A Guimaraes AJ Rodrigues ML Freire-de-Lima CG Almeida IC Nimrichter L Compositional and immunobiological analyses of extracellular vesicles released by Candida albicans Cell Microbiol 2015 17 3 389 407 10.1111/cmi.12374 25287304
Wang Z Zhou Y Han Q Ye X Chen Y Sun Y Liu Y Zou J Qi G Zhou X Cheng L Ren B Synonymous point mutation of gtfB gene caused by therapeutic X-rays exposure reduced the biofilm formation and cariogenic abilities of Streptococcus mutans Cell Biosci 2021 11 1 91 10.1186/s13578-021-00608-2 34001238
Wenzel RP Nosocomial candidemia: risk factors and attributable mortality Clin Infect Dis 1995 20 6 1531 1534 10.1093/clinids/20.6.1531 7548504
Wu G Morris SM Jr Arginine metabolism: nitric oxide and beyond Biochem J 1998 336 ( Pt 1)(Pt 1) 1 17 10.1042/bj3360001 9806879
Wu G Meininger CJ McNeal CJ Bazer FW Rhoads JM Role of L-arginine in nitric oxide synthesis and health in humans Adv Exp Med Biol 2021 1332 167 187 10.1007/978-3-030-74180-8_10 34251644
Xiong K Zhu H Li Y Ji M Yan Y Chen X Chi Y Yang X Deng L Zhou X Zou L Ren B The arginine biosynthesis pathway of Candida albicans regulates its cross-kingdom interaction with actinomyces viscosus to promote root caries Microbiol Spectr 2022 10 4 e0078222 10.1128/spectrum.00782-22 35862976
Yang B Wang J Jiang H Lin H Ou Z Ullah A Hua Y Chen J Lin X Hu X Zheng L Wang Q Extracellular vesicles derived from Talaromyces marneffei yeasts mediate inflammatory response in macrophage cells by bioactive protein components Front Microbiol 2020 11 603183 10.3389/fmicb.2020.603183 33488545
Yawen Z Xiangyun C Binyou L Xingchen Y Taiping L Xuedong Z Jiyao L Lei C Wenyue X Biao R The dynamic landscape of parasitemia dependent intestinal microbiota shifting and the correlated gut transcriptome during Plasmodium yoelii infection Microbiol Res 2022 258 126994 10.1016/j.micres.2022.126994 35220138
Zamith-Miranda D Nimrichter L Rodrigues ML Nosanchuk JD Fungal extracellular vesicles: modulating host-pathogen interactions by both the fungus and the host Microbes Infect 2018 20 9–10 501 504 10.1016/j.micinf.2018.01.011 29471026
Zarnowski R Sanchez H Covelli AS Dominguez E Jaromin A Bernhardt J Mitchell KF Heiss C Azadi P Mitchell A Andes DR Candida albicans biofilm-induced vesicles confer drug resistance through matrix biogenesis PLoS Biol 2018 16 10 e2006872 10.1371/journal.pbio.2006872 30296253
Zarnowski R Noll A Chevrette MG Sanchez H Jones R Anhalt H Fossen J Jaromin A Currie C Nett JE Mitchell A Andes DR Coordination of fungal biofilm development by extracellular vesicle cargo Nat Commun 2021 12 1 6235 10.1038/s41467-021-26525-z 34716343
Zhou Y Yang H Zhou X Luo H Tang F Yang J Alterovitz G Cheng L Ren B Lovastatin synergizes with itraconazole against planktonic cells and biofilms of Candida albicans through the regulation on ergosterol biosynthesis pathway Appl Microbiol Biotechnol 2018 102 12 5255 5264 10.1007/s00253-018-8959-8 29691631
Zhou Y Cheng L Liao B Shi Y Niu Y Zhu C Ye X Zhou X Ren B Candida albicans CHK1 gene from two-component system is essential for its pathogenicity in oral candidiasis Appl Microbiol Biotechnol 2021 105 6 2485 2496 10.1007/s00253-021-11187-0 33635358
Zhu C Liao B Ye X Zhou Y Chen X Liao M Cheng L Zhou X Ren B Artemisinin elevates ergosterol levels of Candida albicans to synergise with amphotericin B against oral candidiasis Int J Antimicrob Agents 2021 58 3 106394 10.1016/j.ijantimicag.2021.106394 34197906
Zou S Wang X Liu P Ke C Xu S Arginine metabolism and deprivation in cancer therapy Biomed Pharmacother 2019 118 109210 10.1016/j.biopha.2019.109210 31330440
| 36441207 | PMC9703431 | NO-CC CODE | 2022-12-15 23:15:19 | no | Appl Microbiol Biotechnol. 2023 Nov 28; 107(1):355-367 | utf-8 | Appl Microbiol Biotechnol | 2,022 | 10.1007/s00253-022-12300-7 | oa_other |
==== Front
J Mater Cycles Waste Manag
J Mater Cycles Waste Manag
Journal of Material Cycles and Waste Management
1438-4957
1611-8227
Springer Japan Tokyo
1554
10.1007/s10163-022-01554-y
Review
4R of rubber waste management: current and outlook
Leong Seng-Yi 1
Lee Siang-Yin 2
Koh Thiam-Young 1
http://orcid.org/0000-0002-4361-3638
Ang Desmond Teck-Chye [email protected]
3
1 Tunku Abdul Rahman University of Management and Technology, Jalan Genting Kelang, Wilayah Persekutuan Kuala Lumpur, 53300 Kuala Lumpur, Malaysia
2 grid.466894.1 0000 0001 0274 7114 Technology and Engineering Division (BTK), RRIM Sungai Buloh Research Station, Malaysian Rubber Board (MRB), 47000 Selangor, Sungai Buloh Malaysia
3 grid.10347.31 0000 0001 2308 5949 Department of Chemistry, Faculty of Science, Universiti Malaya, 50603 Kuala Lumpur, Malaysia
28 11 2022
115
20 5 2022
18 11 2022
© Springer Japan KK, part of Springer Nature 2022, Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Excessive accumulation of rubber waste necessitates the need to revisit the effectiveness of the existing rubber waste management system. This review provides an overview of the legislative frameworks, techniques, challenges, and trends of rubber waste management in various countries. The 4R (reduce, reuse, recycle and recover) framework applied in waste management system in some countries appears to be viable for the processing of rubber waste. Certain countries especially some of the European Union (EU) members have implemented extended producer responsibility (EPR) system to manage the collection of rubber waste, particularly used tires. The processing of rubber waste in each level of the 4R hierarchy was then discussed, with detailed elaboration on the most practiced ‘R’, recycling which encompasses the direct recycling of products, as well as material recycling via physical and/or chemical means. The challenges faced in the implementation of rubber waste management system in different countries were highlighted and recommendations for a more sustainable rubber consumption were provided at the end of this review.
Keywords
Rubber waste treatment
Waste management system
4R policy
Sustainable rubber consumption
Ministry of Higher Education MalaysiaFP037-2021 Ang Desmond Teck-Chye
==== Body
pmcIntroduction to 4R concept of rubber waste management
Rubber manufacturing and production remain strong throughout the years due to high market demand. Based on the statistic, the amount of natural rubber (NR) produced in 2015 has doubled compared to that in 2000. Thailand was the major NR exporting country in 2020, with export values totaling 3.5 billion U.S. dollars, and this accounted for about 30 percent of the global value of NR shipments that year [1]. Demand for rubber is expected to remain strong in the near future as rubber products are still very relevant and in need for various niche applications. An increase in the rubber production and consumption would inevitably increase the production of rubber waste, and improper handling or processing of the waste could harm the environment. This can be perceived as an obstacle to achieve Goal 12 of the United Nations’ Sustainable Development Goals (SDG12) on “Responsible Consumption and Production” which aims to achieve good management of waste and reduce its accumulation through the practice of 3R waste management [2]. Rubber products, especially the synthetic ones are quite resistant to natural degradation due to the vulcanization process [3]. The sulfur crosslinks formed between the polymer chains are meant to improve the mechanical strength, as well as the physical and chemical properties of the rubber. However, the complex three-dimensional structure of the vulcanized rubber also makes it more difficult to be decomposed via biodegradation. Besides, the presence of additives, antioxidants, and fillers also fosters its resistance towards biodegradation [3, 4].
The used tires are one of the largest sources of rubber waste in the environment due to the large proportion of raw rubber that goes into the tire production sector, which is about 70% of the annual NR production [5, 6]. It was forecasted that the amount of discarded waste tires each year could reach 1.2 billion pieces by the year 2030 [7]. Besides, the usage of rubber gloves is also on surge in light of the SARS-CoV-2 pandemic situation beginning March 2020 when personal protective equipment (PPE), such as masks and gloves are widely used. As of June 2020, it was estimated that approximately 129 billion pieces of face masks which ear loops are made of polyisoprene rubber [8], and 65 billion pieces of rubber gloves are disposed each month globally [9]. Other major sources of rubber waste include automotive components, such as shock absorbers and conveyor belts, and municipal solid wastes, such as clothing, shoe soles, toys, electric wires, sofas, and cushions [10].
Traditional methods of rubber waste management are usually landfill disposal or incineration. However, several negative environmental implications may result from the landfill disposal of rubber waste, and this includes the leaching of toxic substances and heavy metals into the groundwater. Besides, the stockpile may cause inextinguishable fire as the rubber can serve as excellent fuel to sustain combustion [11]. An unstoppable fire incident that lasted for 18 days took place at a landfill site containing about 20.5 million kg of tires in Iowa City, Iowa, USA in 2012. The fire caused the release of excessive hazardous air pollutants, such as carbon dioxide, sulfur dioxide, fine particulate mass (PM2.5), polycyclic aromatic hydrocarbons (PAH), etc. [12, 13]. Incineration of rubber waste without proper control will also contribute to the emission of greenhouse gases (GHGs) [14].
To improve the handling of rubber waste, 4Rs (reduce, reuse, recycle, and recover) waste management was introduced by the European Commission on 19 November 2008 to be adopted as the common practice in waste processing [15, 16]. The distinguishable difference between the 4R Framework and the 3R Framework is the fourth R (Recover), which re-defines the incineration of waste materials in terms of the energy recovery efficiency [15]. The most prioritized item in the 4R hierarchy is “Reduce”, which refers to the practices that decrease the generation of waste, in this case, rubber waste [17, 18]. “Reuse” helps to prolong the lifetime of the rubber products for the same purpose of usage. “Recycling” may involve grinding of the rubber waste into small particles (ground rubber), or devulcanization that breaks the crosslinks between polymer chains, and re-purposed for other applications [19]. The last R, “Recover” usually indicates the process of decomposing the waste by heat into fuels and/or other valuable substances with lower molecular weights [19, 20].
Legislation framework and regulations on rubber waste management in different countries
Intervention from federal and/or local governments is essential for effective rubber waste management, and many countries have implemented various policies and regulations on issues related to waste management. Malaysia, being the fifth largest country in the production and export of NR products has a clear legislation framework on waste management [21], shown in Fig. 1. The Solid Waste Management and Public Cleansing Act (Act 672) was implemented in 2011 to improve the waste management service in seven states in the country [23]. Bursa Malaysia has also mandated its members to perform Corporate Social Responsibility Reporting (CSRR) annually from 2007 onwards, which includes concerns on pollution control and waste management, with the hope to better manage the environmental health [24]. In 2010, approximately 245,000 tons of used tire was generated, but its flow distribution was not recorded and there is lack of update in recent years [22].Fig. 1 Government agencies regulations of rubber waste management (particularly tire products) [22]
In the European Union (EU), the member states were obliged to manage rubber waste, specifically waste tires under two EU policies which are the 1999 Directive on the Landfill of Waste 1000/31/EC, and the end-of-life vehicle directive 2000/53/EC [25, 26]. Under these policies, stockpiling of waste tires other than bicycle tires and those with an external diameter > 1.4 m were prohibited, and the tires must be removed from the end-of-life vehicles for recycling [26]. The policy is still in operation and updated to raise the reuse and recycling rate from 80% in 2006 to 85% beginning January 2015, and the rate of reuse and recovery was raised from 85 to 95% [27]. However, the outline of the recycling procedure of the tires was not provided in these policies. Consequently, each member state of the EU had developed its regulations for its operations [26]. This in turn leads to the development of three different systems: extended producer responsibility (EPR), tax system, and free market system [28]. Amongst the three systems, the EPR system which requires the producers or importers to bear the responsibility for waste-tire recycling is adopted in most of the member states in the EU [28]. Implementation of EPR in Portugal has successfully increased the waste recycling rates from 69 to 98%, although improvement in consistency of the management was needed [29]. A comparative study on waste tire management between Italy and Romania showed that different pathways of management were attributed to the different regulations and economic context of the two countries [30]. EPR system adopted in both countries had led to the collection of used tires amounting to 403,000 tons for Italy and 46,000 tons for Romania in 2012. 18.1% of the collected tires in Italy were reused, while the remaining was distributed to recycling (36.1%), energy recovery (57.9%), and landfilling (6%). The end-of-life tires (ELTs) in Romania were channeled to recycling (43.5%) and energy recovery (56.5%) [30]. A study on the comparison of the EPR system between Belgium, The Netherlands, and Italy was also conducted, and the findings showed that EPR helps to increase resource efficiency, reduce illegal stockpiling, and move up the waste hierarchy [31]. For Belgium, a report published by the Public Waste Agency of Flanders stated that 51,375 tons of waste were collected in 2012 in the Flanders region, plus a collection rate of 88% was attained in the previous year. Of all the collected tires in Belgium, 85.0% were directed for recycling and reuse, while 15.0% were intended for energy recovery [32]. Poland which adopted a free market system in waste management implemented the Act of 11 May 2001 which required the entrepreneurs to manage their wastes by means of recovery and recycling [33, 34]. Waste tires were included under this act and the percentage of recovery and recycling was set at 75 and 15%, respectively in 2011 [33, 35]. A recent report in 2017 found that in 2014, the recovery rate of the collected tires had reached 76.3% while the recycling rate achieved 24.6%, which satisfied the set requirements [36]. It was; however, unclear if the recovery rate reported is inclusive of the recycling of rubber materials or exclusive to energy recovery.
Japan has implemented various laws for environmental protection, laws to reduce the consumption of natural resources and its environmental burden, as well as laws for proper waste management by practicing 3Rs in treating municipal and industrial wastes [37, 38]. End-of-life vehicle (ELV) recycling act was implemented in Japan in 2005 [37, 39], and the recycling of ELV follows the EPR concept whereby the responsibilities of ELV processing are on the manufacturers and importers using the fee collected from the vehicle owners [39]. A report in 2020 stated that Japan has collected 937,000 tons of ELTs, in which energy recovery was the dominant process of managing the waste tires, amounting to about 65% from the total [40]. In South Korea, the Plastic Waste Control Plan (PWCP) was established in 2018 by the Korean Ministry of Environment based on the Resource Circulation Act (RCA) and Resource Circulation Master Plan (RCMP). The objective of the plan was to reduce at least 50% of the plastic waste and recycle over 70% of the plastic waste produced by 2030, wherein synthetic rubber waste was included in the term ‘plastic waste’ [41]. South Korea recycled and recovered 70% of the waste tires generated per annum (500,000 metric tons) as tire-derived fuel (TDF) to emphasize energy recovery and control emissions. Collaboration from the relevant authorities with Sumitomo Heavy Industries was made to set up two TDF co-firing Circulating Fluidized Bed Combustion (CFBC) boiler plants to rely less on the use of coal while encouraging the involvement of TDF for power and heat generation [42].
In Turkey, Environmental Law (EL), and Regulation on Control of the End-of-Life Tires (RCELT) are two regulations that are responsible to control and manage the ELTs [34]. The RCELT helps to set up a transport system to collect the ELT. Under this regulation, proper management of ELTs was practiced in which recovery is strongly encouraged while import, disposal at landfill, or open burning were prohibited [34]. Recorded statistics showed that 72.4 kilotons (kt) of ELTs were collected, of which 33.4 kt was recycled and 33.9 kt was recovered in cement plants [34]. Colombia is one of the countries in South America that has adopted EPR as well, since 2007. The implementation of EPR was only partially successful due to governance issues, although the amount of recycled tire wastes did increase following the implementation. The responsibilities of waste tire management, both operational and financial were imposed on the tire producers alone, causing the lack of participation from other players in the tire production chain [43]. A total of 45,800 tons ELTs was collected in the year 2016, of which the main processing method was recycling (70%) followed by recovery (22%) and reusing (8%) [43].
Canada collected an estimate of 508 kt waste rubber tires in 2019, and about 450 kt of ELTs (date not specified) was recycled [44]. EPR was practiced in Canada, and there were more than ten provincial and territorial EPR provincial tire stewardship programs. Waste Diversion Act (WDA) was also enacted to supervise the diversion of target waste streams away from landfills [45]. Such scheme and the cooperation among local authorities managed to divert bulk of the waste tires from landfills, with the average diversion rate of about 98% over the past ten years [46]. In Australia, 69% of 466,000 metric tons of ELTs was managed through reusing (26.3%), recycling into tire-derived products (58.8%), and recovery process (14.9%), respectively [47]. Although the figures reported are rather impressive, Belgium had however outperformed Australia in terms of environmental performance by a factor of 7.9 [48]. Two reasons were provided for the observation; first, the regulatory framework in Australia favors the profit-driven automotive recycling industries which paid more attention to low costing materials, and secondly, the voluntary-based waste policy caused price competition between legitimate and illegitimate sectors. In turn, the non-adherence to the environmental standards and competitive prices offered by some of the illegitimate recyclers could demotivate the legitimate recyclers in the movement of proper ELT treatment [48]. To mitigate this problem, the enactment of ELV legislation was refined in 2021 to prohibit the exportation of unprocessed ELTs and maintain the exportation right exclusive to the licensed exporters [49].
Moving the focus to Brazil, only 10% recycling rate was achieved for annual tire disposal weighing 300 tons despite various enacted laws and regulations [50]. One of the enacted regulations is The Normative Instruction N °001/2010 from the National Council of the Environmental of Brazil (CONAMA), under which the manufacturers and importers are obligated to recycle 100% of the outstanding tires in Brazil, and companies are to bear the responsibilities of ELTs and their destinations [51]. Besides, Brazil has also enacted Publication of Resolution No. 416/09 which specifies the installation of collection points at a frequency of 1 to 100,000 inhabitants in cities to aid the collection of wastes. It however did not produce a practical effect as there were only 4 out of 120 proposed collection points being set up in São Paulo [51]. There is no denying that problems like poor infrastructure, insufficient funding, and attitudinal behavior of the citizen and corporates that are driven by profitability have limited the effectiveness of waste management in many countries, including Brazil [52].
Waste tire management in mainland China appears to be somewhat effective with a recycling rate of about 30%, comparable with countries such as Japan and Poland. However, a report by W. Huang in 2020 suggested that that the figure could have been higher in China and pointed out to several factors that may have retarded the recycling rate, and this includes the lack of regulations, government support, and relevant agencies. It was stated that the recycling system of car tires worked more effectively than that for tires from bicycles and electric motorbikes. The popularity of bicycles and motorbikes in China could have therefore contributed to the country’s lower than expected rubber recycling rate [53]. The increase in disposable income and the reduction in the price of tires contributed to the surge in waste tires from bicycles and electric motorbikes, which exceeded the number of waste tires from cars. The author commented that Vietnam also experienced a similar problem due to higher usage of motorbikes than cars [53]. There is a lack of intervention from government agencies and the recycling of rubber waste is mostly managed by the private sector [54]. In Anh’s findings in 2015, Vietnam generated about 400,000 tons of waste tires per annum. 50% of it was disposed or landfilled, while 40% was converted into thermal energy, and 10% for recycling and reuse [54]. In Taiwan, the Environmental Protection Agency established the Recycling Fund Management Board in 1998 which is responsible for inspection and management of the waste tire recycling process [55]. Collected waste tires were converted through shredding and grinding into rubber chips or powder for recycling or recovery purposes. It was recorded that in 2012, 80% of the rubber pieces from the waste tires was channeled as the auxiliary fuel source, while 15% was for the rubber materials, and 5% was directed to pyrolysis [55].
Techniques in rubber waste management
There are several techniques that have been reported in rubber waste management, and this includes ways to reuse rubber, rubber product and material recycling, as well as recovering energy from the rubber waste. A comparison of the pros and cons of each reported technique is shown in Table 1.Table 1 Comparison of various techniques in rubber waste management [26, 55–60]
Techniques Pros Cons
Retreading Reduce the utilization of rubber resources Require high operator’s technical level
Reduce rubber waste generation
Require high investment on retreading equipment
Product recycling Recycling of entire tires without any treatments Possible risk of leaching of additives and degraded materials into the environment
Pyrolysis Economically competitive Yield fuel may lead to engine performance problem (due to high sulfur, ash, and char content)
Produce wide range of products (hydrocarbons, CB, steel wire, etc.)
Material recycling Retrieve rubber material for the use in production of composite material/blending Requires shredding and granulating equipment
Downsizing Requires controlled conditions to prevent degradation of properties
Reclamation/Devulcanization Chemical usage may cause toxicity or pollution
Retreading and reuse
Retreading technology is applied on the end-of-life tires where the worn treads are replaced with new treads to allow the tires to be reused, and such technology prevents the piling up of waste tires in the environment [61, 62]. The general process involved in the retreading of a used tire is shown in Fig. 2. The carcass of the waste tire will first be inspected for reusability, after which the grinding and repairing procedure will follow to remove the tire crown from the carcass. After repairing, a piece of buffer rubber will be attached prior to the pasting of pre-vulcanized tread rubber. The last step of the process is the vulcanization of the pasted rubber tread. Finally, the products are inspected before they are released into the market [57].Fig. 2 General flow of tire retreading process [57]
Product recycling
Direct recycling of used tires while retaining their original forms was seen in various applications. These include the use in boats as fenders, scrap tires for the construction of artificial reef, as insulation for the foundation of buildings, etc. [26]. Some of the applications are however not widely practiced in recent times due to the adverse impacts they had on the environment. For instance, the initiative of using waste tires to construct an artificial reef in Florida in 1967, with the intention to create a marine habitat was unsuccessful. The dispersion of the waste tires brought by the water currents and storms damaged the coral reef, leading to an extra cost of up to 30 million USD to remove 2 million pieces of waste tires from the coast [58].
Pyrolysis
Pyrolysis normally refers to the breakdown of the rubber waste via incineration in an anaerobic condition, into smaller compounds such as fuel oil, gas, carbon black (CB), sulfur, and metal [63, 64]. The fuel oil produced could be refined to filter out the sulfur, char, and ash to ensure better engine performance. The gas can be used to drive the generation of heat and electricity in power plants [65], meanwhile the CB can be blended into plastic, EVA foam, or converted into activated carbon [55].
Material recycling
(a) Downsizing rubber particles
Description of different methods of downsizing rubber material and the characteristics of the ground rubber particles produced from each method are given in Table 2. Each of the methods has its own pros and cons. Dry ambient grinding of rubber is a relatively simple process that is carried by physical grinding of the rubber at ambient temperature, and the crumbs produced have high surface area to volume ratio. However, the friction between the rubber surface and grinder generates high amount of heat, and the temperature of the rubber may reach as high as 130 °C, resulting in oxidation of the rubber surface. Another relatively simple method, wet ambient grinding is also able to produce ground particles with a high surface area to volume ratio, but such method requires a long post-grinding drying period. Rubber can also be downsized using high water pressure jet, and such technique is environmentally friendly and cost-effective. However, highly trained personnel are needed to ensure smooth operation of the process. Berstoff’s method is an effective method to produce rubber particles with small grain size, but such method requires invested facilities such as an industrial miller and twin-screw extruder. Finally, the cryogenic method has also been adopted to produce ground rubber particles, and one apparent advantage of the technique is the ability to prevent surface oxidation on the rubber. This method is however quite costly due to the high consumption of liquid nitrogen in the process [59].Table 2 Different downsizing methods used in recycling rubber waste [59]
Methods Description Characteristics of the ground rubber particles
Dry ambient Repeated grinding of rubber waste until crumb rubber is obtained Size approximately 300 µm rough and irregular surface
Wet ambient Grind suspension of shredded rubber in the presence of water as coolant Size approximately 100 µm rough and irregular surface
Water jet Uses pressurized water jet (> 2000 bar) to strip and breaks rubber Rough and irregular surface
Berstoff’s method Uses the combination of rolling mill and twin-screw extruder Rough and irregular surface
Cryogenic Rubber cooled in liquid nitrogen and shattered using impact type mill Size approximately 75 µm sharp edge, flat/ smooth surface
The crumb rubber produced could regain its application in different industries. Oil absorptivity of a blend produced from mixing crumb rubber with 4-tert-butylstyrene (tBS) was reported by Wu and Zhou. Results from their study showed that the blend ratio of 60% waste tire crumb rubber to 40% tBS could attain a maximum oil sorption capacity of 24 g/g, a value which is slightly lower than the commercial counterpart used for benchmarking (30 g/g) [66]. A separate study was conducted by Odeh and Okpaire in 2020 to investigate the effect of rubber particle size on the oil absorptivity. They concluded that the optimized mesh size of 0.15 mm could achieve oil absorptivity of 4.71 g/g due to its great surface area to volume ratio [67]. Considering the reusability of the waste tire powder of up to 100 times, the total amount of oil that can be absorbed throughout the service time of the waste tire powder could be more than 200 g/g, a value which is higher than some of the single use commercially available oil sorbent [68]. A barrier made of recycled tires performed the best in soundproofing among all other materials such as concrete, metal, and wood. The barrier effectively reduced the noise level of a residential area near highways from 90 to 53 dBA [69]. Preparation of stone mastic asphalt added with crumb rubber and limestone yielded similar volumetric and mechanical properties, and a superior feature of reducing tire-road noise level as compared to the standard mixture [70]. Incorporation of crumb rubber was also proven to constrict the emission of CO and CH4 significantly [71].
(b) Reclamation and devulcanization of rubber waste
The principle of rubber reclamation and devulcanization lies in the breaking of crosslinks between the polymeric chains of the rubber. It should be made clear that reclamation and devulcanization are distinct from each other in terms of definition. Devulcanization refers to the scission of crosslinks, such as C-S and S–S bonds, whereas reclamation refers to the scission of both the crosslinks and the main chain bonds [72]. The physical method of rubber devulcanization takes the advantage of overcoming the lower bond energies of C-S bond (310 kJ mol−1) and the S–S bond (270 kJ mol−1), while rubber reclamation involves an additional C–C bond which have higher bond energy, 370 kJ mol−1 [73]. In reality however, the main chain scission could still occur in the devulcanization process due to the fact that the carbon–carbon bonds are present in higher abundance than the crosslinking bonds [74]. Different techniques of rubber reclamation, as well as their advantages and disadvantages, are summarized in Table 3.Table 3 Comparison of different reclamation or devulcanization techniques [74–81]
Techniques Advantages Disadvantages
Physical
Thermomechanical High selectivity for crosslink scission Potential main chain scission due to high abundancy of carbon–carbon bond
Easy to scale up for industrial application
Microwave Homogenous heating environmentally friendly Restricted to polar rubber molecule which absorbs microwave, otherwise requires suitable additives
Ultrasound Efficient process that can takes place within a few seconds High energy input required
Costly
Chemical Wide selection of devulcanizing chemicals and high efficiency Harmful to environment (secondary pollution)
Biological High selectivity for crosslink bond by desulfurization enzymes Microbial degradation is very slow. Devulcanization only happen on rubber surface
(i) Thermomechanical method
In thermomechanical devulcanization, the breaking of crosslinks is accomplished using a combination of high temperature and shear force. Many different equipment and techniques used for thermomechanical devulcanization have been reported. Formela et al. in 2014 found out that an increase in the treatment temperature decreases the screw torque required to reclaim ground tire rubber (GTR), and the properties of the reclaimed rubber are also affected by the temperature used. Out of three treatment temperatures (60, 120, and 180 °C), the one produced at 60 °C had the best mechanical properties, suggesting it experienced the least main chain scission. GTR which were reclaimed at 180 °C however incorporates better and more homogeneously in other polymeric matrices such as in styrene-butadiene rubber (SBR) blend to give better mechanical properties [74]. In 2019, Seghar et al. studied the effect of temperature, from 80 to 220 °C on the devulcanization of NR by feeding it at a rate of 5 kg/h into an industrial screw extruder with a screw speed of 240 rpm. They discovered that the degree of reclamation can achieve approximately 90% irrespective of the temperatures, but 80–100 °C range was reported to yield the best devulcanized rubber quality, presumably due to the highest tendency of sulfur bond scission in that temperature range. It was reported that NR has the tendency to undergo self-heating when the shear strain was applied to it, leading to the cleavage of the sulfur bonds. They concluded that relatively low energy consumption is needed to promote the recycling of NR [82]. Another study by Piszczyk et al. in 2015 was conducted to investigate the effect of GTR produced by thermomechanical reclamation on the properties of polyurethane foam. Results showed that the incorporation of reclaimed GTR by up to 30 wt% had resulted with an increase in apparent density, from 290 to 315 kg/m3, and increased the initial degradation temperature of the polyurethane composite by 14 °C. Blending the polyurethane foam with GTR however, did not affect the compressive strength of the composite [83].
(ii) Microwave method
This method utilizes the heating effect of the microwave to cleave the crosslinks in the rubber, provided that the rubber is sufficiently polar to absorb the microwave radiation [60]. For rubber with a poor absorptivity of the radiation such as NR, SBR, and ethylene-propylene-diene rubber (EPDM), polar fillers which absorbs microwave radiation such as silica and CB could be incorporated to improve the devulcanization process [78, 84]. This was reported in a study carried out by de Sousa et al. in 2015 to determine the effect of CB content and length of exposure time to microwave on the degree of devulcanization. It turned out that the gel content of the devulcanized NR showed negative correlations to both variables [84]. Moreover, the efficiency of devulcanization via microwave irradiation also depends on the devulcanizing agent and types of oil used in the process. In 2016, Molanorouzi and Mohaved published their studies on the reclamation of rubber waste using various devulcanizing agents and 2 types of oils (paraffinic and aromatic). Of all the agents, compositions, and conditions experimented, the best result was obtained by using 30 phr of aromatic oil, 6 phr of diphenyl disulfide, DPDS and a temperature of 240 °C. Aromatic oil is preferred probably because it enhanced the solubility of DPDS in it. Interestingly, the polysulfidic crosslinks decreased greatly, but the monosulfidic linkage increased [85].
(iii) Ultrasonic method
The scission of crosslinks (C-S and S–S bonds) in vulcanized rubber can be accomplished by the vibrations induced by wave energy that leads to the formation of cavities in the polymer matrix [77]. A study by Isayev et al. in 2014 established a relationship between the particle size or surface area of the rubber and the efficiency of devulcanization. It was found that under the same amplitude of ultrasonic wave and temperature 250 °C, rubber of size 30 mesh experienced greater extent of devulcanization that led to lower gel content compared to that of size 10 mesh [77]. Another interesting research by Sun and Isayev in 2008 revealed a positive correlation between the ultrasonic amplitude with the degree of devulcanization. It was found that the addition of processing oil could aid the devulcanization of isoprene rubber (IR) and NR. They also carried out a comparison on the degree of devulcanization of CB-filled IR and NR, and the results obtained showed that the effect of CB loadings on devulcanization of both rubbers was different, suggesting the stereoregular structure of rubber could play a role in affecting the extent of devulcanization [86].
(iv) Chemical method
The chemical method of rubber reclamation involves the use of chemicals to break the crosslinks between rubber chains and/or to block the successful recombination of sulfur linkage between the polymer chains. Many types of chemicals have been reported to serve as effective devulcanizing agents for rubber, and they include sulfides, peroxides, amines, deep eutectic solvents, and ionic liquids.
The use of diphenyl disulfide, DPDS as a devulcanizing agent was reported by Vega et al. in 2007, during which they performed devulcanization of rubber using a microwave coupled with DPDS. The incorporation of DPDS facilitated the devulcanization process, evidenced by the evolution of a high amount of squalene, and triterpene [87]. In 2012, Jiang et al. devulcanized butyl rubber with different amounts of DPDS, from 0 to 5 g, in combination with supercritical CO2 to swell the rubber vulcanizate. It was found that the sol fraction increased from about 15 to 98.5% when the amount of DPDS was raised from 0 to 4 g, indicating that the DPDS was involved in the devulcanization process rather than being the result of thermal degradation alone [88]. Thiosalicylic acid was used by Thaichaoroen et al. to reclaim the NR vulcanizate through the mechano-chemical method. The NR vulcanizate was heated at 140 °C for 30 min with different loading of thiosalicyclic acid after it was milled at room temperature. The result indicated that 1 phr thiosalicylic acid produced the optimal devulcanization effect and they claimed that the thiosalicylic acid had comparable devulcanization efficiency as the DPDS [80].
Zhang et al. in 2021 reported the use of hydrogen peroxide (H2O2) to reclaim SBR. The SBR was turned into powder and blended with soybean oil and H2O2 at different ratios. The mixed sample was then heated in a drying oven at 100 °C for 4 h with consistent air supply. After 4 h, the experiment yielded 100% sol fraction, suggesting complete devulcanization of the rubber [89]. Another study by Sabzekar et al. in 2015 investigated the effect of reaction time, temperature, and the amount of benzoyl peroxide as devulcanizing agent on the devulcanization efficiency of sulfur-cured NR. It was shown that cleavage of the crosslink bonds was achieved at lower concentrations of the benzoyl peroxide, from 2 to 8 phr, and a further increase in the concentration of the peroxide led to non-selective cleavage. A shorter reaction time of 2 h was recommended to prevent main chain scission, and it was found that a reaction temperature of 110 °C with low benzoyl peroxide content of up to 4 phr could drastically reduce the crosslink density of the rubber [90].
Walvekar et al. in 2018 performed the devulcanization of waste tire rubber with amines in combination with ultrasonic treatment at different temperatures and rubber: amine ratios. They found that tertiary amine (3-aminopropyltrimethoxysilane) produced better outcome compared to primary amine [(n-diethyl-3-aminopropyl) trimethoxysiloxane] at sonication temperature of 50 °C, with gel content of the treated rubber at 63–77% and 75–87%, respectively [91]. Walvekar et al. also investigated the use of deep eutectic solvent, DES as a devulcanizing agent. Rubber waste was devulcanized via ultrasonic method using zinc chloride: urea at the mole ratio of 2: 7 and 1: 4 at temperatures 30, 130, 150, and 180 °C. DES with ZnCl2: urea in (2: 7) ratio requires temperature of 130 °C for optimal devulcanization, and the higher temperature resulted in bond reformation that decreased the devulcanization efficiency. Temperature higher than 130 °C was however required for ZnCl2: urea ratio (1:4) mixture for better devulcanization effect. Based on the high sol fraction produced, > 85%, it was concluded that DES comprising of ZnCl2 and urea is very effective for desulfuration of rubber [92]. Pyrrolidinium hydrogen sulfate ionic liquid, IL was used in combination with microwave treatment by Seghar et al. in 2015 to devulcanize SBR. Different microwave energy from 0 to 440 Wh/kg and 10 wt % of the IL was used in the experiment. It was observed that the sol fraction is positively correlated to the microwave energy, and the addition of the IL further increased the sol fraction at microwave energy > 220Wh/kg, confirming the role of the IL as devulcanizing agent [93].
(v) Biological method
Unlike chemical method which deals with a variety of harmful chemical solvents, the biological degradation of rubber polymer offers a safer alternative as the devulcanization process is catalyzed by enzymes produced from microorganisms [76]. In 2012, Li et al. used Thiobacillus sp. for devulcanization of GTR and investigated the properties of NR/devulcanized GTR (NR/dGTR) composite. An increase in the wettability of the composite was observed in which the water contact angle decreased from 120.5 to 93.5°. This is attributed to about 30% increase in oxygen content in the dGTR after oxidative desulfurization by the Thiobacillus sp. The tear strength of the composite improved when the rubber waste content was below 30 phr due to better homogeneity, as well as due to the presence of high concentration of active sites for revulcanization to takes place [94]. Yao et al. reported the ability of Alicyclobacillus sp. to devulcanize rubber latex waste (prevulcanized) at a concentration of 5% (w/v) with the aid of Tween 80 as surfactant. A 62.5% decrement in the sulfur content and a 34.9% increment in oxygen content after 10 days of co-culturing were reported, pointing to successful oxidative cleavage of the sulfur crosslinks. This outcome was supported by the change of water contact angle from 104.3 to 85.0° which means an increase in the hydrophilicity of the latex rubber was achieved after desulfurization [95].
(vi) Supercritical devulcanization
Recent devulcanization technology uses CO2 at supercritical state (scCO2) as the reaction medium for devulcanization. CO2 is known for its non-flammability, non-toxicity, and chemical inactivity, all of which are desirable properties to ensure the safety of the process [96]. It was also reported to have a critical point that is easily achieved and good diffusivity which helps the swelling process for the diffusion of devulcanization agent into the rubber material [97]. scCO2 was used as a medium to devulcanize unfilled NR/SBR blend under non-isothermal conditions at 120 °C set-point temperature and 8 MPa pressure, with the use of bulb crude extract (BCE) from Tulbaghia violacea as the devulcanizing agent. It was reported that the use of 1.0 wt % of BCE managed to increase the sol content of the vulcanizate to a maximum of about 45% at reaction time t = 17 min, and this is a big improvement compared to the untreated control (~ 19%) [97]. Devulcanization of GTR was conducted by Mangili et al. in 2014 using DPDS in scCO2 at temperature 180 °C and pressure 15 MPa. The sol fraction of the GTR increased from 1.1 to 8.3 wt %, accompanied by a reduction in the crosslink density from 0.082 to 0.037. The changes were attributed to the crosslink scission whereby the sulfur crosslinks were broken and reacted with radicals produced from DPDS [98]. Liu et al. in 2015 investigated the parameters involved in devulcanization of waste tread rubber by using scCO2 as the medium. Their findings suggested that DPDS concentration of at least 10 g L−1, pressure below 7.38 MPa, and temperature of 140 °C are suitable for the scission of sulfur crosslinks in the waste tread rubber. scCO2 increased the cleavage of the crosslinks as it facilitated the devulcanization agent DPDS to reach the crosslink bonds [99]. In another study, GTR was produced by jet pulverization of scCO2, along with the addition of DPDS as the devulcanization agent. Application of scCO2 led to the swelling of the waste tire rubber, enhancing the penetration of the DPDS into the rubber for efficient devulcanization. Result from FTIR analysis showed that the use of scCO2 allowed the radical reactions to incorporate benzene rings and cleave the sulfur crosslinks [96].
Biodegradability of NR and its usefulness in biological recycling of waste rubber
NR, in chemistry context, is the polymeric product of the isoprene unit connected through carbon-1 and -4 at cis-configuration [100]. The raw material of NR can be identified from over 2000 plant species, but the major contributor of NR in the rubber industry comes from Hevea brasiliensis. The naturally occurring polymer has been reported to be degradable by native microbes.
Kasai et al. published an article in 2017 regarding the biodegradability of deproteinized natural rubber, DPNR by a soil bacterium, Rhizobacter gummiphilus NS21. NS21 was incubated on the DPNR-overlay agar plate for 3 days, followed by Schiff’s staining and gel permeation chromatography (GPC) analysis. The color development indicated the cleavage of the long poly(cis-1,4-isoprene) chain to produce short-chain products with terminal aldehyde groups. Result from GPC analysis agreed with the suggestion that R. gummiphilus degraded the DPNR, with notable decrease in the peak height at MW = 1300 kDa and increase in peak height at MW = 110 kDa [101, 102]. Degradation of various NR products inclusive of fresh latex stripes, latex condom, latex glove, and car tires was also observed after incubation with Streptomyces sp. CFMR7 which was isolated from rubber plantation site. Detection of the carbonyl group in the fungus-treated fresh latex pieces using FTIR-ATR spectroscopy underpinned the oxidative cleavage of the C = C bond by the Streptomyces sp. CFMR7. The transition of the MW distribution from higher to lower value was verified through GPC analysis, and collectively the results suggests that NR could be biodegraded by the microorganisms found in the native environment [103, 104].
Nguyen et al. in 2020 performed a study on the bacteria consortia enriched from rubber-processing factory’s waste in Vietnam to compare the biodegradation of NR and DPNR film. The consortia were enriched using the soil and wastewater samples from the factory and incubated with NR or DPNR with constant shaking at 150 rpm, and temperature 30 °C for 2 weeks. It was found that the highest weight loss achieved for NR was 35.4% which was lower than DPNR’s 48.4%, and this is due to the presence of protein as the natural antioxidant to protect the NR from oxidative degradation. The degradation of proteins by Proteobacteria and Bacteroidetes caused the NR to be susceptible to the chain cleavage reaction from microorganisms. Among the genus in the consortia, Gordonia and Mycobacterium in the phylum Actinobacteria were identified to be the main bacteria mediating the biodegradation reaction of the NR [105].
18 actinobacterial strains were reported by Basik et al. in 2021 to have the NR degradation capability through random screening on NR latex agar. Surprisingly, two rare strains related to Microtetraspora glauca (strain AC03309) and Dactylosporangium sucinum (strain AC04546) were discovered amongst the 18 strains. Both strains were tested for their capability to degrade latex gloves and tire granules by incubation in 50 mL of mineral salts medium, MSM and 0.5% (w/v) rubber material, with constant shaking at 180 rpm and temperature 28 °C for 30 days. SEM analysis showed that both strains were able to colonize and utilize the rubber material, among which the AC04546 strain showed a good utilization of rubber from tire samples. The authors suspected that the rubber-degrading ability of these strains may be stimulated by their exposure to the rubber particles deposited in the environment [106].
Streptomyces sp. AC04842 identified from a soil sample was found to be able to attach and colonize latex pieces, latex glove strips, and tire granules, a deduction made based on SEM images that shows the extension of mycelia to the surface of the materials. Degradation on the latex pieces and glove strips by Streptomyces sp. was recognized through the presence of holes and cavities in the polymer under SEM images after they were cultivated in the fungal culture for 60 days at 28 °C. Besides, the formation of oxygen-carrying bonds after incubation with Streptomyces sp. supported the authors’ claim that degradation process had taken place. [107]. Lactobacillus plantarum (strain LOCK 1145) is another bacterium that is capable of degrading NR, discovered by Olejnik et al. in 2022. In the experiment, various rubber materials were cultured with the bacteria at 30 °C for 14 days, and the average carbon content of the samples has reduced from the range of 89.1–95.4% to 56.5–65.7%. Pore formation observed in the NR vulcanizate along with the recorded mass loss of 1–5% strongly implies that the bacterial strain could degrade the rubber material [108].
Collectively, the findings summarized above suggest that natural rubber can be biodegraded by selected microbial strains. The ability of the microbes to degrade NR should be improved further through biotechnological research to enhance the biodegradation of waste rubber. While Hevea rubber offers the advantage of being an environmentally friendly source for rubber, the sole dependence on this tropical plant might not be the best option. Rapid expansion of Hevea rubber tree plantation may leave a negative environmental impact such as biodiversity loss and climate change [109]. Other sources of NR should be exploited and among the promising ones include those from Guayule or Kazakh dandelion. Although there are various challenges to the industrial production of NR from these plants, they are nevertheless candidates with great potential to increase the global natural rubber supply. These plants are native to temperate regions which potentially are the emerging sources of NR other than Hevea rubber [110].
Future outlook on rubber waste issues: prospects and challenges
The generation of rubber waste will continue to increase as the demand for rubber products continues to remain strong in the near future. Although the most common rubber product being discussed was rubber tires, the volume of rubber products in the medical field such as rubber gloves, ear-loop bands, and catheters are increasing significantly in light of the COVID-19 pandemic. Effective implementation of rubber waste management to resolve the growing amount of rubber waste requires close cooperation between the government and the industry stakeholders. Such cooperation is also crucial to ensure that sustainable development can take place parallel with economic development. Unfortunately, it remains a challenge to maintain a balance between economy and the environmental protection as these two aspects are often of causal relationship [111]. Poorer countries may not have much option but to prioritize the economic development of their respective countries over environmental preservation, especially during this time when the global economy is badly affected by the ongoing Covid-19 pandemic. The use of face mask and other rubber products such as rubber gloves led to the generation of additional rubber waste, which is normally incinerated or sent to licensed landfill sites after sterilization [112, 113]. It is likely that the current waste management system inclusive of rubber waste will remain as it is, and any improvement to the system will only take place when the global economy has recovered from the pandemic.
To develop an effective and efficient rubber waste management system, there are a few challenges that need to be considered. First, it is undeniable that the legislative framework in a country is important for good rubber waste management. The lack of laws and regulations in controlling the disposal of rubber waste, along with the lack of action in the waste management industry and support from the government are believed to be some of the main reasons behind the poor recycling and recovery of rubber from the waste. The case study in Colombia had shown that the government did not incentivize the waste processors and the end-users in their EPR model, and this has resulted in a lack of participation from the industry players and consumers in recycling the waste tires [43]. In contrast, the rubber waste management system in the EU seems to be relatively well managed, with the presence of EPR system in managing the collection and recycling of rubber waste. It is plausible that the system will continue to support the recycling of rubber waste products beyond waste tires in the future.
Rubber waste management is an industry where the initial capital needed is very high as it requires the construction of the recycling and recovery facilities such as the pyrolytic treatment plant which are often very costly. Since a huge amount of capital must be invested for the setting of the treatment plant, giant reactors, efficient rubber waste collection system, etc., not many entrepreneurs have the capacity to venture into the waste management business. In addition to the cost factor, the risk from the uncertain return of investment, ROI could shy away investors too. It is noteworthy that the quality of the recycled products may vary due to the differences in the composition of the rubber wastes, as well as due to the processing condition such as the temperature, pressure, catalysts, and types of reactors used [114, 115]. The inconsistency may affect the properties of the recycled products, and this makes commercialization much more difficult.
Conclusion and recommendations
4R waste management framework (Reduce, Reuse, Recycling, and Recover) seems to be a viable and practical approach to address the excessive rubber waste accumulation issue and relieve the environmental consequences brought by it. Out of the 4R, ‘Recycling’ is the most adopted practice, and material recycling which encompasses grinding of rubber waste into smaller particles, and devulcanization techniques to retrieve the rubber component appear to be some of the most applied techniques in the rubber waste management industry. This is probably because the techniques allow greater utilization of waste material, and the retrieved rubber can be used to generate added value products with decent qualities.
Based on our findings from reviewing the current development in rubber waste management and its related issues, the followings are some of the recommendations to address the issues and promote more sustainable rubber consumption:The policy for rubber waste management should be carefully devised to increase the recycling rate. The rubber waste management model and relevant legislations that have been successfully implemented in some of the EU member countries should be adopted by others to improve the rubber recycling rate. From economic standpoint, rubber waste management is a challenging industry that requires high capital, high operating cost, and uncertain return on investment, leading to a lack of participation from the industry players and investors. Government should incentivize the industry to promote it and encourage more participation. Studies on the cost return of rubber waste management could be helpful for the government to devise a framework for rubber waste management system to maximize the profit from rubber waste recycling.
From a long-term perspective, it is important to establish rubber recycling and recovery systems in all countries. Considering the effectiveness and encouraging results reported in many countries, these initiatives should be prioritized in rubber waste management. However, in some of the developing countries which currently lack the necessary basic infrastructures for rubber recycling and recovery exercise, the use of NR may be considered over inert synthetic rubbers, where possible. As elaborated in the earlier section, natural rubber has the potential to be degraded by various native microbes leading to wide range of weight loss, while synthetic ones such as nitrile rubber tend to be highly resistant to biodegradation. Besides, prioritizing naturally derived products could help to reduce the dependency on petrochemicals which are non-renewable and known to add carbon footprint to the environment. However, more detailed studies on the effectiveness of biodegradation of NR in landfill and its potential impact to the surrounding ecosystem are necessary before any recommendation can be made on its suitability to be part of mainstream rubber waste management system. The effect of wide range of additives formulated in commercial NR products on their biodegradability need to be carefully evaluated too. For now, recycling and recovery remains as the best options when it comes to treatment of rubber waste.
Author contributions
All authors contributed to the study conception and design. The first draft of the manuscript was written by Seng-Yi Leong and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript. Supervision: Desmond Teck-Chye Ang and Siang-Yin Lee.
Funding
This project is funded by Fundamental Research Grant Scheme (FRGS/1/2021/STG04/UM/02/4) by Ministry of Higher Education Malaysia.
Data availability
The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.
Declarations
Conflict of interests
The authors have no relevant financial or non-financial interests to disclose.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
1. TiseoI (2022) Leading natural rubber exporting countries worldwide based on export value in 2020, by major country. In: Statista. https://www.statista.com/statistics/652791/leading-natural-rubber-exporters-by-country/. Accessed 7Aug2022
2. CernevT FennerR The importance of achieving foundational sustainable development goals in reducing global risk Futures 2020 115 102492 10.1016/J.FUTURES.2019.102492
3. Aboelkheir MG Bedor PB Leite SG Biodegradation of vulcanized SBR A comparison between bacillus subtilis 2019 Pseudomonas aeruginosa and Streptomyces sp Sci Reports
4. Fazli A Rodrigue D Waste rubber recycling: a review on the evolution and properties of thermoplastic elastomers Mater 2020 13 782 10.3390/MA13030782
5. Forrest MJ Recycling and re-use of waste rubber Recycl Re-use Waste Rubber 2019 10.1515/9783110644142
6. SchmidtM DG Palekhov D Hansmann B Sustainable global value chains 2019 Cham Springer International Publishing
7. RashadAM, A comprehensive overview about recycling rubber as fine aggregate replacement in traditional cementitious materials Int J Sustain Built Environ 2016 5 46 82 10.1016/J.IJSBE.2015.11.003
8. SelvaranjanK NavaratnamS RajeevP RavintherakumaranN Environmental challenges induced by extensive use of face masks during COVID-19: A review and potential solutions Environ Challenges 2021 3 100039 10.1016/J.ENVC.2021.100039
9. SilvaAL P PrataJC WalkerTR Increased plastic pollution due to COVID-19 pandemic: challenges and recommendations Chem Eng J 2021 405 126683 10.1016/J.CEJ.2020.126683 32834764
10. Abraham E, Cherian BM, AEP, et al (2011) Editors: A. Fainleib and O.Grigoryeva 2. Recent advances in the recycling of rubber waste. Transw Res Netw. 37:47–100
11. Mohajerani A Burnett L Smith JV Recycling waste rubber tyres in construction materials and associated environmental considerations: a review Resour Conserv Recycl 2020 155 104679 10.1016/J.RESCONREC.2020.104679
12. Downard J Singh A Bullard R Uncontrolled combustion of shredded tires in a landfill—Part 1: characterization of gaseous and particulate emissions Atmos Environ 2015 104 195 204 10.1016/J.ATMOSENV.2014.12.059
13. Singh A Spak SN Stone EA Uncontrolled combustion of shredded tires in a landfill—Part 2: population exposure, public health response, and an air quality index for urban fires Atmos Environ 2015 104 273 283 10.1016/J.ATMOSENV.2015.01.002
14. Ghorai S Mondal D Hait S Devulcanization of waste rubber and generation of active sites for silica reinforcement ACS Omega 2019 4 17623 17633 10.1021/ACSOMEGA.9B01424/ASSET/IMAGES/LARGE/AO9B01424_0009.JPEG 31681869
15. Grosso M Motta A Rigamonti L Efficiency of energy recovery from waste incineration, in the light of the new waste framework directive Waste Manag 2010 30 1238 1243 10.1016/J.WASMAN.2010.02.036 20347289
16. Kirchherr J Reike D Hekkert M Conceptualizing the circular economy: an analysis of 114 definitions Resour Conserv Recycl 2017 127 221 232 10.1016/J.RESCONREC.2017.09.005
17. Ichi SS Yano J Hirai Y Waste prevention for sustainable resource and waste management J Mater Cycles Waste Manag 2017 19 1295 1313 10.1007/S10163-017-0586-4/FIGURES/3
18. KholilA AJ Waste management based on 3r in mutiara waste banks Bekasi City Indonesia World Environ 2018 8 71 76 10.5923/J.ENV.20180803.02
19. Nuzaimah M Sapuan SM Nadlene R Jawaid M Recycling of waste rubber as fillers: a review IOP Conf Ser Mater Sci Eng. 2018 368 012016 10.1088/1757-899X/368/1/012016
20. Imbernon L Norvez S From landfilling to vitrimer chemistry in rubber life cycle Eur Polym J 2016 82 347 376 10.1016/J.EURPOLYMJ.2016.03.016
21. TiseoI (2022) Natural rubber top exporting countries distribution 2020. https://www.statista.com/statistics/652791/leading-natural-rubber-exporters-by-country/. Accessed 28Jul2022
22. Chemsain Konsultant Sdn Bhd (2011) A Study on Scrap Tyre Management for Peninsular Malaysia: Final Report. https://jpspn.kpkt.gov.my/resources/index/user_1/Sumber_Rujukan/kajian/Tyre Study_Final Report_Eng Version.pdf. Accessed 28Jul2022
23. Abas MA Wee ST Municipal solid waste management in malaysia: an insight towards sustainability SSRN Electron J 2014 10.2139/SSRN.2714755
24. Zainal D Zulkifli N Saleh Z Corporate social responsibility reporting in Malaysia: a comparison between Shariah and non-Shariah approved firms Middle East J Sci Res 2013 15 1035 1046 10.5829/IDOSI.MEJSR.2013.15.7.11034
25. Parliament E Directive 2000/53/EC of the European parliament and of the council of 18 September 2000 on end-of life vehicles Off J Eur Union 2000 279 34 43
26. Sienkiewicz M Kucinska-Lipka J Janik H Balas A Progress in used tyres management in the European Union: a review Waste Manag 2012 32 1742 1751 10.1016/J.WASMAN.2012.05.010 22687707
27. Karger-Kocsis J Mészáros L Bárány T (2012) Ground tyre rubber (GTR) in thermoplastics, thermosets, and rubbers J Mater Sci 2012 481 48 1 38 10.1007/S10853-012-6564-2
28. Dabic-Miletic S Simic V Karagoz S (2021) End-of-life tire management: a critical review Environ Sci Pollut Res 2021 2848 28 68053 68070 10.1007/S11356-021-16263-6
29. Niza S Santos E Costa I Extended producer responsibility policy in Portugal: a strategy towards improving waste management performance J Clean Prod 2014 64 277 287 10.1016/j.jclepro.2013.07.037
30. Torretta V Rada EC Ragazzi M Treatment and disposal of tyres: Two EU approaches A review Waste Manag 2015 45 152 160 10.1016/J.WASMAN.2015.04.018 25943287
31. WinternitzK HeggieM BairdJ, Extended producer responsibility for waste tyres in the EU: Lessons learnt from three case studies – Belgium, Italy and the Netherlands Waste Manag 2019 89 386 396 10.1016/J.WASMAN.2019.04.023 31079753
32. Public Waste Agency of FlandersO Extended producer responsibility . The case of used tyres 1 in
33. Adamczyk J, Gulba M, Sąsiadek M, et al. (2019) Rubber waste management. Sci Pap Silesian Univ Technol Organ Manag Ser. 10.29119/1641-3466.2019.137.1
34. Karaağaç B Ercan Kalkan M Deniz V End of life tyre management: Turkey case J Mater Cycles Waste Manag 2017 19 577 584 10.1007/S10163-015-0427-2/FIGURES/7
35. Alwaeli M End-of-life vehicles recovery and recycling and the route to comply with EU directive targets Environ Prot Eng 2016 42 191 202 10.5277/EPE160114
36. Godlewska J Recovery and recycling of waste tires in Poland Procedia Eng. 2017 182 229 234 10.1016/J.PROENG.2017.03.173
37. Ministry of the Environment of Japan (2014) History and Current State of Waste Management in Japan. 32
38. UNEP (2011) The Japanese Industrial: Waste Experience: Lessons for rapidly industrializing countries
39. Hiratsuka J Sato N Yoshida H Current status and future perspectives in end-of-life vehicle recycling in Japan J Mater Cycles Waste Manag 2014 16 21 30 10.1007/S10163-013-0168-Z/TABLES/3
40. JATMA (2021) Tyre Industry of Japan 2021, Report
41. Shin SK Um N Kim YJ New policy framework with plastic waste control plan for effective plastic waste management Sustain 2020 12 6049 10.3390/SU12156049
42. KleanIndustries Circulating Fluidized Bed Combustion Technologies | Klean Industries. https://kleanindustries.com/waste-processing-projects/tire-pyrolysis-recycling/kumho-petrochemical-korea/. Accessed 13Aug2022
43. Park J Díaz-Posada N Mejía-Dugand S Challenges in implementing the extended producer responsibility in an emerging economy: the end-of-life tire management in Colombia J Clean Prod 2018 189 754 762 10.1016/J.JCLEPRO.2018.04.058
44. Andrews L, Skinner D, Harding J, et al (2021) Final Report-Exploring Circular Economy for Rubber in Canada
45. Cocker J Graham K Circular economy in Canada Circular Economy: Global Perspective 2020 Singapore Springer Singapore 87 122
46. CATRACA of TRA CATRA: annual report 2020 AIMS Math 2021 6 14064 14068
47. TSATSA Tire Stewardship Australia (TSA) (2020). “69% of Australia’s end-of-life tires recovered for further use—2018–19 Australian Tire Consumption & Recovery”. https://www.tyrestewardship.org.au/news/69-of-australias-end-of-life-tyres-recovered-for-further-use-2018-19-australian-tyre-consumption-recovery/. Accessed 14Aug2022
48. Soo VK Peeters J Compston P Comparative study of end-of-life vehicle recycling in Australia and Belgium Procedia CIRP 2017 Netherlands Elsevier 269 274
49. Numfor SA, Halog A, Matsubae K (2022) A Review of End-of-Life Tire Recycling in Australia, Japan, South Africa and Cameroon Environmental Management and National Development Planning View project. 2
50. Bittencourt ES de Oliveira Fontes CH Rodriguez JLM Modeling the socioeconomic metabolism of end-of-life tires using structural equations: a brazilian case study Sustain 2020 12 2106 10.3390/SU12052106
51. de Oliveira Neto GC Chaves LEC Pinto LFR Economic, environmental and social benefits of adoption of pyrolysis process of tires: a feasible and ecofriendly mode to reduce the impacts of scrap tires in Brazil Sustain 2019 10.3390/SU1102076
52. Felix IA, Oluseyi AO, Oyawale F, Akinlabi SA. 2018 Sustainable End-of-Life Tyre (EOLT) Management for Developing Countries-A Review
53. Huang W Sustainable management of different systems for recycling end-of-life tyres in China Waste Manag Res 2021 39 966 974 10.1177/0734242X20976976 33342391
54. Anh P Options for environmental sustainability of scrap tire in vietnam from discharge to reuse and recylce: case study in the Southeast of Vietnam J Glob Ecol Environ. 2017 7 21 26
55. Tsai WT Chen CC Lin YQ Status of waste tires’ recycling for material and energy resources in Taiwan J Mater Cycles Waste Manag 2017 19 1288 1294 10.1007/S10163-016-0500-5/TABLES/3
56. de Sousa FDB Scuracchio CH Hu GH Hoppe S Devulcanization of waste tire rubber by microwaves Polym Degrad Stab 2017 138 169 181 10.1016/J.POLYMDEGRADSTAB.2017.03.008
57. Qiang W Li J Yunlong W Discussion on tire retreading and reuse technology IOP Conf Ser Earth Environ Sci. 2020 512 012146 10.1088/1755-1315/512/1/012146
58. Morley DM Sherman RL Jordan LKB Environmental enhancement gone awry: characterization of an artificial reef constructed from waste vehicle tires WIT Trans Built Environ. 2008 99 73 87 10.2495/CENV080071
59. Ramarad S Khalid M Ratnam CT Waste tire rubber in polymer blends: a review on the evolution, properties and future Prog Mater Sci 2015 72 100 140 10.1016/J.PMATSCI.2015.02.004
60. Asaro L Gratton M Seghar S Aït Hocine N Recycling of rubber wastes by devulcanization Resour Conserv Recycl 2018 133 250 262 10.1016/J.RESCONREC.2018.02.016
61. Dabić-Ostojić S Miljuš M Bojović N Applying a mathematical approach to improve the tire retreading process Resour Conserv Recycl 2014 86 107 117 10.1016/J.RESCONREC.2014.02.007
62. Ounsaneha W Buadit T Rattanapan C Assessment of human health impact based on life cycle assessment: a case study of Thai retread tire IOP Conf Ser Mater Sci Eng. 2020 773 012038 10.1088/1757-899X/773/1/012038
63. AlsalehA SattlerML Waste tire pyrolysis: influential parameters and product properties Curr Sustain Energy Reports 2014 1 129 135 10.1007/S40518-014-0019-0/TABLES/1
64. Deng N Wang WW Chen GW Pyrolysis characteristics of rubber compositions in medical waste J Cent South Univ 2013 20 2466 2471 10.1007/S11771-013-1758-6
65. Čabalová I Ház A Krilek J Recycling of wastes plastics and tires from automotive industry Polym 2021 13 2210 10.3390/POLYM13132210
66. Wu B Zhou MH Recycling of waste tyre rubber into oil absorbent Waste Manag 2009 29 355 359 10.1016/J.WASMAN.2008.03.002 18455384
67. Odeh AO Okpaire LA Modelling and optimizing the application of waste tyre powder (WTP) as oil sorbent, using response surface methodology (RSM) African J Heal Saf Environ. 2020 1 1 12 10.52417/ajhse.v1i2.75
68. Lin C Huang CL Shern CC Recycling waste tire powder for the recovery of oil spills Resour Conserv Recycl 2008 52 1162 1166 10.1016/J.RESCONREC.2008.06.003
69. Mokhtar FN Abdel Rehim IV Mahmoud EA Applicability of using recycled rubber-Tire materials for acoustic insulation in barriers of residential areas in Egypt ARPN J Eng Appl Sci. 2017 12 806 820
70. Sangiorgi C Tataranni P Simone A Stone mastic asphalt (SMA) with crumb rubber according to a new dry-hybrid technology: a laboratory and trial field evaluation Constr Build Mater 2018 182 200 209 10.1016/J.CONBUILDMAT.2018.06.128
71. Stout D, Stout D, Douglas PE, Carlson D. 2003 Stack Emissions With Asphalt Rubber A Synthesis of Studies
72. Shi J Jiang K Ren D Structure and performance of reclaimed rubber obtained by different methods J Appl Polym Sci 2013 129 999 1007 10.1002/APP.38727
73. Diaz R Colomines G Peuvrel-Disdier E Deterre R Thermo-mechanical recycling of rubber: relationship between material properties and specific mechanical energy J Mater Process Technol 2018 252 454 468 10.1016/J.JMATPROTEC.2017.10.014
74. Formela K Cysewska M Haponiuk JT Thermomechanical reclaiming of ground tire rubber via extrusion at low temperature: efficiency and limits J Vinyl Addit Technol 2016 22 213 221 10.1002/VNL.21426
75. Dubkov KA Semikolenov SV Ivanov DP Reclamation of waste tyre rubber with nitrous oxide Polym Degrad Stab 2012 97 1123 1130 10.1016/J.POLYMDEGRADSTAB.2012.04.006
76. Ghavipanjeh F Ziaei Rad Z Pazouki M Devulcanization of ground tires by different strains of bacteria: optimization of culture condition by taguchi method J Polym Environ 2018 26 3168 3175 10.1007/S10924-017-1169-0/TABLES/7
77. Isayev AI Liang T Lewis TM Effect of particle size on ultrasonic devulcanization of tire rubber in twin-screw extruder Rubber Chem Technol 2014 87 86 102 10.5254/RCT.13.87926
78. Movahed SO Ansarifar A Zohuri G Devulcanization of ethylene–propylene–diene waste rubber by microwaves and chemical agents J Elast Plast 2014 48 122 144 10.1177/0095244314557975
79. Pirityi DZ Pölöskei K Thermomechanical devulcanisation of ethylene propylene diene monomer (EPDM) rubber and its subsequent reintegration into virgin rubber Polym 2021 13 1116 10.3390/POLYM13071116
80. Thaicharoen P Thamyongkit P Poompradub S Thiosalicylic acid as a devulcanizing agent for mechano-chemical devulcanization Korean J Chem Eng 2010 274 27 1177 1183 10.1007/S11814-010-0168-9
81. Valdés C Hernández C Morales-Vera R Andler R Desulfurization of vulcanized rubber particles using biological and couple microwave-chemical methods Front Environ Sci 2021 9 271 10.3389/FENVS.2021.633165/BIBTEX
82. Seghar S Asaro L Rolland-Monnet M Aït Hocine N Thermo-mechanical devulcanization and recycling of rubber industry waste Resour Conserv Recycl 2019 144 180 186 10.1016/J.RESCONREC.2019.01.047
83. Piszczyk Ł Hejna A Formela K Effect of ground tire rubber on structural, mechanical and thermal properties of flexible polyurethane foams Iran Polym J (English Ed). 2015 24 75 84 10.1007/S13726-014-0301-4/TABLES/4
84. De Sousa FDB Scuracchio CH The role of carbon black on devulcanization of natural rubber by microwaves Mater Res 2015 18 791 797 10.1590/1516-1439.004915
85. Molanorouzi M Mohaved SO Reclaiming waste tire rubber by an irradiation technique Polym Degrad Stab 2016 128 115 125 10.1016/J.POLYMDEGRADSTAB.2016.03.009
86. Sun X Isayev IA Continuous ultrasonic devulcanization: comparison of carbon black filled synthetic isoprene and natural rubbers Rubber Chem Technol 2008 81 19 46 10.5254/1.3548195
87. Vega B Montero L Lincoln S Control of vulcanizing/devulcanizing behavior of diphenyl disulfide with microwaves as the heating source J Appl Polym Sci 2008 108 1969 1975 10.1002/APP.27578
88. Jiang K Shi J Ge Y Complete devulcanization of sulfur-cured butyl rubber by using supercritical carbon dioxide J Appl Polym Sci 2013 127 2397 2406 10.1002/APP.37542
89. Zhang Z Li J Wan C Understanding H2O2-induced thermo-oxidative reclamation of vulcanized styrene butadiene rubber at low temperatures ACS Sustain Chem Eng 2021 9 2378 2387 10.1021/ACSSUSCHEMENG.0C08867/SUPPL_FILE/SC0C08867_SI_001.PDF
90. Sabzekar M Chenar MP Mortazavi SM Influence of process variables on chemical devulcanization of sulfur-cured natural rubber Polym Degrad Stab 2015 118 88 95 10.1016/J.POLYMDEGRADSTAB.2015.04.013
91. Walvekar R Kunju K Saputra R Parametric study for devulcanization of waste tire rubber utilizing deep eutectic solvent (DES) MATEC Web Conf 2018 152 01005 10.1051/MATECCONF/201815201005
92. Walvekar R Afiq ZM Ramarad S Khalid S Devulcanization of waste tire rubber using amine based solvents and ultrasonic energy MATEC Web Conf 2018 152 01007 10.1051/MATECCONF/201815201007
93. Seghar S Aït Hocine N Mittal V Devulcanization of styrene butadiene rubber by microwave energy: effect of the presence of ionic liquid Express Polym Lett 2015 9 1076 1086 10.3144/EXPRESSPOLYMLETT.2015.97
94. Li Y Zhao S Wang Y Improvement of the properties of natural rubber/ground tire rubber composites through biological desulfurization of GTR J Polym Res 2012 19 1 7 10.1007/S10965-012-9864-Y/FIGURES/11
95. Yao C Zhao S Wang Y Microbial desulfurization of waste latex rubber with Alicyclobacillus sp Polym Degrad Stab 2013 98 1724 1730 10.1016/J.POLYMDEGRADSTAB.2013.06.002
96. Wang Z Pan C Hu Y High-quality ground tire rubber production from scrap tires by using supercritical carbon dioxide jet pulverization assisted with diphenyl disulfide Powder Technol 2022 398 117061 10.1016/J.POWTEC.2021.117061
97. Gumede JI Hlangothi BG Mabuto B Devulcanization of natural rubber/styrene-butadiene rubber unfilled blend in supercritical carbon dioxide using Tulbaghia violacea crude extract J Clean Prod 2022 362 132478 10.1016/J.JCLEPRO.2022.132478
98. Mangili I Collina E Anzano M Characterization and supercritical CO2 devulcanization of cryo-ground tire rubber: Influence of devulcanization process on reclaimed material Polym Degrad Stab 2014 102 15 24 10.1016/J.POLYMDEGRADSTAB.2014.02.017
99. Liu Z Li X Xu X Devulcanizaiton of waste tread rubber in supercritical carbon dioxide: operating parameters and product characterization Polym Degrad Stab 2015 119 198 207 10.1016/J.POLYMDEGRADSTAB.2015.05.017
100. Ferreira M Mendonça RJ Coutinho-Netto J Mulato M Angiogenic properties of natural rubber latex biomembranes and the serum fraction of Hevea brasiliensis Brazilian J Phys 2009 39 564 569 10.1590/S0103-97332009000500010
101. Kasai D Imai S Asano S Identification of natural rubber degradation gene in Rhizobacter gummiphilus NS21 Biosci Biotechnol Biochem 2017 81 614 620 10.1080/09168451.2016.1263147 28110615
102. Imai S Ichikawa K Muramatsu Y Isolation and characterization of Streptomyces, Actinoplanes, and Methylibium strains that are involved in degradation of natural rubber and synthetic poly(cis-1,4-isoprene) Enzyme Microb Technol 2011 49 526 531 10.1016/J.ENZMICTEC.2011.05.014 22142727
103. Nanthini J Sudesh K Biodegradation of natural rubber and natural rubber products by streptomyces sp. strain CFMR 7 J Polym Environ 2017 25 606 616 10.1007/S10924-016-0840-1/FIGURES/9
104. Chia KH Nanthini J Thottathil GP Identification of new rubber-degrading bacterial strains from aged latex Polym Degrad Stab 2014 109 354 361 10.1016/J.POLYMDEGRADSTAB.2014.07.027
105. Nguyen LH Nguyen HD Tran PT Biodegradation of natural rubber and deproteinized natural rubber by enrichment bacterial consortia Biodegradation 2020 31 303 317 10.1007/S10532-020-09911-0/FIGURES/8 32914250
106. Basik AA Nanthini J Yeo TC Sudesh K Rubber degrading strains: microtetraspora and dactylosporangium Polym 2021 13 3524 10.3390/POLYM13203524
107. Basik AA Trakunjae C Yeo TC Sudesh K Streptomyces sp AC04842: genomic insights and functional expression of its latex clearing protein genes (lcp1 and lcp2) when cultivated with natural and vulcanized rubber as the sole carbon source Front Microbiol 2022 10.3389/FMICB.2022.854427
108. Olejnik TP Pietras M Sielski J The process of natural and styrene–butadiene rubbers biodegradation by lactobacillus plantarum Appl Sci 2022 12 5148 10.3390/APP12105148
109. Ahrends A Hollingsworth PM Ziegler AD Current trends of rubber plantation expansion may threaten biodiversity and livelihoods Glob Environ Chang 2015 34 48 58 10.1016/J.GLOENVCHA.2015.06.002
110. Soratana K Rasutis D Azarabadi H Guayule as an alternative source of natural rubber: a comparative life cycle assessment with Hevea and synthetic rubber J Clean Prod 2017 159 271 280 10.1016/J.JCLEPRO.2017.05.070
111. Dong L Fujita T Promotion of low-carbon city through industrial and urban system innovation: Japanese experience and China’s practice World Sci Ref Asia World Econ 2015 10.1142/9789814578622_0033
112. ADB (2020) Managing Infectious Medical Waste during the COVID-19 Pandemic | Asian Development Bank. https://www.adb.org/publications/managing-medical-waste-covid19. Accessed 7Aug2022
113. Sangkham S Face mask and medical waste disposal during the novel COVID-19 pandemic in Asia Case Stud Chem Environ Eng 2020 2 100052 10.1016/J.CSCEE.2020.100052
114. Czajczyńska D Anguilano L Ghazal H Potential of pyrolysis processes in the waste management sector Therm Sci Eng Prog 2017 3 171 197 10.1016/J.TSEP.2017.06.003
115. Zabaniotou A Antoniou N Bruton G Analysis of good practices, barriers and drivers for ELTs pyrolysis industrial application Waste Manag 2014 34 2335 2346 10.1016/J.WASMAN.2014.08.002 25148927
| 36466440 | PMC9703434 | NO-CC CODE | 2022-11-29 23:21:09 | no | J Mater Cycles Waste Manag. 2022 Nov 28;:1-15 | utf-8 | J Mater Cycles Waste Manag | 2,022 | 10.1007/s10163-022-01554-y | oa_other |
==== Front
Chirurgie (Heidelb)
Chirurgie (Heidelb)
Chirurgie (Heidelberg, Germany)
2731-6971
2731-698X
Springer Medizin Heidelberg
36441200
1772
10.1007/s00104-022-01772-y
Leitthema
Chirurgie des alten Menschen – Thoraxchirurgie
Surgery of old people—Thoracic surgeryEhrsam Jonas Peter
Aigner Clemens [email protected]
grid.477805.9 0000 0004 7470 9004 Abteilung Thoraxchirurgie und thorakale Endoskopie, Ruhrlandklinik, Tüschener Weg 40, 45239 Essen, Deutschland
Redaktion U. Settmacher, Jena
28 11 2022
110
27 10 2022
© The Author(s), under exclusive licence to Springer Medizin Verlag GmbH, ein Teil von Springer Nature 2022
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Hintergrund
Eine Vielzahl thoraxchirurgisch relevanter Krankheitsbilder häuft sich mit steigendem Alter. Hohes Alter wird aber häufig als Kontraindikation für kurative Eingriffe oder größere operative Eingriffe per se gesehen.
Ziel der Arbeit
Darlegung der aktuellen relevanten Studienlage, Ableitung von Empfehlungen zur PatientInnenselektion sowie prä-, peri- und postoperativen Optimierung.
Material und Methoden
Auswertung der aktuellen Studienlage.
Ergebnisse
Wie aktuelle Daten zeigen, ist das Alter für die meisten thorakalen Erkrankungen allein kein Grund eine chirurgische Behandlung vorzuenthalten. Bei der Selektion gilt es viel mehr, Komorbiditäten, Gebrechlichkeit, Mangelernährung und kognitive Einschränkungen zu berücksichtigen. Eine Lobektomie oder Segmentresektion bei nichtkleinzelligem Lungenkrebs (NSCLC) im Stadium I kann bei gut selektionierten > 80-jährigen PatientInnen zu akzeptablen bis sogar vergleichbar guten Kurz- und Langzeitergebnissen führen wie bei jüngeren. Ausgewählte > 75-jährige PatientInnen mit NSCLC im Stadium II–IIIA können sogar von einer adjuvanten Chemotherapie profitieren. Nach guter Selektion sind Hochrisikoeingriffe wie eine Pneumonektomie bei > 70-Jährigen und pulmonale Endarteriektomie bei > 80-Jährigen ohne erhöhte Mortalitätsraten möglich. Und auch die Lungentransplantation kann bei gut selektionierten > 70-jährigen PatienInnen zu einem guten Langzeitergebnis führen. Minimal-invasive Operationsverfahren und nichtintubierte Anästhesie tragen zur Risikoreduktion bei marginalen PatientInnen bei.
Diskussion
In der Thoraxchirurgie ist nicht das numerische Alter, sondern das biologische Alter entscheidend. Zur Optimierung der Selektion, der Eingriffe, der Vor- und Nachbehandlung sowie der Lebensqualität sind angesichts der immer älter werdenden Bevölkerung dringend weitere Studien nötig.
Background
The incidence of a large number of diseases relevant to thoracic surgery increases with age; however, old age is still frequently considered a contraindication per se for curative interventions and extensive surgical procedures.
Objective
Overview of the current relevant literature, derivation of recommendations for patient selection as well as preoperative, perioperative and postoperative optimization.
Material and methods
Analysis of the current study situation.
Results
Recent data show that for most thoracic diseases, age alone is not a reason to withhold surgical treatment. Much more important for the selection are comorbidities, frailty, malnutrition and cognitive impairment. A lobectomy or segmentectomy for stage I non-small cell lung cancer (NSCLC) in carefully selected octogenarians can provide acceptable to even comparably good short-term and long-term results as in younger patients. Selected > 75-year-old patients with stages II-IIIA NSCLC even benefit from adjuvant chemotherapy. With appropriate selection high-risk interventions, such as pneumonectomy in > 70-year-old patients and pulmonary endarterectomy in > 80-year-old patients can be performed without an increase in mortality rates. Even lung transplantation can lead to good long-term results in carefully selected > 70-year-old patients. Minimally invasive surgical techniques and nonintubated anesthesia contribute to risk reduction in marginal patients.
Discussion
In thoracic surgery the biological age rather than the chronological age is decisive. In view of the increasingly older population, further studies are urgently needed to optimize patient selection, type of intervention, preoperative planning and postoperative treatment as well as the quality of life.
Schlüsselwörter
Alter
Thorakale Erkrankungen
Lungenchirurgie
Nichtkleinzelliges Lungenkarzinom
Lungentransplantation
Keywords
Age
Thoracic disease
Lung surgery
Non-small cell lung cancer
Lung transplantation
==== Body
pmcThoraxchirurgie ist für den alten Menschen von großer Relevanz. Bedeutsame Krankheitsbilder im Alter reichen von Onkologie über Empyeme, Rippenfakturen bis hin zu funktionellen Lungenerkrankungen. Entgegen der historischen Meinung ist das Alter allein für die meisten thorakalen Erkrankungen kein Grund, eine chirurgische Behandlung vorzuenthalten oder sie nur als vereinfacht palliative Form anzubieten. Eine Vielzahl neuerer Arbeiten weist darauf hin, dass nicht das numerische Alter entscheidend ist, sondern das biologische Alter, welches die Komorbiditäten und den Grad der Gebrechlichkeit berücksichtigt.
Hintergrund
Thoraxchirurgie ist für die aufgrund steigender Lebenserwartung immer älter werdende Bevölkerung von großer Relevanz. Die Wahrnehmung des „alten Menschen“ hat sich durch die erwartbare gesunde Lebensspanne in den letzten Jahrzehnten stark gewandelt. Die Definition des „alten Menschen“ ist daher in der Literatur sehr inhomogen und beginnt oft bei einem chronologischen Alter über 65 Jahren, jedoch wird auch teilweise 70 oder sogar 80 Jahre als Grenzwert herangezogen. Thoraxchirurgisch relevante Erkrankungsbilder des alten Menschen finden sich vor allem in der Onkologie, in funktionellen Lungenerkrankungen im Endstadium und in der Akutbehandlung von durch Osteoporose und Gebrechlichkeit aggravierten Thoraxtraumen sowie einer erhöhten Anfälligkeit für parapneumonische Komplikationen.
Dieses Review hat das Ziel, die lungenspezifischen und thoraxchirurgisch relevanten pathophysiologischen Verändern des alternden Menschen zu erläutern, die aktuelle Studienlage über die Morbidität und Mortalität nach Thoraxchirurgie aufzuzeigen sowie Empfehlungen für die optimale präoperative Selektion, die Optimierung der Operation, des perioperativen Managements und der postoperativen Rekonvaleszenz abzugeben.
Pathophysiologische Veränderungen von Thorax und Lunge im Alter
Das Auftreten altersbedingter physiologischer Veränderungen hat sich in den letzten 20 Jahren um 5 bis 10 Jahre nach hinten verlagert [1]. Der physiologische Alterungsprozess des Menschen stellt spezifische Herausforderungen und Risiken an die Thoraxchirurgie. Zum einen reduziert sich die Brustwand-Compliance durch zunehmende Kalzifikation der Knorpelgelenke sowie durch Verengung der Interkostalräume wegen Degeneration der Intervertebralgelenke. Zusammen mit der im Alter atrophierenden respiratorischen Muskulatur verringert sich die Fähigkeit zur intrathorakalen Volumenänderung. Während die totale Lungenkapazität über das Leben hinweg etwa gleich bleibt, steigt das Residualvolumen zwischen 20 und 65 Jahren von 20 auf 40 % an. Begleitend sinkt die Vitalkapazität mit dem Alter [2]. Auch die Lunge selbst verliert an Elastizität wodurch sich der negative intrapleurale Druck verringert, teilweise kleine Atemwege nicht mehr richtig offen gehalten werden können und sich das Totraumvolumen vergrößert. Das forcierte exspiratorische Volumen in einer Sekunde sinkt um mindestens 10 % zwischen 30 und 70 Jahren und die maximale Atemkapazität um 50 % [2].
Die Reizschwelle für den Atemreiz sowie für den Husten- und Schluckreflex sinkt
Degeneration von Alveolen über die Jahrzehnte führt zu einer geringeren Gasaustauschfläche. Auch die alveolokapilläre Membran verdichtet sich, wodurch die Sauerstoffpermeabilität abnimmt, was schlussendlich zu einem Oxygenationsdefizit mit progredientem Rechts-links-Shunt führt [2]. Zudem sinkt die Reizschwelle im zentralen Nervensystem für die Atmungsinitiative auf Hypoxie und Hyperkapnie [3] sowie für den Husten- und Schluckreflex, was das Risiko für Atemwegsinfekte und Aspiration steigert [4]. Die Muskulatur von Hypopharynx und Genioglossus atrophiert auch zunehmend, was zur Obstruktion der oberen Atemwege und Schlafapnoesyndrom führen kann [5]. All diese Faktoren tragen dazu bei, dass sich der Patient schneller erschöpft oder erschwert vom Respirator entwöhnt.
Ab dem 40. Lebensjahr sinkt auch der kardinale Auswurf jährlich um ca. 1 %, wodurch er bei einem 80-Jährigen nur noch rund 50 % beträgt [2]. Gleichzeitig sinkt die maximale Sauerstoffverwertbarkeit (VO2 max) des Körpers um 10 % pro Dekade [2]. Mit dem Alter kommt es auch zu zunehmender Glomerulosklerose mit Reduktion der Kreatininclearance. Ein schlechter kardinaler Auswurf hat zusätzliche negative renale Auswirkungen [2].
Nebst diesen physiologischen Alterungsprozessen, welche von Individuum zu Individuum stark in Ausprägung und Chronologie variieren, können sich durch Genetik, ungesunden Lebensstil und Umwelt über die Jahre zahlreiche Komorbiditäten entwickeln. Auch führt der Rückgang der funktionellen Reserven und Organfunktionen zu zunehmender körperlicher Schwäche, Gebrechlichkeit, Mangelernährung bis Sarkopenie sowie einem Immundefizit. Zudem häufen sich eine beeinträchtigte Kognition und Depression. Hierdurch entsteht ein sehr heterogener Pool an potenziellen PatientInnen, die es in der Thoraxchirurgie zu berücksichtigen gibt.
Domänen und aktuelle Ergebnisse der Thoraxchirurgie im Alter
Lungenkrebschirurgie
Lungenkrebschirurgie ist im Alter die zahlenmäßig mit Abstand größte thoraxchirurgische Domäne. Lungenkrebs wird in Industriestaaten in einem medianen Alter von 71 Jahren diagnostiziert und stellt ab einem Alter von 80 Jahren unter allen Malignomen die Haupttodesursache dar [6].
Die ersten kleinen Einzelzentrums- und Multizentrumsstudien aus den 1950er-, 1960er- und 1970er-Jahren, welche ihre Daten zu Lungenkrebschirurgie bei > 70-Jährigen präsentierten, zeigten abschreckend hohe perioperative Mortalitätsraten von 12,5 % [7] bis 26,6 % [8]. 5‑Jahres-Überlebensraten reichten von weniger als 6 % [9] bis maximal 36 % für gewisse Lobektomiegruppen [10], wobei die damals prädominant durchgeführten Pneumonektomien teilweise 0 % Überlebende aufwiesen [10]. Fortgeschrittenes Alter wurde daher in weiten Kreisen als Hochrisiko eingestuft und entsprechende Eingriffe gemieden. Durch neue funktionelle Abklärungsmethoden, detailliertere Bildgebungen zum Tumorstaging, bessere perioperative Monitorisierungsmöglichkeiten, sanftere Beatmungsgeräte, bessere kardiovaskuläre Medikamente sowie schonendere Operationstechniken konnten über die weiteren Jahrzehnte die Morbidität, Mortalität und Krebsrezidive deutlich gesenkt werden.
Für NSCLC im Stadium I ist altersunabhängig eine kurativ intendierte Resektion zu prüfen
Aktuelle Studien (Tab. 1) zeigen für gut selektionierte > 80-jährige PatientInnen mit nichtkleinzelligem Lungenkarzinom im Frühstadium I (NSCLC Stadium I) In-Hospital-Mortalitäten von nur noch 0–3,9 %. Trotz Selektion liegt die perioperative Komplikationsrate alter PatientInnen aber noch höher als für Jüngere. Auch unter Anwendung minimal-invasiver Verfahren wird bei > 80-Jährigen von Komplikationsraten von 23–54 % berichtet (Tab. 1). Gehäuft bei Älteren sind Vorhofflimmern, Myokardinfarkt, prolongierte Luftleckage, akutes respiratorisches Distresssyndrom und postoperatives Delir (Tab. 1). Diese Komplikationen sind wiederum starke Risikofaktoren für einen langen Hospitalisationsaufenthalt, Morbidität und Mortalität. Das aktuelle 5‑Jahres-Überleben > 80-Jähriger liegt bei 36–49 % (Tab. 2). Ohne Therapie sinkt das mittlere Überleben aber auf noch 1,2 Jahre [11]. In Anbetracht dessen, dass die Lebenserwartung in Industrieländern für 80-Jährige heute bei 89 Jahren liegt, also im Mittel noch 9 Jahre beträgt [12], sind diese chirurgischen Ergebnisse auch im erhöhten Alter gut zu rechtfertigen. Für NSCLC im Stadium I empfehlen aktuelle Guidelines daher, altersunabhängig eine kurativ intendierte Resektion zu prüfen [13]. Zaatar et al. 2020 [18] Bongiolatti et al. 2021 [19] Berry et al. 2011 [20] Fiorelli et al. 2016 [21] Sarkari et al. 2019 [22] Mimae et al. 2021 [23]
2016–2018 2014–2019 2000–2009 2006–2012 2011–2015 2015–2016
Universitätsmedizin Essen, Deutschland, retrospektiv Datenbank, Italien, retrospektiv Duke Universität, USA, retrospektiv Multizentrumstudie, Italien, retrospektiv National Premiere Database, USA, retrospektiv JACS1303, Multizentrumstudie, Japan, randomisiert kontrolliert
S + L, offen und VATS L‑VATS K + L, offen und VATS K + L, offen und VATS Roboter‑L VATS‑L L offen K S + L
Alter (Jahre) < 70 ≥ 70 ≤ 80 > 80 ≥ 80 > 75 ≥ 80 ≥ 80 ≥ 80 ≥ 80 ≥ 80
Mittel (Jahre) 60,4 75,2 67,1 82,5 82 78 82,3 82,6 82,4 82,5 82
Anzahl 315 190 6694 329 109 239 232 562 562 66 90
Leichte Komplikationen
Luftleckage > 5 Tage (%) 9,5 10,5 0,1 – 15 4,2 9,2 9,3 8,2 – –
Atelektase (%) 1,3 0,5 2,9 4 – 2,9 14,2 14,9 17,1 – –
Chylothorax (%) 1,3 1 0.6 0.3 – – – – – – –
Nervus-recurrens–Parese (%) 0,3 1,6 2,1 5,5 1 – – – – – –
Weichteilemphysem (%) 0,3 0 – – – – – – – – –
Wundinfkektion (%) 0,3 0 – – – – – – – – –
Sekretverhalt (%) – – 1,3 2,7 6 – – – – – –
Schwere Komplikationen – – – – – 10 – – – – –
Vorhofflimmern (%) 1,6 5,3 0,2 – 17 3,8 15,1 18,3 21,5 – –
Myokardinfarkt (%) 0 0 8,3 10,3 1 0,8 1,3 1,4 2,0 – –
Bronchopleurale Fistel (%) 0,3 0,5 0,3 – – – – – – – –
Tiefe Venenthrombose (%) 0,3 0 – – – – 0,9 0,4 1,2 – –
Multiorganversagen (%) 0,3 0,5 – – – – – – – – –
Pneumonie (%) 0,6 1,6 0,4 0,9 – 1,3 8,2 9,6 11,2 – –
Empyem (%) 0,3 1 – – – – – – – – –
Lungenembolie (%) 0,,3 0 0,,4 1 – 0,,4 0 1,,1 0,,9 – –
Reintubation (%) – – 2,2 1,2 4 – 7,3 10,9 12,4 – –
ARDS (%) – – 3,4 5,8 – 6,3 – – – – –
Nierenversagen (%) 0,3 0,5 0,5 1,2 2 – – – – – –
Reoperation bei Blutung (%) 0,9 1 0,3 0,6 – 1,3 7,3 12,8 14,9 – –
Delir (%) – – – – 11 0,4 – – – – –
Total Komplikationsrate (%) 22 33 25,5 31,9 – – 43,1 49,8 53,9 10,6 23,4
Hospitalisationstage – – 6,8 8 4,9 10,4 7,4 8,1 9,7 7 9
In-Hospital-Mortalität (%) 0,6 2,6 – – 3,7 – 3,4 3,9 3,4 0 1,1
ARDS „acute respiratory distress syndrome“, K Keilresektion, L Lobektomie, S Segmentresektion, VATS videoassistiere Thorakoskopie
Sun et al. 2017 [24] Zaatar et al. 2020 [18] Fiorelli et al. 2016 [21] Chan et al. 2022 [25] Mimae et al. 2021 [23]
2004–2013 2016–2018 2006–2012 2004–2016 2015–2016
Surveillance, Epidemiology, and End Results Database, USA, retrospektiv Universitätsmedizin Essen, Deutschland, retrospektiv Multizentrumstudie, Japan, retrospektiv National Cancer Database, USA, retrospektiv JACS1303, Multizentrumstudie, Japan, randomisiert kontrolliert
Resektion L + K + S S + L L K L K + S L + S K
Alter (Jahre) 65–69 70–74 75–79 80–84 ≥ 85 < 70 ≥ 70 > 75 > 75 ≥ 80 ≥ 80 ≥ 80 ≥ 80
Mittel (Jahre) – – – – – 60,4 75,2 77,9 79,2 82,0 82,0 82,0 82,5
Anzahl 6147 6061 5115 2725 734 315 190 149 90 14.594 6370 90 66
30-Tage-Mortalität (%) – – – – – 0,6 2,6 – – – – 1,9 0
3‑Jahres-Überleben (%) – – – – – – – – – – – 83,7 86,5
5‑Jahres-Überleben (%) 64,9 58,5 51,1 43,1 36,3 – – 60,5 45 48,5 41,1 – –
S Segmentresektion, K Keilresektion, L Lobektomie
Für NSCLC im höheren Stadium II–IIIA ist in den Guidelines prinzipiell auch eine kurative Resektion empfohlen, nach welcher aber nur bei < 75-Jährigen eine adjuvante Chemotherapie angeschlossen werden soll [13]. Älteren wird von einer adjuvanten Chemotherapie aufgrund befürchteter Intoleranz und Toxizität abgeraten [13]. Erste Studien an gut selektionierten ≥ 75-jährigen PatientInnen, die eine adjuvante Chemotherapie erhielten [14–16], darunter eine aktuelle Propensity-Score-Matched-Analyse [16], welche nur PatientInnen mit fehlender Niereninsuffizienz und höchstens mäßiger Leistungseinschränkung einschloss, zeigen aber auch für diese Altersgruppe einen deutlichen Vorteil bezüglich Rezidivfreiheit und Gesamtüberleben.
Obwohl die kurative Chirurgie zumindest für NSCLC im Stadium I seit langem der Goldstandard ist, werden PatientInnen mit steigendem Alter bezüglich einer solchen immer noch substanziell benachteiligt. Ein aktueller Vergleich von drei medizinisch hochentwickelten europäischen Ländern [17] hat gezeigt, dass die Behandlungsansätze des NSCLC im Stadium I in der alten Population stark variieren (Abb. 1). So hatte die Altersgruppe 70 bis 79 in 41–56 % und die Altersgruppe ≥80 in 13–23 % eine kurative Resektion erhalten. Für PatientInnen, die nicht für Chirurgie qualifiziert eingestuft wurden oder keine Chirurgie wünschten, wurde in 14–47 % der 70- bis 79-Gruppe und in 16–60 % der ≥ 80-Gruppe eine stereotaktische Bestrahlung durchgeführt. Keine oder zumindest keine kurative Behandlung erhielten 9–24 % der 70-bis 79-Jährigen und 23–54 % der ≥80-Jährigen.
Metastasenchirurgie
Die Lunge ist einer der häufigsten Metastasierungsorte solider Tumoren; im Alter ist dabei das Kolonkarzinom prädominant. Metastasenchirurgie der Lunge ist ein etabliertes, potenziell kuratives Verfahren, das bei selektionierten PatientInnen eine Lebensverlängerung bewirken kann, wenn der Primärtumor unter Kontrolle ist, keine weiteren extrathorakalen Metastasen vorliegen, die Lungenfunktion genügend ist und die Metastasen resektable sind. Höheres Alter ist derzeit eine signifikante Hemmschwelle für eine Überweisung an die Thoraxchirurgie [26, 27]. Eine aktuelle retrospektive Einzelzentrumsstudie, welche pulmonale Metastasektomien ≥ 70-jähriger PatientInnen (n = 222) mit < 70-jährigen (n = 538) verglich und hierbei in der älteren Gruppe auf eine überschaubare Anzahl Metastasen und möglichst thorakoskopische Resektion bedacht war, zeigte aber keine Unterschiede in Komplikationsraten und Hospitalisationsdauer und ein nichtsignifikant unterschiedliches 5‑Jahres-Überleben [27]. Zu gleichen ermutigenden Ergebnissen ist kürzlich auch eine kleinere retrospektive Einzelzentrumsstudie gekommen [26].
Pulmonale Endarteriektomie
Das Risiko für Lungenembolien steigt mit höherem Alter und hierdurch auch die Zahl an PatientInnen mit chronisch thromboembolischer pulmonaler Hypertonie. Der Goldstandard zur Heilung ist die pulmonale Endarteriektomie. Dieser Eingriff gilt per se als Hochrisikoeingriff und bedarf am kardiopulmonalen Bypass eines Kreislaufstillstands in tiefer Hypothermie. Mehrere aktuelle Einzelzentrumsstudien an hoch selektionierten > 70-jährigen [28] bzw. ≥ 80-jährigen PatientInnen [29, 30] zeigen, dass diese Gruppen im Vergleich zu jüngeren PatientInnen postoperativ länger intubiert und invasiv beatmet bleiben und einen längeren Hospitalisationsaufenthalt haben. Die Mortalitätsrate während des stationären Aufenthaltes schwankte zwischen 2,1 % für > 70-Jährige bis 8–17 % für ≥ 80-Jährige (Tab. 3). Das Langzeitüberleben war aber nicht signifikant unterschiedlich, zumindest nach Vergleich mit einer geschlechts- und altersidentischen populationsbasierten Referenzkohorte [29]. Auch wird von einem funktionellen Profit für alle Gruppen berichtet, wobei aber in einer der Studien [30] bei ≥ 80-Jährigen eine geringere Verbesserung des Herzindex detektiert wurde. Wiedenroth et al. 2022 [28] Newnham et al. 2017 [29] Grazioli et al. 2021 [30]
2014–2016 2006–2016 1994–2016
Justus-Liebig Universität, Deutschland, prospektiv Royal Papworth Hospital, England, retrospektiv Universität Pavia, Italien, retrospektiv
Alter (Jahre) ≤ 50 51–70 > 70 < 80 ≥ 80 < 60 60–79 ≥ 80
Mittleres Alter (Jahre) 38,9 60,5 75,7 61 81 – – –
Anzahl 89 154 143 1115 32 259 352 24
Komplikationsrate (%) – – – – – 35 54 71
Spitalsmortalität (%) 0 2,6 2,1 4 8 4 10 17
1‑Jahres-Überleben (%) 98,9 96,8 93,7 91,8 83,5 – – –
5‑Jahres-Überleben (%) – – – 84,4 69,4 95 85 70
Chirurgische Lungenvolumenreduktion
Das durch Noxen, Genetik und mutmaßlich weitere Einflüsse entstehende Lungenemphysem und die miteinhergehede chronische obstuktive Lungenerkrankung gehören zu den Hauptmorbiditäts- und -mortalitätsursachen im Alter. Die meist thorakoskopisch durchgeführte Lungenvolumenreduktion zielt auf die Reduktion der Hyperinflation und hierdurch die Verbesserung der Atemmechanik. Im randomisiert-kontrollierten National Empyhsema Treatment Trial von 1998 bis 2002 wurde für ≥ 70-Jährige eine Prärehabilitation vorausgesetzt [31]. Zudem fand diese Studie, dass steigendes Alter ein unabhängiger Faktor für postoperative schwerwiegende pulmonale und kardiale Komplikationen innerhalb der ersten 30 Tage ist [32]. Auch eine neuere US-Datenbank-Analyse von 2001 bis 2017, welche 59,9- ± 11,7-jährige PatientInnen beinhaltet, kam für steigendes Alter zum selben Schluss [33]. Eine ältere Einzelzentrumsstudie an 9 über 75-jährigen PatientInnen berichtete über gemischte Kurz- und Langzeitresultate [34]. Neuere Studien, welche die effektiven Risiken innerhalb der alten Population beleuchten, fehlen. Datenbanken aus England, wo Alter derzeit nicht als Ausschusskriterium gehandhabt wird, zeigen, dass ältere PatientInnen trotz erfüllter Kriterien derzeit deutlich seltener operiert werden [35].
Septische Thoraxchirurgie
Die Prävalenz parapneumonischer Empyeme, superinfizierter kardial, hepatisch oder renal bedingter Pleuraergüsse sowie superinfizierter Hämatothoraces nehmen mit höherem Alter zu. Sepsisbedingt resultiert hierdurch eine erhebliche Morbidität und Mortalität. Entgegen historischen Studien war Alter in einer Gruppe 80- bis 95-Jähriger (n = 49) im Vergleich zu Jüngeren (n = 286) kein Risikofaktor in einer neueren Einzelzentrumsstudie mit chirurgischer parapneumonischer Empyemsanierung [36]. Obwohl die alte Gruppe ein höheres Komorbiditätsprofil und gleichverteilte Empyemstadien aufwies, traten unter bevorzugt thorakoskopischem Verfahren keine höheren Raten an Sepsis, sepsisassoziierten Komplikationen und respiratorischen Dekompensationen auf.
Rippenosteosynthese
Rippenfakturen häufen sich durch steigende Prävalenz von Osteoporose und Sturzneigung im Alter. Obwohl Rippenfrakturen im Alter häufig durch leichtgradige Traumata verursacht werden, haben ≥ 65-jährige PatientInnen gemäß einem nationalen Register aus den Niederlanden [37] eine hohe Mortalitätsrate von 11,7 %, während es bei Jüngeren nur 3,0 % sind. Rippenosteosynthese kann in gewissen Fällen von Rippenserienfakturen und insbesondere instabilem Thorax zur Reduktion oder Vermeidung einer Beatmungsbedürftigkeit sowie der Kurz- und Langzeitmorbidität und -mortalität beitragen. In einer retrospektiven Einzelzentrumsstudie wurden bei ≥ 70-jährigen PatientInnen (n = 295) nach Plattenosteosynthese keine gehäuften Komplikationen gefunden [38].
Lungentransplantation
Die Nachfrage nach Lungentransplantation als lebensverlängernde Maßnahme für Lungenerkrankungen im Endstadium ist insbesondere bei älteren PatientInnen seit Jahren ein anhaltender Trend. Den aktuellen Guidelines zufolge gilt ein Alter > 65 Jahren mit niedrigen physiologischen Reserven als relative Kontraindikation zur Transplantation [39]. Dennoch macht beispielsweise in den USA der Anteil an > 65-Jährigen inzwischen schon mehr als 38 % der Transplantierten aus [40].
Eine neuere Analyse der großen United Network for Organ Sharing Datenbank konnte zwar für ≥ 70-Jährige ein vergleichbares 1‑Jahres-Überleben wie für 60- bis 69-Jährige nachweisen, zeigte aber für die ältere Gruppe ein deutlich schlechteres 5‑Jahres-Überleben (Tab. 4; [41]). Im Gegensatz dazu berichtete eine aktuelle große Einzelzentrumsstudie, dass prinzipiell ein vergleichbar gutes 5‑Jahres-Überleben für ≥ 70-jährige wie für < 65-jährige PatientInnen möglich ist [42]. Wie eine Einzelzentrumsstudie demonstrierte, scheinen die zugrunde liegende Lungenerkrankung, das Komorbiditäsprofil und die Spenderlungenqualität bei der Lungentransplantation für das Kurz- und Langzeitüberleben deutlich wichtigere Faktoren zu sein als das steigende Alter an sich [43]. Olson et al. 2021 [42] Hayanga et al. 2015 [41]
2012–2016 2006–2012
St. Joseph’s Hospital and Medical Center, Phoenix, USA United Network for Organ Sharing, USA
Alter (Jahre) 18–64 65–69 ≥ 70 60–69 ≥ 70
Mittel (Jahre) 53,9 66,9 71,1 – –
Anzahl 221 109 41 4327 543
30-Tage-Überleben (%) – – – 96,2 96,8
1‑Jahres-Überleben (%) 93,2 83,5 86,7 81,7 78,6
3‑Jahres-Überleben (%) 70,1 59,6 64,4 63,7 49,3
5‑Jahres-Überleben (%) 58,8 44,0 57,8 47,5 28,2
Ob die Allokation von Spenderlungen an ältere PatientInnen angesichts chronischen Organmangels gerechtfertigt ist, ist eine anhaltende große ethische Debatte. Zudem werden ältere PatientInnen mit eigentlichen Ausschlusskriterien wie Herzerkrankungen, durch perkutane Koronarintervention und Transkatheter-Aortenklappenersatz immer häufiger zu Lungentransplantationskandidaten. Ob sich solche kurzzeitige Verbesserungen unter diesem Hochrisikoeingriff und der anschließenden Immunsuppression nicht vorschnell wieder aggravieren, bedarf weiterer Forschung [44].
Des Weiteren besteht noch viel Forschungsbedarf bezüglich Gebrechlichkeit. Im Vergleich zu beispielsweise älteren onkologischen PatientInnen erholen sich viele gebrechliche PatientInnen mit einer chronischen Lungenerkrankung innerhalb der ersten 6 Monate nach Transplantation deutlich [45, 46]. Auf der anderen Seite geht Gebrechlichkeit mit einer erhöhten Wartelistenmortalität einher [45].
Postoperative Lebensqualität
Obwohl von zentraler Bedeutung für die PatientInnen, wurde die Lebensqualität nach Thoraxchirurgie im Alter bisher kaum untersucht. Eine kürzlich veröffentlichte kleinere, etwas inhomogen aufgebaute Studie [47], die Altersgruppen von < 70 (n = 33), 70–79 (n = 25) und ≥ 80 (n = 48) mehrere Monate postoperativ befragte, fand in der ältesten Altersgruppe sowohl global als auch funktionell, physikalisch und emotional die höchste Lebensqualität. Das Operationsausmaß hatte hierbei keinen Einfluss. Als mögliche Erklärung für dieses Ergebnis wurde eine bessere Akzeptanz von Funktionsverlust und postoperativen Symptomen mit höherem Alter herangezogen.
Identifikation alter RisikopatientInnen
Die Selektion alter PatientInnen, die von thoraxchirurgischen Interventionen profitieren, die Identifikation von RisikopatientInnen, bei denen die chirurgische Therapie adaptiert werden muss, und der Ausschluss von HochrisikopatientInnen sind Herausforderungen, die einer ausführlichen, teils interdisziplinären präoperativen Abklärung bedürfen.
Ein steigendes Komorbiditätsprofil gemessen am altersunabhängigen Charlson-Deyo-Index wurde in der Lungenkrebschirurgie als unabhängiger Prädiktor für progrediente Morbidität und Mortalität bei > 80-Jährigen [25] und der reguläre Charlson-Index als Risikofaktor für Langzeitüberleben identifiziert [48]. Unter den Komorbiditäten waren in der Lungenkrebschirurgie vorbestehende Herzerkrankungen [49] und in der Metastasenchirurgie Lungenemphysem [27] und präoperative Arhythmie [26] von hoher Bedeutung.
Weitere Prädiktoren aus der Lungenkrebschirurgie sind männliches Geschlecht [48] und ein bereits fortgeschrittenes Tumorstadium [48, 49].
Bereits eine geringe mentale Einschränkung (Montreal Cognitive Assessments) wurde mit dem Auftreten eines postoperativen Delirs und Folgekomplikationen assoziiert [50].
Mangelernährung ist mit Immundefizit und Wundheilungsstörung verbunden
Mangelernährung, gemessen am Prognostic Nutritive Index, war bei ≥ 75-jährigen Patienten nach Lungenkrebschirurgie mit Immundefizit und Wundheilungsstörung wie prologierter Lungenleckage, bronchopulmonaler Fistel, Pneumonie und Empyem verbunden. Diese Komplikationen hatten wiederum Einfluss auf die Mortalität [51].
Das Verhältnis aus C‑reaktivem Protein und Albumin (Glasgow Prognostic Score), welches das Verhältnis aus systemischer Inflammation und Malnutrition widerspiegelt, wurde als prognostisch für das krebsfreie und Langzeitüberleben nach Lungenkrebschirurgie beschrieben [48, 52].
Das Torque-Teno-Virus, ein ubiquitärer nichtpathogener Virus, wird in immunkompromittierten Individuen repliziert. Erste Studien [53] weisen darauf hin, dass dieses Virus bei alten PatientInnen in Korrelation zu einer veränderten Immunlage eine veränderte Dynmaik zeigt und parallel keine Assoziation zu Tumorerkrankung und Stadium besteht.
In der Transplantationschirurgie wird derzeit versucht, die besonders gebrechlichen Kandidaten, das biologische Alter und die immunologische Kompetenz anhand von Bildgebungen, blutbasierten Biomarkern, Zellfunktion, Entzündung und Telomerlänge besser zu eruieren [46]. Mehrere geriatrische Fragebögen und Bewertungsskalen wurden inzwischen auch auf Prädiktion des postoperativen Verlaufs alter gebrechlicher PatientInnen getestet [54, 55]. Unter ihnen hat sich aber noch keiner/keine durchgesetzt.
Prä-, peri- und postoperative Optimierungsmöglichkeiten
Präoperative Optimierung
Präoperativ am bedeutendsten ist, ein auf die PatientInnen individuell zugeschnittenes Behandlungskonzept festzulegen und Risiken und mögliche Konsequenzen klar zu benennen. In gewissen Fällen ist es sogar wichtig, eine Diskussion über Erwartungen über das Lebensende zu führen, um Angehörigen eine gegebenenfalls später notwendig werdende Entscheidung abzunehmen.
Nebst der Behandlung von Infekten, Anämie und bei aktiven RaucherInnen der Initiation einer Nikotinkarenz soll prinzipiell auch Malnutrition optimiert werden [56]. Im Vergleich zu einer Kontrollgruppe hatten PatientInnen, welche hochkalorische Getränke über 5 Tage vor der Operation einnahmen, nach Lungenmalignomresektion signifikant niedrigere Komplikationsraten und Hospitalisationsaufenthalte [57]. Die wenigen existierenden Studien zur präoperativen Rehabilitation vor Lungenkrebschirurgie konnten bisher nicht alle einen postoperativen Benefit nachweisen [58–60]. Ein Hauptproblem bleibt für onkologische Eingriffe oft auch die hierfür nur limitiert vertretbare Zeit.
Perioperative Optimierung
Notfalleingriffe, ausgedehnte Resektionen [20, 49] und lange Operationszeit [48, 51] wurden wiederholt als Risikofaktoren für postoperative Morbidität und Mortalität alter PatientInnen identifiziert.
Resektionsausmaß
Obwohl seit der randomisierten Studie der Lung Cancer Study Group [61] 1995 für NSCLC im Stadium I eine Lobektomie nachweislich einer simplen Keilresektion bezüglich Rezidivrate als überlegen gilt und als Goldstandard etabliert wurde, wird der Nutzen im fortgeschrittenen Alter kontrovers diskutiert. Erste größere Datenbankanalysen [62] zeigten keinen Unterschied. Seither gingen zahlreiche retrospektive und prospektive Studien dieser Frage nach [63]. Neueste Studien [21, 23, 25] finden sich in Tab. 2. Fehlende oder heterogene Baseline-Charakteristika erschweren aber die abschließende Interpretation, ob Keilresektion, Segmentresektion oder Lobektomie für alte PatientInnen besser ist.
Die Pneumonektomie gilt generell als Hochrisikoeingriff und wird im Alter wegen historisch hoher Morbiditäts- und Mortalitätsraten gemieden. In einer neuen retrospektiven Multizentrumsstudie an 136 gut selektionierten > 70-jährigen Patienten, die wegen eines zentralen Lungenkarzinoms eine Pneumonektomie erhielten, kam es zwar in 45 % zu postoperativen Komplikationen, die In-Hospital-Mortalität betrug jedoch nur 1,5 % und die 30-Tage-Mortalität 3,7 %. Zudem war eine neoadjuvante Therapie mit keiner erhöhten Komplikations- oder Mortalitätsrate verbunden [64]. Die beschriebenen Ergebnisse zeigen, dass bei hoch selektionierten PatientInnen so niedrige Kurzzeitmortalitätsraten wie für Lobektomien erzielt werden können.
Operationszugang
Obwohl mittels Thorakotomie eine Resektion an der Lunge meist schneller erfolgen kann als mittels videoassistierter Thorakoskopie, zeigen aktuelle Studien [65] für Letztere auch bei älteren PatientInnen eine deutliche Überlegenheit. Durch die Thorakoskopie wird die Brustwandintegrität und somit die Atemmuskulatur weniger verletzt. Hierdurch wurde nach Thorakoskopie eine geringer Inflammation, ein kürzerer Hospitalisationsaufenthalt sowie durch die Videooptik eine radikalere Lymphadenektomie beobachtet [64]. Wird für die Resektion dennoch ein größerer Zugang benötigt, soll dieser möglichst muskelerhaltend erfolgen.
Anästhesie
Die Adaptation der Anästhesie an den Körper des alten Menschen ist für die Operation unerlässlich. Die mechanische Beatmung und besonders die Einlungenventilation müssen so sanft wie möglich gehaltenen werden, um gealterte oder vorgeschädigte Lungen keinem Barotrauma auszusetzen [66]. Wegen des geringeren Wasseranteils des Körpers müssen Medikamente vorsichtiger dosiert und wegen Muskelatrophie die Körpertemperatur penibler überwacht werden. Des Weiteren muss die Anästhesietiefe stetig überwacht werden, da durch eine zu tiefe Sedierung Hypotonie mit konsekutiver Minderperfusion und Schädigung der im Alter oft kompromittierten Organe droht [66].
Bei der nichtintubierten Thoraxchirurgie wird der Patient spontan atmend operiert
Ein neuerer Ansatz mit zunehmender Popularität ist die nichtintubierte Thoraxchirurgie. Hierbei wird der Patient spontan atmend, mit thorakaler Epiduralanästhesie oder einem Interkostalnervenblock wach oder unter leichter Sedition operiert. Der Hustenreiz wird dabei durch einen Vagusblock unterdrückt. Dieses Verfahren ermöglicht es, auch Patienten mit stark kompromittierter kardialer Pumpfunktion zu operieren, welche keine Vollnarkose tolerieren würden. Zudem ist die Spontanatmung schonender für das Lungengewebe und es kommt weniger zu einem Ventilations-Perfusions-Ungleichgewicht.
Mit diesem Ansatz wurden im Alter inzwischen nebst Pleuraergüssen, Pleurabiopsien und Empyeme [67] in einem Einzelzentrum 36 Lobektomien bei 65- bis 87-Jährigen mit NSCLC im Stadium I und II durchgeführt [68]. Dasselbe Zentrum hat in einer aktuellen Propensity-Matched-Studie ihre nichtintubierten Thorakoskopien (n = 79) mit ihren herkömmlichen Thorakoskopien (n = 158) bei gut selektionierten ≥ 75-jährigen PatientInnen verglichen [69]. Bei der nichtintubierten Thorakoskopie wurden zwar weniger Lymphknoten reseziert, jedoch deutlich weniger und kürzere Intensivstationsaufenthalte sowie tendenziell weniger Komplikationen und kürzere Hospitalisationsaufenthalte gesehen. Überleben und Rezidivfreiheit waren vergleichbar zwischen den Gruppen.
Postoperative Optimierung
Alte wie auch junge PatientInnen sollten so früh wie möglich extubiert, mobilisiert und Drainagen so bald wie möglich entfernt werden. Alte PatientInnen sind häufig unfähig, ihre Schmerzen adäquat zu äußern, und werden daher fälschlicherweise oft als weniger schmerzsensibel eingestuft. Zu geringe Analgesie hemmt die Atemmechanik und erhöht das Risiko für Atelektasen, Pneumonie und Hypoxie [70]. Anschlussrehabilitationen, Anschlusslösungen oder Hilfen für zu Hause sollten für diese Patientengruppe frühzeitig organisiert werden.
Fazit für die Praxis
Alter per se ist in der Thoraxchirurgie kein Ausschlusskriterium.
Selbst große Eingriffe wie Pneumonektomien, pulmonale Endarteriektomien und Lungentransplantationen können bei guter Selektion mit geringen Morbiditätsraten und guten Kurz- und Langzeitresultaten durchgeführt werden.
Die Selektion soll nach biologischem und nicht nach chronologischem Alter unter Berücksichtigung des Komorbiditätsprofils, der Gebrechlichkeit, der Mangelernährung und kognitiver Einschränkung erfolgen.
Durch minimal-invasive Operationsverfahren und nichtintubierte Anästhesie können auch marginalere PatientInnen von Operationen profitieren.
Zur Optimierung der Selektion, der Eingriffe, der Vor- und Nachbehandlung sowie der Lebensqualität sind angesichts der immer älter werdenden Bevölkerung dringend weitere Studien notwendig.
Einhaltung ethischer Richtlinien
Interessenkonflikt
J.P. Ehrsam und C. Aigner geben an, dass kein Interessenkonflikt besteht.
Für diesen Beitrag wurden von den Autor/-innen keine Studien an Menschen oder Tieren durchgeführt. Für die aufgeführten Studien gelten die jeweils dort angegebenen ethischen Richtlinien.
QR-Code scannen & Beitrag online lesen
==== Refs
Literatur
1. https://data.oecd.org/healthstat/life-expectancy-at-65.htm. Zugegriffen: 1. Sept. 2022
2. Tonner PH Kampen J Scholz J Pathophysiological changes in the elderly Best Pract Res Clin Anaesthesiol 2003 17 2 163 177 10.1016/S1521-6896(03)00010-7 12817912
3. García-Río F Villamor A Gómez-Mendieta A The progressive effects of ageing on chemosensitivity in healthy subjects Respir Med 2007 101 10 2192 2198 10.1016/j.rmed.2007.04.015 17643972
4. Meyer KC Lung infections and aging Ageing Res Rev 2004 3 1 55 67 10.1016/j.arr.2003.07.002 15163102
5. Hoch CC Reynolds CF 3rd Monk TH Comparison of sleep-disordered breathing among healthy elderly in the seventh, eighth, and ninth decades of life Sleep 1990 13 6 502 511 10.1093/sleep/13.6.502 2126391
6. https://seer.cancer.gov/statfacts/html/lungb.html. Zugegriffen: 1. Sept. 2022
7. Bates M Results of surgery for bronchial carcinoma in patients aged 70 and over Thorax 1970 25 1 77 78 10.1136/thx.25.1.77 5418012
8. Evans EW Resection for bronchial carcinoma in the elderly Thorax 1973 28 1 86 88 10.1136/thx.28.1.86 4685214
9. Harviel JD McNamara JJ Straehley CJ Surgical treatment of lung cancer in patients over the age of 70 years J Thorac Cardiovasc Surg 1978 75 6 802 805 10.1016/S0022-5223(19)39591-1 661348
10. Kirsh MM Rotman H Bove E Major pulmonary resection for bronchogenic carcinoma in the elderly Ann Thorac Surg 1976 22 4 369 373 10.1016/S0003-4975(10)64969-7 985655
11. McGarry RC Song G des Rosiers P Timmerman R Observation-only management of early stage, medically inoperable lung cancer: poor outcome Chest 2002 121 4 1155 1158 10.1378/chest.121.4.1155 11948046
12. Arias E Xu J United States life tables, 2019 Natl Vital Stat Rep 2022 70 19 1 59 35319436
13. Postmus PE Kerr KM Oudkerk M Early and locally advanced non-small-cell lung cancer (NSCLC): ESMO clinical practice guidelines for diagnosis, treatment and follow-up Ann Oncol 2017 28 4 iv1 iv21 10.1093/annonc/mdx222 28881918
14. Pepe C Hasan B Winton TL Adjuvant vinorelbine and cisplatin in elderly patients: national cancer institute of Canada and intergroup study JBR.10 J Clin Oncol 2007 25 12 1553 1561 10.1200/JCO.2006.09.5570 17442999
15. Wisnivesky JP Smith CB Packer S Survival and risk of adverse events in older patients receiving postoperative adjuvant chemotherapy for resected stages II-IIIA lung cancer: observational cohort study BMJ 2011 343 d4013 10.1136/bmj.d4013 21757436
16. Blasi M Eichhorn ME Christopoulos P Major clinical benefit from adjuvant chemotherapy for stage II-III non-small cell lung cancer patients aged 75 years or older: a propensity score-matched analysis BMC Pulm Med 2022 22 1 255 10.1186/s12890-022-02043-6 35761214
17. Damhuis RAM Senan S Khakwani A Harden S Helland Å Strand TE Age-related treatment patterns for stage I NSCLC in three European countries J Geriatr Oncol 2021 12 8 1214 1219 10.1016/j.jgo.2021.05.005 33994330
18. Zaatar M Stork T Valdivia D Minimal-invasive approach reduces cardiopulmonary complications in elderly after lung cancer surgery J Thorac Dis 2020 12 5 2372 2379 10.21037/jtd.2020.03.73 32642142
19. Bongiolatti S Gonfiotti A Borgianni S Crisci R Curcio C Voltolini L Post-operative outcomes and quality of life assessment after thoracoscopic lobectomy for non-small-cell lung cancer in octogenarians: analysis from a national database Surg Oncol 2021 37 101530 10.1016/j.suronc.2021.101530 33548589
20. Berry MF Onaitis MW Tong BC Harpole DH D’Amico TA A model for morbidity after lung resection in octogenarians Eur J Cardiothorac Surg 2011 39 6 989 994 10.1016/j.ejcts.2010.09.038 21276728
21. Fiorelli A Caronia FP Daddi N Sublobar resection versus lobectomy for stage I non-small cell lung cancer: an appropriate choice in elderly patients? Surg Today 2016 46 12 1370 1382 10.1007/s00595-016-1314-8 27085869
22. Sarkaria IS Gorrepati ML Mehendale S Oh DS Lobectomy in octogenarians: real world outcomes for robotic-assisted, video-assisted thoracoscopic, and open approaches J Thorac Dis 2019 11 6 2420 2430 10.21037/jtd.2019.05.52 31372279
23. Mimae T Saji H Nakamura H Survival of octogenarians with early-stage non-small cell lung cancer is comparable between wedge resection and lobectomy/segmentectomy: JACS1303 Ann Surg Oncol 2021 28 12 7219 7227 10.1245/s10434-021-09835-w 33900499
24. Sun F Ma K Yang X A nomogram to predict prognosis after surgery in early stage non-small cell lung cancer in elderly patients Int J Surg 2017 42 11 16 10.1016/j.ijsu.2017.04.024 28428062
25. Chan EY Amirkhosravi F Nguyen DT Chihara RK Graviss EA Kim MP Lobectomy provides the best survival for stage I lung cancer patients despite advanced age Ann Thorac Surg 2022 114 5 1824 1832 10.1016/j.athoracsur.2022.03.031 35351425
26. Sponholz S Schirren M Oguzhan S Schirren J Morbidity, mortality, and survival in elderly patients undergoing pulmonary metastasectomy for colorectal cancer Int J Colorectal Dis 2018 33 10 1401 1409 10.1007/s00384-018-3136-1 30056558
27. Hassan M Ehle B Passlick B Grapatsas K Lung resections for elderly patients with lung metastases: a comparative study of the postoperative complications and overall survival Curr Oncol 2022 29 7 4511 4521 10.3390/curroncol29070357 35877217
28. Wiedenroth CB Bandorski D Ariobi K Does age matter? Pulmonary endarterectomy in the elderly patient with CTEPH Thorac Cardiovasc Surg 2022 10.1055/s-0041-1740559 35038757
29. Newnham M Hernández-Sánchez J Dunning J Age should not be a barrier for pulmonary endarterectomy in carefully selected patients Eur Respir J 2017 50 6 1701804 10.1183/13993003.01804-2017 29217606
30. Grazioli V Ghio S Pin M Pulmonary endarterectomy in the octogenarian population: safety and outcomes J Cardiovasc Med (Hagerstown) 2021 22 7 567 571 33186233
31. Criner GJ Cordova F Sternberg AL Martinez FJ The national emphysema treatment trial (NETT) part II: lessons learned about lung volume reduction surgery Am J Respir Crit Care Med 2011 184 8 881 893 10.1164/rccm.201103-0455CI 21719757
32. Platz JJ Naunheim KS Critical analysis of the national emphysema treatment trial results for lung-volume-reduction surgery Thorac Surg Clin 2021 31 2 107 118 10.1016/j.thorsurg.2021.01.004 33926665
33. Abdelsattar ZM Allen M Blackmon S Contemporary practice patterns of lung volume reduction surgery in the United States Ann Thorac Surg 2021 112 3 952 960 10.1016/j.athoracsur.2020.08.058 33161015
34. Kaiwa Y Kurokawa Y Lung volume reduction surgery for chronic pulmonary emphysema in elderly patients Kyobu Geka 2005 58 8 709 713 16097623
35. Whittaker HR Connell O Campbell J Elbehairy AF Hopkinson NS Quint JK Eligibility for lung volume reduction surgery in patients with COPD identified in a UK primary care setting Chest 2020 157 2 276 285 10.1016/j.chest.2019.07.016 31381881
36. Schweigert M Solymosi N Dubecz A Surgery for parapneumonic pleural empyema—what influence does the rising prevalence of multimorbidity and advanced age has on the current outcome? Surgeon 2016 14 2 69 75 10.1016/j.surge.2014.05.002 24930000
37. Peek J Beks RB Hietbrink F Epidemiology and outcome of rib fractures: a nationwide study in the Netherlands Eur J Trauma Emerg Surg 2022 48 1 265 271 10.1007/s00068-020-01412-2 32506373
38. Cooper E Wake E Cho C Wullschleger M Patel B Outcomes of rib fractures in the geriatric population: a 5-year retrospective, single-institution, Australian study ANZ J Surg 2021 91 9 1886 1892 10.1111/ans.17064 34268852
39. Leard LE Holm AM Valapour M Consensus document for the selection of lung transplant candidates: an update from the international society for heart and lung transplantation J Heart Lung Transplant 2021 40 11 1349 1379 10.1016/j.healun.2021.07.005 34419372
40. Valapour M Lehr CJ Skeans MA OPTN/SRTR 2020 annual data report: lung Am J Transplant 2022 22 2 438 518 10.1111/ajt.16991 35266615
41. Hayanga AJ Aboagye JK Hayanga HE Contemporary analysis of early outcomes after lung transplantation in the elderly using a national registry J Heart Lung Transplant 2015 34 2 182 188 10.1016/j.healun.2014.09.028 25447584
42. Olson MT Elnahas S Biswas Roy S Outcomes after lung transplantation in recipients aged 70 years or older Clin Transplant 2022 36 1 e14505 10.1111/ctr.14505 34634161
43. Ehrsam JP Benden C Seifert B Lung transplantation in the elderly: influence of age, comorbidities, underlying disease, and extended criteria donor lungs J Thorac Cardiovasc Surg 2017 154 6 2135 2141 10.1016/j.jtcvs.2017.07.032 28823801
44. Shigemura N Toyoda Y Elderly patients with multiple comorbidities: insights from the bedside to the bench and programmatic directions for this new challenge in lung transplantation Transpl Int 2020 33 4 347 355 10.1111/tri.13533 31554020
45. Montgomery E Newton PJ Chang S Frailty measures in patients listed for lung transplantation Transplantation 2022 106 5 1084 1092 10.1097/TP.0000000000003823 34028384
46. Schaenman JM Diamond JM Greenland JR Frailty and aging-associated syndromes in lung transplant candidates and recipients Am J Transplant 2021 21 6 2018 2024 10.1111/ajt.16439 33296550
47. Asemota N Saftic I Tsitsias T King J Pilling J Bille A Quality of life in octogenarians after lung resection compared to younger patients Clin Lung Cancer 2022 3 2 e118 e30 10.1016/j.cllc.2021.05.008
48. Hino H Karasaki T Yoshida Y Risk factors for postoperative complications and long-term survival in lung cancer patients older than 80 years Eur J Cardiothorac Surg 2018 53 5 980 986 10.1093/ejcts/ezx437 29272371
49. Sirbu H Schreiner W Dalichau H Busch T Surgery for non-small cell carcinoma in geriatric patients: 15-year experience Asian Cardiovasc Thorac Ann 2005 13 4 330 336 10.1177/021849230501300408 16304220
50. Tong C Huang C Wu J Xu M Cao H The prevalence and impact of undiagnosed mild cognitive impairment in elderly patients undergoing thoracic surgery: a prospective cohort study J Cardiothorac Vasc Anesth 2020 34 9 2413 2418 10.1053/j.jvca.2020.03.011 32381306
51. Okada S Shimomura M Ishihara S Ikebe S Furuya T Inoue M Clinical significance of postoperative pulmonary complications in elderly patients with lung cancer Interact CardioVasc Thorac Surg 2022 35 2 ivac153 10.1093/icvts/ivac153 35640579
52. Miyazaki T Saji H Nakamura H The C-reactive protein to albumin ratio is a prognostic factor for stage I non-small cell lung cancer in elderly patients: JACS1303 Surg Today 2022 52 10 1463 1471 10.1007/s00595-022-02485-9 35211804
53. Stefani D Hegedues B Collaud S Torque Teno virus load in lung cancer patients correlates with age but not with tumor stage PLoS ONE 2021 16 6 e0252304 10.1371/journal.pone.0252304 34077485
54. Cooper L Gong Y Dezube AR Thoracic surgery with geriatric assessment and collaboration can prepare frail older adults for lung cancer surgery J Surg Oncol 2022 126 2 372 382 10.1002/jso.26866 35332937
55. Dezube AR Cooper L Jaklitsch MT Prehabilitation of the thoracic surgery patient Thorac Surg Clin 2020 30 3 249 258 10.1016/j.thorsurg.2020.04.004 32593358
56. Shepherd SJ Klein AA Martinez G Enhanced recovery for thoracic surgery in the elderly Curr Opin Anaesthesiol 2018 31 1 30 38 10.1097/ACO.0000000000000537 29120930
57. Robinson LA Tanvetyanon T Grubbs D Preoperative nutrition-enhanced recovery after surgery protocol for thoracic neoplasms J Thorac Cardiovasc Surg 2021 162 3 710 720.e1 10.1016/j.jtcvs.2020.06.016 32713631
58. Benzo R Wigle D Novotny P Preoperative pulmonary rehabilitation before lung cancer resection: results from two randomized studies Cancer Treat Res 2011 74 3 441 445
59. Rosero ID Ramírez-Vélez R Lucia A Systematic review and meta-analysis of randomized, controlled trials on preoperative physical exercise interventions in patients with non-small-cell lung cancer Cancers (Basel) 2019 11 7 944 10.3390/cancers11070944 31284372
60. Lai Y Huang J Yang M Su J Liu J Che G Seven-day intensive preoperative rehabilitation for elderly patients with lung cancer: a randomized controlled trial J Surg Res 2017 209 30 36 10.1016/j.jss.2016.09.033 28032568
61. Ginsberg RJ Rubinstein LV Randomized trial of lobectomy versus limited resection for T1 N0 non-small cell lung cancer. Lung cancer study group Ann Thorac Surg 1995 60 3 615 622 10.1016/0003-4975(95)00537-U 7677489
62. Mery CM Pappas AN Bueno R Similar long-term survival of elderly patients with non-small cell lung cancer treated with lobectomy or wedge resection within the surveillance, epidemiology, and end results database Chest 2005 128 1 237 245 10.1378/chest.128.1.237 16002941
63. Stamatis G Leschber G Schwarz B Survival outcomes in a prospective randomized multicenter phase III trial comparing patients undergoing anatomical segmentectomy versus standard lobectomy for non-small cell lung cancer up to 2 cm Cancer Treat Res 2022 172 108 116 10.1016/j.lungcan.2022.08.013
64. Minervini F Kocher GJ Bertoglio P Pneumonectomy for lung cancer in the elderly: lessons learned from a multicenter study J Thorac Dis 2021 13 10 5835 5842 10.21037/jtd-21-869 34795932
65. Mao Y Gao Z Yin Y Complete video-assisted thoracoscopic surgery and traditional open surgery for elderly patients with NSCLC Front Surg 2022 9 863273 10.3389/fsurg.2022.863273 35372482
66. Kozian A Kretzschmar MA Schilling T Thoracic anesthesia in the elderly Curr Opin Anaesthesiol 2015 28 1 2 9 10.1097/ACO.0000000000000152 25517623
67. Katlic MR Facktor MA Non-intubated video-assisted thoracic surgery in patients aged 80 years and older Ann Transl Med 2015 3 8 101 26046042
68. Wu CY Chen JS Lin YS Feasibility and safety of nonintubated thoracoscopic lobectomy for geriatric lung cancer patients Ann Thorac Surg 2013 95 2 405 411 10.1016/j.athoracsur.2012.10.082 23245449
69. Chen PH Chuang JH Lu TP Non-intubated versus intubated video-assisted thoracic surgery in patients aged 75 years and older: a propensity matching study Front Surg 2022 9 880007 10.3389/fsurg.2022.880007 35586501
70. Loran DB Zwischenberger JB Thoracic surgery in the elderly J Am Coll Surg 2004 199 5 773 784 10.1016/j.jamcollsurg.2004.08.008 15501120
| 36441200 | PMC9703435 | NO-CC CODE | 2022-11-29 23:21:42 | no | Chirurgie (Heidelb). 2022 Nov 28;:1-10 | utf-8 | Chirurgie (Heidelb) | 2,022 | 10.1007/s00104-022-01772-y | oa_other |
==== Front
Arch Pharm Res
Arch Pharm Res
Archives of Pharmacal Research
0253-6269
1976-3786
Pharmaceutical Society of Korea Seoul
36441472
1419
10.1007/s12272-022-01419-w
Review
Non-alcoholic fatty liver disease and liver secretome
Khan Muhammad Sohaib [email protected]
Lee Choongho [email protected]
http://orcid.org/0000-0001-5266-1722
Kim Sang Geon [email protected]
[email protected]
grid.255168.d 0000 0001 0671 5021 College of Pharmacy and Integrated Research Institute for Drug Development, Dongguk University-Seoul, Ilsandong-gu Dongguk-ro 32, Goyang-si, Gyeonggi-do 10326 South Korea
28 11 2022
126
28 9 2022
15 11 2022
© The Pharmaceutical Society of Korea 2022, Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Metabolism of carbohydrates and lipids and protein degradation occurs in the liver and contributes to the body's homeostasis by secreting a variety of mediators. Any imbalance in this homeostasis due to excess fat consumption and the pathologic events accompanying lipotoxicity, autophagy dysregulation, endoplasmic reticulum stress, and insulin resistance may cause disturbances in the secretion of the proteins from the liver and their physiologic modifications and interactions with others. Since the liver secretome plays a role in the regulation of fuel metabolism and inflammation not only in the liver per se but also in other organs, the proteins belong to the utmost targets for treating metabolic and inflammatory diseases (e.g., COVID-19), depending on the available and feasible approaches to controlling their biological effects. However, in this era, we still come across new liver-derived proteins but are yet unable to entirely understand the pathologic basis underlying disease progression. This review aims to provide an updated overview of liver secretome biology with explanatory mechanisms with regard to the progression of metabolic and inflammatory liver diseases.
Keywords
Liver disorders
Hepatokines
ER stress
COVID-19
Autophagy
http://dx.doi.org/10.13039/501100014188 Ministry of Science and ICT, South Korea 2017K1A1A2004511 Kim Sang Geon
==== Body
pmcMetabolic disorders and non-alcoholic fatty liver disease
Metabolic syndrome can be diagnosed if three or more of the following factors are present: fasting glucose ≥ 100 mg/dL, blood pressure ≥ 130/85 mm Hg, triglyceride level ≥ 150 mg/dL, high-density lipoprotein cholesterol level < 40 mg/dL in men or < 50 mg/dL in women, and waist circumference (for Westerners, > 100 cm in men or 88 cm in women; for Asians, > 88 cm in men or > 80 cm in women) (Carr et al. 2016). Metabolic disorders such as diabetes and hyperlipidemia are becoming increasingly common in modern society. Adults in Western countries suffer from obesity (90%), diabetes (50%), and hyperlipidemia (90%) (Le et al. 2017). The major reason for this high prevalence of metabolic disorders is due to increased sedentary lifestyles and excess calorie intake, leading to energy imbalances. Humans maintain their health by regulating lipid metabolic rate via lipogenesis, lipoprotein absorption, and secretion.
The liver is considered the metabolic hub of the body because all ingested nutrients pass through the liver after intestinal absorption. Hence, the liver quickly senses any nutritional changes, which then alters metabolic activities to maintain homeostasis. Therefore, any disturbances in fuel consumption of the liver (e.g., steatosis and steatohepatitis) are often indicators of a metabolic disorder. The hypothalamus is also an essential regulator of energy and weight homeostasis. Evidently, mutant genes contribute to the underlying basis of metabolic disorders (Hochberg and Hochberg 2010). Therefore, metabolic disorders result from various combined factors, including genetic variations, nutritional alternation, and hormonal impairment. Conditions commonly associated with metabolic syndrome include obesity (Cornier et al. 2008), diabetes, and nonalcoholic fatty liver diseases (Eckel et al. 2010), which will be discussed in more detail.
Obesity
Obesity is defined as abnormal or excessive fat accumulation that presents a risk to health. According to the World Health Organization (WHO), individuals with a body mass index (BMI) over 30 kg/m2 are considered obese. This cut-off value was determined from the point where the typical metabolic complications of obesity increased twofold. However, among the Asian population, who have a high prevalence of metabolic complications with lower BMI compared with White, Hispanic, or Black populations, the BMI cut-off value is 25 kg/m2.
A WHO report (2021) mentioned that nearly 2.2 billion adults (40%) worldwide are obese. Another study subdivided obesity by gender, showing that 200 million men and 300 million women worldwide are obese (Polyzos et al. 2019). Data on obesity prevalence among Asians in 2014 showed that about 40% of adults in China, 30% in Japan and India, and 27% in Korea were considered obese (Fan et al. 2017). Of them, more than 39 million were considered morbidly obese (BMI > 40 kg/m2). Surveillance data from the United States for 2009 and 2010 revealed that more than 15% of children and adolescents were obese, with a consequent increase in pre-diabetic rates (Roth 2015). Data from the Global Obesity Observatory (2021) revealed that seven out of every ten Indian adults and four out of every eleven children are obese (Observatory 2021a), while in Pakistan, six out of every ten adults and five out of every eleven children are obese (Observatory 2021b). A cohort study conducted on 975 Chinese children aged between 6 and 13 years and continued for 6 years afterward showed that obesity increased 2.8 times during adolescence, while another study included 204 Chinese children aged between 6 and 17 years and followed up for 13 years revealed that obesity prevalence increased 5.8 times (Pan et al. 2021). Therefore, obesity will continue to be a severe global health problem due to its rapid increase among children and adolescents.
Studies have shown a linear increase in NAFLD prevalence with increased obesity. In a study of 181 morbidly obese patients with severe average BMI (= 45.1 ± 8.3 kg/m2), 126 patients (69.6%) exhibited symptoms of NAFLD (Ooi et al. 2021). The study's results showed that more than 33% of obese patients suffered from NAFLD, with more than one-third having bariatric surgery (Machado et al. 2006). NAFLD diagnoses are increasing exponentially in relation to obesity in the United States. For example, 4.17 million cases of NAFLD were reported in 2008, rising to 83.1 million cases in 2015.
Diabetes
Diabetes mellitus (DM) is a chronic disorder that causes abnormal metabolic regulation of glucose as well as vascular and neuropathic complications (Crandall and Shamoon 2020). DM diagnosis depends on the detection of elevated fasting blood glucose levels (≥ 126 mg/dL). There are two types of diabetes: type 1 diabetes (T1D, insulin-dependent or juvenile-onset diabetes) and type 2 diabetes (T2D, non-insulin-dependent or adult-onset diabetes). The pathophysiology of DM can be explained by insulin deficiency and reduced carbohydrate metabolism. Insulin maintains glucose homeostasis by promoting glucose storage in the fed state and releasing it in the fasting state (Crandall and Shamoon 2020). Insulin resistance (IR) is defined as when the signaling pathways of insulin are impaired in principal target organs and tissues such as muscle, fat, and the liver. IR can result in compensatory hyperinsulinemia to maintain normal glucose homeostasis. If compensation is not adequate, hyperglycemia and T2D can eventually develop.
Metabolic syndrome, which includes diabetes, is considered one of the significant life-threatening conditions of the twenty-first century because it can cause serious complications such as cardiovascular disease and stroke. In 2040, it is estimated that 642 million people worldwide will have T2D (Yang et al. 2020), which has shown a linear relationship with elevated BMI, reflecting increasing obesity rates. A meta-analysis using observational data from 20 countries revealed that more than 50% of T2D patients also suffered from NAFLD (Younossi et al. 2019; Targher et al. 2021). The literature suggests a bi-directional relationship between NAFLD and T2D. Moreover, T2D has been proven to be one of the main risk factors for patients developing NAFLD and HCC (Anstee et al. 2013; Powell et al. 2021). The underlying mechanism for high co-incidence of T2D and NAFLD is explained as follows; under IR conditions, adipose tissue becomes dysfunctional, reducing its ability to uptake circulating lipids and enhance lipolysis, even in high-fat diet (HFD) conditions (Yaribeygi et al. 2019). Therefore, these dysfunctional adipose tissues release large amounts of circulating free fatty acids (FFA), which can accumulate in the liver and cause NAFLD (Hammoutene and Rautou 2019).
Non-alcoholic fatty liver disease (NAFLD)
NAFLD is clinically diagnosed if the liver consists of > 5% fat, as monitored by liver imaging or biopsy in the absence of secondary causes of fat accumulation such as chronic alcohol abuse (defined as more than one drink per day for women or two for men) (Carr et al. 2016). NAFLD can encompass a wide variety of liver diseases ranging from simple fatty liver (i.e., simple steatosis) with no inflammation to non-alcoholic steatohepatitis (NASH) with accompanying steatosis, inflammation, and hepatocyte injury. This can manifest as hepatocytes ballooning with or without fibrosis, which may further proceed to liver fibrosis or liver cancer (Piccinin et al. 2019; Makri et al. 2021). Fibrosis can be histologically categorized into four stages, ranging from stage 0 (no fibrosis) to stage 4 (cirrhosis) (Powell et al. 2021).
NAFLD is currently the most common liver disease globally and has been reported to affect 30% of people over the age of 18 (Hou et al. 2021). A cohort study of 139,056 Koreans between 2011 and 2013 showed an association between a sedentary lifestyle and NAFLD prevalence in young and middle-aged people (Ryu et al. 2015). Another study showed that NAFLD is strongly associated with obesity and T2D (Lonardo et al. 2019). As a consequence of this strong correlation, NAFLD has also become known as metabolic dysfunction-associated fatty liver disease (MAFLD) (Eslam et al. 2020; Makri et al. 2021). As metabolic syndrome becomes more common, so does the incidence of NAFLD (i.e., a worldwide incidence of ~ 25%, ranging from 13% in Africa to 42% in Southeast Asia). There are expected to be an estimated 100.9 million cases of NAFLD by 2030, a 21% increase from 2015, with a 33.5% prevalence in people above the age of 15 and 28.4% for people of all ages (Estes et al. 2018). Data monitored by national healthcare providers between 1998 and 2015 were gathered, and a new comprehensive analysis was conducted to investigate the prevalence of NASH in various countries and regions. The results of the above-mentioned study revealed that the annual medical costs for treating NAFLD exceeded $100 billion in the United States alone (Mundi et al. 2020).
Pathogenesis of liver diseases
NAFLD, newly named MAFLD, is certainly associated with metabolic dysfunction. Here, the pathogenesis of metabolic diseases will be discussed in the context of lipotoxicity, autophagy dysregulation, endoplasmic reticulum stress, IR, and other targets. In a recent study, the analysis of single-cell RNA transcriptome has been used to find a cell type-specific role in gene expression for the progression of liver diseases including NAFLD (Su et al. 2021).
Lipotoxicity
One of the most well-known disease progression mechanisms in NAFLD is steatosis. When a hepatocyte’s ability to synthesize triglycerides overwhelms its ability to dispose of them, triglycerides will accumulate inside them as fat. Although triglycerides are not toxic per se, their precursors, such as fatty acids and other metabolic byproducts, such as reactive oxygen species (ROS), are toxic to hepatocytes. The accumulation of these byproducts is known as lipotoxicity (Yoon et al. 2021). Because of impaired lipid metabolism, NAFLD patients experience inter- and intrahepatic lipid buildups such as enhanced hepatic FFA intake and very-low-density lipoprotein synthesis, dysregulation of triglyceride export, and reduced levels of high-density lipoproteins and cholesterol in the blood (Katsiki et al. 2016). Inflammation also promotes cytokine production, gut-derived products (e.g., lipopolysaccharide), and hepatotoxic mediators, which can aggravate NAFLD if hepatocytes are exposed to them (Diehl and Day 2017).
FFAs are hydrophobic, which increases their permeability across the cell membranes. However, a few transport proteins facilitate their transport (e.g., plasma membrane fatty acid-binding protein) and fatty acid translocases such as CD36 (Rada et al. 2020). The fatty acid translocase CD36 has a high-affinity receptor for long-chain FFAs, contributing to enhanced fat surge, excessive lipid storage, and metabolic dysfunction. These proteins are also involved in lipid metabolisms such as fat intestinal absorption and fatty acids consumption by muscle, adipose tissue, and liver (Rey et al. 2020). An exome-wide association study revealed that increased levels of VLDL were adversely found in T2D patients (Liu et al. 2017).
NADP + -dependent aldo–keto reductase family 1, member 10 (AKR1B10), also named as ARL-1 protein, is mainly expressed in the small intestine and colon (Gallego et al. 2007). The levels of AKR1B10 were found to be higher in patients with HCC (Heringlake et al. 2010) or adenocarcinoma of the lung (Fukumoto et al. 2005), renal cancer, and breast cancer (Ma et al. 2012; Kanno et al. 2019). The studies have proven that AKR1B10 is involved in regulating lipotoxicity and de novo-lipogenesis; lipid peroxidation produces electrophilic carbonyls, aggravating DNA damage by interacting with nucleophiles and causing carcinogenesis and apoptosis (Luo et al. 2011; Ye et al. 2019).
Autophagy dysregulation
The liver is the primary organ involved in the detoxification of chemicals within the body. Maintaining homeostasis between the generation of new proteins and the destruction of damaged proteins in eukaryotic cells involves two main pathways: the ubiquitin–proteasome system (UPS) for short-lived proteins and the autophagy-lysosomal pathway for longer-lived proteins (Martinet et al. 2009). Several articles have reported on the correlation between autophagy and lipid metabolism. Autophagy causes the transfer of intracellular materials, such as denatured proteins, fat droplets, and dysfunctional mitochondria, to the lysosomes for their destruction. As a “housekeeper” of cellular contents, autophagy not only inhibits the progression of steatosis and fatty hepatitis but also prevents hepatocyte injury (Kwanten et al. 2014). However, the fizzy lifestyle and intense caloric food intake have increased the obesity ratio, negatively affecting the regulation of autophagy.
Before analyzing the possible pathogenic mechanism of NAFLD driven by dysfunctional autophagy, it is necessary to review how the intracellular contents are controlled. Lipid accumulation in the hepatocytes could result in decreased autophagic activity, bile acid fluctuations, increased endoplasmic reticulum (ER) stress, inflammatory response, and disturbed gut microbiota, all of which can contribute to NAFLD progression (Friedman et al. 2018; Yueh et al. 2020). The relationships between autophagic imbalance and hepatic diseases have been studied (Kwanten et al. 2014; Kim and Kim 2020). In addition, the consequential excessive storage of lipids in hepatocytes due to impaired autophagy has been shown to cause apoptosis, exacerbating NAFLD (Tanaka et al. 2016).
There are three types of autophagy: macro-autophagy, micro-autophagy, and chaperone-mediated autophagy (Amir and Czaja 2011). Lipid droplets (LDs) of various sizes are metabolized by macroautophagic engulfment (Singh et al. 2009). Macroautophagy occurs when autophagosomes and lysosomes fuse together (Amir and Czaja 2011). Autophagosomes, submerged cytosolic double-membrane structures attached with lysosomal enzymes, degrade the cellular constituents, and then autophagy-related genes (Atgs) are responsible for regulating the overall process (Czaja 2011). Atg knockout mice exhibited a fourfold increase in liver mass due to the failure to degrade appropriate cellular components (Czaja 2011). While lysosomal lipase degrades lipoproteins via endocytosis, macroautophagy activates the cleavage of triglycerols and cholesterols stored in hepatocytes and releases FFAs through a process known as “lipophagy.” In addition, chaperone-mediated autophagy stimulates LDs' metabolism, resulting in lipolysis via either cytosolic lipases or macroautophagy (Zhang et al. 2020).
To function properly as the primary initiative for autophagy, autophagosomes need to be formed. This step is mediated by the unc-51-like kinase 1 (ULK1), serine/threonine-protein kinase. One of the possible mechanisms underlying autophagy dysfunction in NAFLD is due to the inhibition of ULK1 by mTOR. Research has shown that chronic caloric intake is directly related to mTOR activation, which leads to the complex formation of mTOR complex 1 (mTORC1) and mTOR complex 2 (mTORC2) (Chung and Chung 2019). Phosphorylation by mTORC1 of ULK1 at Ser758 and Ser757 in human and mouse cells, respectively, interferes with AMP-activated protein kinase (AMPK)-binding ULK1 phosphorylation and inhibits its activation. Since autophagy initiation by ULK1 is inhibited via mTORC1 activation, the action of autophagy is reciprocally regulated by mTORC1 (Fig. 1) (Kim and Guan 2015). Studies have shown that Atg7 knockdown and reduced LC3-II cause decreased levels of autophagic flux with hepatomegaly (Kim et al. 2013; Tsai et al. 2017). Regarding protease enzymes, calpain 2 expression was found to increase compared to that of calpain 1. Research has also shown that calpain 2 activation leads to the loss of Atg3 and Atg7 activities, decreasing autophagy in hepatocytes with the progression of fatty liver and IR (Kim et al. 2008) (Fig. 1).Fig. 1 On the left, increased caloric intake affects the mTORC1, AMPK and ULK1 network and inhibits autophagy. On the right, the activation of calpain-2 results in the loss of Atg3 and Atg7 required to activate autophagy. The loss of function would result in decreased autophagy activity. mTORc1 mammalian target of rapamycin complex 1, ULK1 Unc-51-like kinase 1, Atg3 autophagy related 3, and Atg7 autophagy related 3
Another mechanism proposed for the inhibition of macroautophagy in hepatocytes results from the dysregulation of a protease called autophagy-related protein 4B (ATG4B) and RAS-related protein-8b (Rab-8b), which is mediated with liver X receptor α (LXRα) (Kim et al. 2020). In this event, LXRα transcriptionally induces the MIRLET7A and MIR34A genes to inhibit ATG4B and Rab-8B, suppressing mitochondria biogenesis and fuel consumption. Persistent over-activation of LXRα (due to HFD feeding and/or excessive calorie intake), therefore, worsens NAFLD (Kim et al. 2021).
Alcoholic liver disease (ALD), including alcoholic hepatitis, is the most prevalent liver disease worldwide. ALD is defined by the accretion of neutral lipids and lipid metabolism disruption prior to liver damage. Obesity is another risk factor for ALD development; the incidence of ALD increases by 2–3 times in individuals with steatosis (Parker et al. 2019). Alcohol-induced, kynurenine-mediated AhR activation in hepatocytes is responsible for autophagy inhibition, exacerbating liver steatosis. The importance of different types of cells involved in the NAFLD progression has also been demonstrated at the molecular levels (Jin 2020; Kumar et al. 2021). Studies have provided detailed insights into the role of metabotropic interactions in hepatic parenchymal (hepatocytes) and non-parenchymal cells. These interactions negatively affect autophagy, and therefore mitochondrial activity and biogenesis, via various nuclear receptors as lipid or sensors/or amino acid metabolites in aggravating alcoholic and non-alcoholic liver diseases, either pathological or non-pathological pathways (Choi et al. 2019, 2021).
Endoplasmic reticulum (ER) stress
Hepatocytes have numerous ERs, similar to other secretory cells, because of their protein synthesis capability. ER is involved in the folding of secreted and transmembrane proteins, a process achieved with the assistance of chaperone proteins. The ER also houses enzymes that synthesize cholesterol and triacyl-glycerides (TAG) for energy storage (Little et al. 2007). However, increased levels of saturated fatty acids trigger the excess storage of misfolded or unfolded proteins in the ER lumen, a process known as ER stress. In order to restore homeostasis, ER stress accelerates the unfolded protein response (UPR), a signal transduction pathway located in the ER lumen, which is also known as the regulator of ER proteostasis surveillance (Wang and Kaufman 2016). UPR adaptively stimulates the increased expression of ER proteins, including ER membrane proteins, to extend the organelle space and produce more chaperone proteins required for protein folding. Additionally, UPR activation reduces the total protein synthesis, thereby lessening the workload of the ER, enhancing the secretion of folded proteins, and eliminating misfolded proteins via autophagy and ER-associated protein degradation (ERAD) (Hetz et al. 2020). The ERs’ physiological activity shears the inter-progressive relationship between ER stress and fatty acid synthesis. Chronic ER stress is also associated with NAFLD as it contributes to lipid accumulation, inflammation, and hepatocyte apoptosis (Liu et al. 2021).
The UPR acts as an ER stress sensor through various major pathways: protein kinase RNA-like ER kinase (PERK), eukaryotic translation initiation factor 2α(eIF2α), inositol-requiring protein 1 α(IRE1α), X-box binding protein 1 (XBP1), and activating transcription factor 6α(ATF6) (Xu et al. 2021). Both PERK and IRE1α are widely known as type I transmembrane proteins with the same ER luminal and cytosolic Ser/Thr kinase domains. However, ATF6α is a type II transmembrane protein and has a cytosolic cyclic AMP response element-binding protein (CREB) ATF with a basic leucine zipper domain (Oslowski and Urano 2011). In the resting state, an ER chaperone known as immunoglobin binding protein (BiP) binds to IRE1α or PERK. This binding deactivates the ER stress sensors pathway. When the ERs are stressed, unfolded or misfolded proteins accumulate. BiP binds to these unfolded or misfolded protein peptides, thereby deactivating the ER stress sensors pathway. Alternatively, the unfolded proteins bind directly to IRE1α or PERK activated after unbinding from BiP (Hetz et al. 2020).
PERK activation results in the phosphorylation of eukaryotic translation initiation factor 2α (eIF2α), which reduces the general protein translation to relieve the ER workload. eIF2α selectively enhances the production of the stress-inducible transcription factor, ATF4. It activates to express the genes associated with amino acid metabolism, antioxidative response, autophagy, and ER protein folding. The chronic activation of ATF4 stimulates the expression of transcription factor C/EBP homologous protein (CHOP) (Li et al. 2018). Usually, CHOP remains dormant; however, under persistent stresses such as increased toxins, metabolic inhibitors, and nutrient deprivation, CHOP is activated and arrests the growth and induction of DNA damage-inducible gene 153 (GADD153) (Batchvarova et al. 1995). The overexpression of CHOP sensitizes hepatocytes to apoptosis by promoting ER stress, whereas the opposite was shown with decreased CHOP expression. Therefore, decreased CHOP levels attenuate ER stress-induced apoptosis in the liver (McCullough et al. 2001).
IRE1α contains ribonuclease and kinase domains within the cytosolic region. Under ER stress conditions, the activation of IRE1α leads to the stimulation of the ribonucleolytic activity of itself, resulting in a small intron being excised from the XBP1 mRNA (Fig. 2). This process is known as non-conventional splicing. This excision/splicing causes a shift in the translational reading frame, leading to the production of an active XBP1 transcription factor, XBP1s. It also upregulates genes related to protein folding, translocation, and secretion, as well as degradation (Calfon et al. 2002). Additionally, IRE1α exerts its ribonucleolytic activity on mRNAs in the ER membrane, encoding specific secretory proteins such as proinsulin or IRE1α itself. This process, which reduces the abundance of mRNA and the protein folding load, is known as regulated IRE1-dependent decay (RIDD) (Deng et al. 2013). Under chronic and excessive ER stress, IRE1α activates Jun amino-terminal kinase (JNK) and apoptosis signal kinase 1 (ASK1) by engaging the adaptor protein and tumor necrosis factor receptor-associated factor 2 (TRAF2) (Calfon et al. 2002). JNK phosphorylation stimulates proapoptotic Bcl-2 only-like protein 11 (Bim), which ultimately induces apoptosis. Studies have shown that nuclear factor kappa B (NF-κB) inhibits JNK activation, preventing the induction of apoptosis in normal cells. However, prolonged stress conditions mean that apoptosis through JNK activation prevails in the antiapoptotic function by NF-κB. Stimulating proapoptotic BH3-only proteins transcriptionally or post-transcriptionally leads to proapoptotic Bax and Bak stimulation by antagonizing antiapoptotic members (Fig. 2) (Hetz et al. 2006).Fig. 2 On the left, chronic caloric intake causes TG accumulation and then activates IRE1α, causing cleavage of XBP1 to XBP1s, which promotes unfolded protein response in early phase. On the right during late phase, IRE1α activate TRAF2 which activate ASK, leading to JNK phosphorylation. NF-κB inhibits the phosphorylation of JNK. Persistent activation of JNK, however, leads to Bim activation, which triggers apoptosis. TG triglyceride, IRE1α inositol-requiring transmembrane kinase/endoribonuclease 1α, TRAF2 tumor necrosis factor receptor-associated factor 2, ASK1 apoptosis signal kinase 1, NF-κB nuclear factor-κB, JNK c-Jun N-terminal kinase, and Bim bcl-2-interacting mediator of cell death
ATF6 is translocated from the ER to the Golgi apparatus under ER stress conditions. In the Golgi apparatus, ATF6 is split, and a fragment known as the basic leucine zipper domain (bZip) transcription factor is released. This transcription factor induces gene expression after being translocated inside the nucleus. Both bZip and XBP1s act similarly to the one described above (Lee et al. 2002).
Gα12 overexpression is promoted by ER stress via the IRE1-Xbp1 pathway, which subsequently feeds forward an ER stress-induced vicious cycle in the hepatocytes. Thus, ER stress-induced Gα12 induction may cause hepatocyte death, leading to drug-induced liver disease symptoms. This process is notable because Gα12 overexpression can initiate arachidonate 12-lipoxygenase (ALOX12)-dependent lipid peroxide generation via Rho-associated kinase 1 (ROCK1), facilitating polyunsaturated fatty acids (PUFA) peroxidation, hepatocyte ferroptosis, and eventually fibrosis (Tak et al. 2022). The dysregulation of miR-15a aids in the induction of ALOX-12. ER stress has also been shown to cause liver fibrosis in activated hepatic stellate cells (HSCs), as indicated by the significant association between ER stress and HSC activation in animal models and patients (Gupta et al. 2010). At the molecular level, ER stress-induced dysregulation of primary-miR-18a processing leads to SMAD2 overexpression via the direct phosphorylation of hnRNPA1 at the Thr51 site by PERK (Koo et al. 2016). In cancer biology, sorafenib resistance is attained by ER stress via the upregulation of PMK2 by miR-188-5p/hnRNPA2B (Zhou et al. 2021).
Collectively, under physiological ER stress levels, the UPR sensors are activated to maintain homeostasis, resulting in a reduction of protein synthesis, increased protein folding, promotion of autophagy, and increased degradation of misfolded proteins. However, when the ER stress level exceeds the threshold, ER stress-mediated cell death and apoptosis are initiated (Hughes and Mallucci 2019).
Insulin resistance
IR is primarily responsible for the pathogenesis of T2D, NAFLD, and its more severe form, NASH (Holt et al. 2006). The hepatic IR augmented by FFA influx and the overstimulation of pro-inflammatory cytokines and lipid intermediates in the liver is explained by the impediment to the phosphorylation of insulin receptors (Petito-da-Silva et al. 2019). However, the exact mechanism of IR is still debatable and is hypothesized on the consequences of following events in the body. As per the traditional “two-hit” hypothesis, hepatic lipid deposition is secondary to an inactive/desk-bound lifestyle, having HFD intake that leads to obesity, and a consequent triggering of IR, which serves as a first fundamental hit to sensitize the liver, and subsequently activates inflammatory cascades (Rada et al. 2020).
Others
Excessive caloric intake unbalances hepatic physiological functioning due to the dysfunction of organelles, which can cause a variety of metabolic syndromes, including diabetes, fatty liver disease, and obesity. An electron shift occurs between substrates to oxygen, whereby protons are eliminated from the mitochondrial complex and maintain a chemiosmotic gradient that further boosts ATP production due to ATP synthase activity (Li et al. 2000). Recent studies have revealed that patatin-like phospholipase domain-containing protein 3 (PNPLA3) is associated with central fat accumulation (Trepo et al. 2016). Central fat accumulation also releases chronic inflammatory mediators, such as cytokines, and disrupts the insulin-glucagon balance (Liu et al. 2020). Another recent study has confirmed that phosphorylation of the iroquois homeobox gene 3 (IRX3, a protein involved in tissue/organ patterning or development) by the JNK leads to obesity and macrophage infiltration (Yao et al. 2021). To date, it is well established that macrophagic infiltration is involved in the progression of NAFLD and liver steatosis (Lefere et al. 2020).
The literature has shown that the most commonly used mouse models for obesity and NAFLD (Koo et al. 2017). HFD-induced obesity and hyperglycemia in animals result in elevated levels of Gα13 in skeletal muscle. In addition, a new scientific explanation for Gα13 has recently provided a new molecular mechanism for diabetes when the liver is compromised. In response to hyperglycemic stimuli, the challenged liver tissues show a decrease in Gα13 levels in both mice and humans. Secretome analysis has revealed that a decrease of Gα13 promotes the production of inter-a-trypsin inhibitor heavy chain 1 (ITIH1) in the liver. The circulation of ITIH1 is then associated with IR in peripheral tissues, including skeletal muscle and adipose tissue. Mechanistically, the reduced Gα13 levels in hepatocytes activate O-GlcNAc transferase induction, which is responsible for IR, via the stabilization of ITIH1 and its binding with HA (Kim et al. 2019).
The ligands that specifically activate G protein-coupled receptor (GPCR) coupling to Gα12 members (the ligands of which include sphingosine-1-phosphate (S1P), lysophosphatidic acid, angiotensin II (Ang II), thrombin, and endothelin-1) enhance liver fibrosis (Alexander et al. 2021). Of the Gα12 members, Gα12 has the potential transforming ability, cell proliferation, migration, and inflammation (Suzuki et al. 2009). Since Gα12 acts through GPCRs, resulting in enhanced signaling cascades (Suzuki et al. 2009), changes in Gα12 levels amplify or dampen the biological and physiological processes (Okashah et al. 2020). Gα12 overexpression in activated HSCs promotes liver fibrosis because of the downregulation of miR-16 and miR-29a (Huang et al. 2015; Kim et al. 2018), which is directly related to JNK-dependent ATG12-5 (Kim et al. 2018; Wible et al. 2019). Therefore, GPCR substrates, G proteins, and related dysregulation of microRNA mediators can all potentially contribute to NAFLD, ALD, and liver fibrosis.
In another study, E2 and ERα were found to be mutually associated with Gα12 in patients with HCC and their overall prognoses. However, ERα expressions were reported to have an inverse relationship to Gα12 in cell-based experiments and human tissue (Yun et al. 2022). Ligand-mediated activation of ERα restrains Gα12 gene transactivation, leading to microRNA-141 and -200a downregulation via the Gα12–RhoA axis (Yun et al. 2022) and promotes the amoeboid movement of cancer cells. In this paper, Gα12 antagonism by ERα can be explained by the gender discrepancy in HCC prognosis.
Roles of liver secretome
In association with the above pathologic factors, recent attention was paid to the roles of liver secretory proteins in liver disease progression since they play diverse roles in regulating fuel metabolism and inflammatory processes in different cells and organs. In this section, we will discuss the representative liver secretory proteins associated with metabolic and inflammatory liver diseases (Table 1).Table 1 Representative liver secretory proteins and their functions
Proteins Reported activities Expression References
Proteins associated with acute phase response protein and proinflammation
Fetuin-A ∙ Raised serum levels in liver fibrosis, NAFLD and T2D
∙ Fetuin-A promotes IR and raises blood glucose by activating TLR4
∙ Gα13 hepatic knockout leads to the surge of fetuin-A
∙ Fetuin-A levels were reported lowest in patients with severe alcoholic liver cirrhosis
Hepatocytes Pal et al. (2012), Meex et al. (2015), Prystupa et al. (2016), Peter et al. (2018) and Kim et al. (2019)
Fetuin B ∙ Elevated concentrations reported in T2D and hepatic steatosis
∙ Fetuin B inhibits LXR-SREBP1c and increases fatty acid oxidation, enhances IR and blood sugar levels
Hepatocytes Meex et al. (2015), Peter et al. (2018) and Zhou et al. (2019)
Serum Amyloid A-2 protein
(SAA2)
∙ 1.3-fold plasma concentrations elevated after treating mice with the HFHCD diet
∙ Serum SAA2 levels increased in obese patients
∙ Gα13 loss results in SAA2 upregulation
Liver Chiba et al. (2009), Samsoondar et al. (2017), Kim et al. (2019)
Ceruloplasmin
(Cp)
∙ Aceruloplasminemia positively related to diabetic and HCC patients
∙ Mutation in Cp genes resulted in hyper-ferritinemia and fibrosis
∙ Cp levels were found lower in children with higher NAFLD score due to hepatic inability to produce Cp
∙ Gα13 loss results in Cp overexpression
Liver and lungs Niederau et al. (1996), Cairo et al. (2001), Nobili et al. (2013), Kim et al. (2019) and Corradini et al. (2021)
Alpha-1-acid glycoprotein (AGP) ∙ In NAFLD patients, AGP shows a positive correlation with collagen
∙ AGP desialyation raises its expression in liver diseases
∙ Levels of AGP showed a severity-dependent increase in patients with mild to severe steatosis
∙ HFD-fed primary hepatocytes of Gα13 showed a pronounced surge of AGP
Hepatocytes Serbource‐Goguel et al. (1983), Younossi et al. (2017), Kim et al. (2019) and Liu et al. (2022)
Hemopexin (HPX) ∙ Increased heme toxicity in animals and humans was observed with reduced levels of HPX
∙ Loss of Gα13 positively regulates HPX levels
∙ HPX knockout mice exhibit overstimulation of ROS and TLR4 receptors resulting in chronic inflammation
∙ HPX levels were found elevated in HCC patients compared to those with cirrhosis without HCC or with fibrosis
Liver Debruyne et al. (2010), Lin et al. (2015), Vinchi et al. (2016) and Kim et al. (2019)
Binding carrier of hormones or lipids
RBP4 ∙ Its levels were found to be raised in obesity and T2D
∙ Serum RBP4 showed controversial results in NAFLD patients
∙ RBP4 increases IR by the stimulation of JNK and TLR4
∙ The levels of RBP4 were found elevated by the loss of Gα13
Hepatocytes and adipocytes Graham et al. (2006), Wu et al. (2008), Nobili et al. (2009), Norseen et al. (2012), Kim et al. (2019), Zhong et al. (2019) and Wang et al. (2020)
Precursors of receptor ligands or hormones
Angiopoietin-like protein 1
(ANGPTL1)
∙ In-vitro studies showed that ANGPTL1 inhibits the hepatocyte's growth factor-induced MET phosphorylation
∙ It was also found to be involved in the suppression of metastasis of hepatoma cells, whereas its serum levels were reduced in HCC
∙ ANGPTL1 expressions were reduced in colorectal carcinoma
Vascularized tissue and colon cells Kuo et al. (2013), Chen et al. (2016) and Chang et al. (2022)
ANGPTL2 ∙ ANGPL2 expressed in patients with HCC, also in thyroid cancer and lung carcinoma Hepatocytes Gao et al. (2015), Wei et al. (2017) and Yang et al. (2019)
ANGPTL3 ∙ The serum concentration of ANGPTL3 reduced in obese patients
∙ Single or multiple dose treatments of ANGPTL3 reduced plasma HDL, LDL and increased plasma TAG
Hepatocytes Musso et al. (2012) and Graham et al. (2017)
ANGPLT4 ∙ Obese patients showed enhanced serum levels
∙ Loss of ANGPTL4 results in inhibition of lipoprotein lipase and raised HDL levels
∙ HCC and ER stress leads to diminished ANGPTL4 levels
Liver and adipose tissues Romeo et al. (2007), Romeo et al. (2009), Lichtenstein et al. (2010), Singh et al. (2021) and Spitler et al. (2021)
ANGPTL6 ∙ ANGPTL6 knockout mice showed weight gain, increases in rectal temperature, basal metabolic rate, food intake, enhanced HDL and TAG, and aggravated IR and blood glucose levels
∙ ANGPTL6 serum concentrations were raised in NAFLD and T2D
Hepatocyte-induced circulating factors Oike et al. (2005), Ebert et al. (2009) and Ma et al. (2019)
ANGPTL8/betatrophin ∙ ANPTL8 is commonly known as HCC-associated protein and increased in T2D and liver steatosis with elevated TAG, IR, and glucose levels
∙ ANGPTL8 inhibits AMPK-mediated activation of SREBP1c and results in HCC progression
Liver von Loeffelholz et al. (2017) and Wang et al. (2018)
Fibroblast growth factor 1 (FGF1) ∙ Elevated in obese patients as well as in liver fibrosis and biliary proliferation
∙ Administration of FGFR1 antagonist results in the protection against hepatic fibrosis
Hepatic stellate cells O’Brien et al. (2022)
FGF2 ∙ In BDL mice, administration of FGF2 ameliorates liver fibrosis
∙ FGF2 levels were found elevated in patients with HCC
Hepatic stellate cells and adipocytes Jim-No et al. (1997) and Sato-Matsubara et al. (2017)
FGF 19, FGFR4 and beta-klotho ∙ FGFR4 impaired signaling results in diminished expression of FGF19
∙ FGF19 serum levels were in parallel to bile acid in NASH patients
∙ ER stress elevates the b-klotho and FGF19 in patients with HCC
Ileum, liver, and cholangiocytes Miura et al. (2012) and Jiao et al. (2018)
FGF21 ∙ FGF21 analog leads to amelioration of steatosis, inflammation, and IR
∙ Deficiency of CYP2E1 activates the PPARα-FGF21 axis, increasing the adipose browning with reduced obesity
Liver and Pancreas Zarei et al. (2020), Su et al. (2021) and Zhang et al. (2022)
Pigment epithelium-derived factor (PEDF) ∙ PEDF synergizes the breast cancer proliferation
∙ Hepatic metastasis was negatively regulated with PEDF
∙ PEDF levels were higher in T2D with CKD, and with the loss of Gα13 LKO, PEDF levels were increased
Hepatocytes and adipocytes Fitzgerald et al. (2012), Hui et al. (2014), Protiva et al. (2015) and Kim et al. (2019)
Hepassocin
(HPS)
∙ Serum analysis showed elevated in patients with T2D and NAFLD
∙ ER-stress in primary hepatocytes increased HPS in a dose-dependent manner
Liver Wu et al. (2013), Jung et al. (2018), Abdelmoemen et al. (2019) and Watt et al. (2019)
Angiotensinogen ∙ Gα13 LKO primary hepatocyte analysis showed a 2.583-fold increase in angiotensinogen
∙ Angiotensinogen-induced ER stress and ROS are primitive culprits in HTN
Liver, kidney, and heart Furmanik and Shanahan (2017), Dikalov and Dikalova (2019) and Kim et al. (2019)
Coagulation factors & proinflammatory mediators
Plasma Protease C1 inhibitors
(C1 INH)
∙ Deficiency of C1 INH causes congenital hereditary angioedema type I
∙ Excessive bradykinin production is a sign of genetic deficiency of C1 INH
∙ Primary hepatocytes fed up with HFD in Gα13 LKO showed elevated expression of C1 INH
∙ C1 INH levels were reduced in patients with COVID-19, T2D, and steatosis
Liver Clermont et al. (2011), Oschatz et al. (2011), Ivanov et al. (2019), Kim et al. (2019), Medjeral-Thomas et al. (2021), Karnaukhova (2022) and Subudhi et al. (2022)
Xanthine oxidase/dehydrogenase
XOR/D
∙ CDAHFD-fed NAFLD mice showed elevated XO plasma levels
∙ Primary hepatocytes treated with HFD showed a rise in XOR/D levels
∙ GLUT9 and SLC2A9 deficiency showed increased XOR/D activity
Liver and kidney DeBosch et al. (2014), Kim et al. (2019) and Kawachi et al. (2021)
Mannose Binding Protein C
(MBP-C)
∙ MBP-C aggravates the complement system and promotes tissue damage (thromboembolic system) in COVID-19 patients, and hepatic loss of Gα13 results in MBP-C enhanced expressions Liver Kim et al. (2019) and Asselta et al. (2022)
Structural proteins in association with ECM
Inter-α-trypsin inhibitor heavy chain 1
(ITIH1)
∙ HFD-fed to transgenic Alzheimer’s disease mice showed a 1.54-fold increase in ITIH1 levels
∙ The deficiency of ITIH1/ITIH3 altered the mice's behavior
∙ LPS primed HFD mice showed reduced ITIH1 levels
∙ ITIH1 levels were found higher in patients with T2D and NAFLD
Liver Goulding et al. (2019), Kim et al. (2019), Manuel et al. (2019) and Wang et al. (2021)
ITIH2 ∙ HFD-fed mice showed a positive relation with ITIH2 in Alzheimer’s disease
∙ ITIH2 was found reduced in patients with breast cancer
∙ Significantly diminished ITIH2 levels were observed in LPS primed HFD mice showed
∙ Gα13 loss results in enhanced expression of ITIH2
Liver Hamm et al. (2008), Kim et al. (2019), Manuel et al. (2019) and Wang et al. (2021)
ITIH4 ∙ COVID-19 patients showed decreased ITIH1, ITIH2, and ITIH4, whereas ITIH1 and 2 were overexpressed in COVID-19 patients
∙ ITIH4 levels decreased in patients with T2D and NAFLD
Liver Kim et al. (2019), Demichev et al. (2021a) and Geyer et al. (2021)
Proteins associated with acute phase response proteins and proinflammation
Fetuins A and B
Fetuins belong to the cystatin family and protease inhibitors and are considered acute phase response proteins (Brown and Dziegielewska 1997). Fetuin-A (α2-Heremans-Schmid glycoprotein, AHSG, Alpha-2-HS-glycoprotein) is efficiently expressed in serum, liver, tongue, and placenta (Denecke et al. 2003). Fetuin-B is majorly expressed in the liver, with serum concentrations of about 0.01 g/l and 0.3 g/l in humans and mice, respectively (Denecke et al. 2003). Fetuin levels were positively related in patients with elevated glucose contents, obesity (Peter et al. 2018), NAFLD, and T2D. Hence, fetuin was proposed as a potential biomarker for IR (Meex et al. 2015). Interestingly, enhanced fetuin-A levels were directly related to the loss of hepatic Gα13 and were related to chronic inflammation (Kim et al. 2019). Fetuin-A and B were also elevated in patients with liver steatosis. In contrast, fetuin A was found lowest in the patients with the final stage of alcoholic liver cirrhosis as compared to the initial stages (Prystupa et al. 2016). The circulating levels of fetuin-A were thus directly related to hepatic fat content, while both fetuin-A and B positively correlated with glucose area under the curve and oral glucose tolerance test results (Peter et al. 2018). In the same study and another one, fetuin-A stimulated IR in association with FFA through TLR4 (Pal et al. 2012; Peter et al. 2018). Fetuin-B is primarily released from the hepatocytes, and the levels were also elevated in T2D patients; consistently, HepG2 cells treated with fetuin-B showed enhanced lipid accumulation (Meex et al. 2015; Zhou et al. 2019). Mechanistically, fetuin-B decreases the phosphorylation of 5'-adenosine monophosphate-activated protein kinase (AMPK) (Zhou et al. 2019).
Serum amyloid A-2 protein
Serum amyloid A (SAA) belongs to four homologous alpha-helical amphipathic proteins encoded at chromosomes 7 and 11 in mice and humans, respectively. C-reactive proteins (CRP) and SAA subtypes are increased > 1000 times in the case of inflammation. CRP and SAA are primarily regulated by IL-1 and TNF-α in the presence of IL-6. SAA4 is mainly expressed in the liver. Apoe–/– mice fed with a high fat, high cholesterol diet (HFHCD) for 12 weeks showed elevated levels of SAA by inhibiting liver-specific ATP-citrate lyase. HFHCD feeding for 12 weeks resulted in chronic systemic inflammation and raised 1.3-fold plasma concentrations of SAA (Samsoondar et al. 2017). In another study, enhanced SAA levels were reported in obese humans and mice, and their respective livers showed elevated SAA2 mRNA levels (Chiba et al. 2009). Similar results were obtained using Apoe−/− mice (Chiba et al. 2009). The loss of hepatic Gα 13 in mice resulted in elevated ITIH1 and SAA2 levels. Both ITIH1 and SAA2 seem to be related to obesity, T2D, and NAFLD (Kim et al. 2019).
A number of studies showed a the strong association of SSA with the severity of COVID-19, emphasizing its prognostic value for COVID-19 (Goncalves and Sesterheim 2021; Zinellu et al. 2021). Two meta-analysis studies also validated its positive correlation in COVID-19 patients (Zhang et al. 2021; Zinellu et al. 2021). Mechanistically, one paper described the enhancement of amyloid formation of SAA in vitro in its nine-residue segment located at the C-terminus of the envelope protein of SARS-CoV-2 (Jana et al. 2021). It remains to be established whether the virus-induced upregulation of amyloid formation aggravates COVID-19.
Ceruloplasmin
Ceruloplasmin (Cp) belongs to glycoproteins (~ 150 kDa) and is primarily produced in the liver; Cp acts as the eighth binding atom of copper ions to the apo ceruloplasmin (Wolf and Griffiths 1982). Thus, Cp serves as a serum ferroxidase and transporter for copper (Wolf and Griffiths 1982). The amount of transferrin (FeIII) is a primitive sign to assess the iron concentrations in serum. The oxidation of FeII to FeIII by serum ferroxidase follows a zero-order reaction, and therefore a reduced Cp level lowers iron content in the blood (Roeser et al. 1970; Vachette et al. 2002). Studies have also shown that inherited Cp loss and low hepcidin serum levels lead to aceruloplasminemia (Kono 2012); Aceruloplasminemia is associated with diabetes and liver cancer because of intrusion in glucose metabolism due to hemochromatosis (iron toxicity) (Niederau et al. 1996; Cairo et al. 2001). The findings of another study showed that aceruloplasminemia is associated with increased iron storage in the liver and brain with low serum Cp levels (Loréal et al. 2002; Finkenstedt et al. 2010). A cohort study including 328 NAFLD patients showed that Cp gene mutation was related to hyper-ferritinemia, liver siderosis, and fibrosis (Corradini et al. 2021). Cp variants-associated hyperferritinemia and specific mutation of gene rs61733458 were reported in NAFLD patients (Pelucchi et al. 2021). The levels of Cp were lower in children with higher NAFLD scores, which may be due to the inability of the liver to produce Cp in the patients (Nobili et al. 2013). In another study, however, Cp levels were increased in the states of mild to severe steatosis (Liu et al. 2022). HFD-fed Gα13 LKO mice showed elevated Cp levels (i.e., a 2.2-fold change) (Kim et al. 2019). In addition, copper and Cp showed positive correlations in COVID-19 patients (Hackler et al. 2021). In particular, surviving patients exhibited much higher mean serum copper and CP levels compared to non-survivors (Hackler et al. 2021), suggestive of their potential use as prognostic markers for COVID-19 progression.
Alpha-1-acid glycoprotein-1
Alpha-1-acid glycoprotein (AGP), also known as orsomucoid (ORM), is a 44 kDa acute phase response protein and is the most abundantly occurring protein. AGP is mainly secreted by the hepatocytes, and the consequential human serum levels vary between 0.5 and 1.2 g/l (Ceciliani and Pocacqua 2007). It exists in two forms, AGP1 and AGP2. AGP has anti-inflammatory and immunomodulatory, anti-neutrophil, and anti-complementary effects in cases of inflammation, infection, and tissue grievance; however, the exact mechanism involved in this activity is still in debate (Ceciliani et al. 2002). Previously, it was believed that cytokines release is the major factor for the elevated expression of the AGP and its release from the hepatocytes. In the study, AGP levels were elevated after treatment with phenobarbital. Interestingly, endogenous secretion of IL-1 and IL-6 does not play a major role in the induction of AGP (Gauldie et al. 1987). After asialyation, AGP levels were increased in patients suffering from severe liver diseases (Serbource‐Goguel et al. 1983). More than a four-fold AGP increase was found in mild steatosis, whereas a sevenfold increase was observed in patients with severe steatosis (Liu et al. 2022). Of note, primary hepatocytes from HFD-fed Gα13 LKO mice showed enhanced expression of AGP1 (Kim et al. 2019), suggesting its association with the Gα13/12 signaling pathway. In another study, AGP1 positively correlated with the percent changes of collagen in the liver of NAFLD patients, whereas a negative relationship was found with apolipoprotein C-II in the fatty liver (Younossi et al. 2017).
Hemopexin
Hemopexin (HPX, 60 kDa) is the plasma glycoprotein with heme binding capability. HPX is majorly found in the liver and belongs to the acute phase responsive proteins. In the case of injury with inflammation, their levels are found to be significantly higher (Fiorito and Tolosano 2022). Analysis of 163 patients showed raised hemopexin articulation with lymph node ratio, venous invasion, and lymphatic invasion (Suzuki et al. 2020). Sickle mice (Hx-null) showed increased ROS and stimulation of Toll-like receptor 4 signaling mechanisms as well as cytokines, whereas the administration of HPX attenuated inflammatory and macrophage-activating pathways (Vinchi et al. 2016). The consequences of experiments using animal and patient samples exhibited reduced levels of hemopexin and decreased neutralized heme in patients with acute respiratory distress syndrome, burns, or premature infants; however, beneficial outcomes were observed after treatment with hemopexin (Lin et al. 2015).
The results of another study show that long-term subarachnoid hemorrhage is the primitive culprit for the cytotoxicity of heme and lower levels of HPX (Garland et al. 2016). HPX was elevated in patients with HCC, compared to those with either cirrhosis without HCC or fibrosis, and healthy volunteer groups (Debruyne et al. 2010). A mouse model with hemorrhagic shock was protected by treatment with either haptoglobin or hemopexin. Moreover, these treatments protect the kidney from injury associated with a high level of plasma hemoglobin (Graw et al. 2016).
Binding carrier of hormones or lipids
Retinol binding protein 4
Retinol-binding protein 4 (RBP4) is a polypeptide chain having a molecular weight of 21 kD and belongs to the lipocalin family. RBP4, majorly produced in the liver, acts as a serum carrier protein for vitamin A transport. Patients suffering from obesity or those with impaired glucose metabolism, IR (Graham et al. 2006; Haider et al. 2007), and T2D (Graham et al. 2006; Wu et al. 2008) showed elevated serum RBP4 levels, whereas this change was reversed with diet-associated weight loss, bariatric surgery, and exercise (Haider et al. 2007). However, in the case of NAFLD, some study results showed direct relation of RBP4 to NAFLD (Zhong et al. 2019), but others did an inverse relationship of RBP4 with NAFLD (Nobili et al. 2009; Wang et al. 2020). RBP4 leads to the activation of pro-inflammatory cytokines in mice and humans and interrupts insulin signaling by the stimulation of JNK and TLR4 molecular pathways (Norseen et al. 2012). Adipose-Glut4−/− mice showed a 2.5-fold increase in serum concentration of RBP4 compared to the control. The resultant increase of RBP4 activates hepatic expression of a gluconeogenic enzyme (i.e., phosphoenolpyruvate carboxykinase), disturbing muscle insulin signaling (Yang et al. 2005).
Precursors of receptor ligands or hormones
Angiopoietin-like proteins
Angiopoietin-like proteins (ANGPTLs), highly hydrophobic paracrine factors, are ramified into eight members. Of them, ANGPTL 1–7 share structural resemblance and serve as ligands for Tie receptors (TieI or Tie1) (Oike et al. 2003). ANGPTLs are majorly expressed in various organs, such as the liver, kidneys, vascular system, and hematopoietic system, and are involved in the regulation of angiogenesis, inflammation, and lipid metabolism (Tabata et al. 2009; Chen et al. 2016). Studies have proved that ANGPTL 1 and 2 were related to hepatocellular carcinomas (Chen et al. 2016; Carbone et al. 2018). In addition, ANGPTLs are found to have a regulatory impact on lipid metabolism and angiogenesis, being considered therapeutic candidates for metabolic syndrome (Li and Teng 2014). Serum ANGPTL4 levels were raised in obese patients with or without T2D; correspondingly, the levels of ANGPTL3 were decreased respectively (Cinkajzlová et al. 2018). Both ANGPTL 3 and 4 are highly expressed in the liver, whereas ANGPTL4 hepatic expression is 10% of the adipose tissue (Koishi et al. 2002; Romeo et al. 2009). ANGPTL4 is also known as hepatic fibrinogen/angiopoietin-related protein. Studies have proven that the deletion of ANGTL4 in mice and its mutational loss in patients leads to decreased triglycerides and elevated high-density lipoproteins levels via inhibition of lipoprotein lipase activity, protecting patients against obesity, T2D, NAFLD, and steatosis (Romeo et al. 2007, 2009; Singh et al. 2021; Spitler et al. 2021). Physiologically ANGPTL4 expression is raised because of fasting, cold, exercise, and fatty acid-activated peroxisome proliferator-activated receptors (Lichtenstein et al. 2010). In another study, ANGPTL4 significantly diminished foam cell formation, inflammatory gene expression, and ER stress (Lichtenstein et al. 2010). A study focused on the effect of ANGPTL6 (also known as an angiopoietin-related growth factor) against obesity and IR reveals that ANGPTL6−/− mice showed a significant increase in the body weight leading to obesity on a normal chow diet. Interestingly, loss of ANGPTL6 raised the rectal temperature, basal metabolic rate, and food intake, consequently raising serum cholesterol and TAG levels. Furthermore, mice have developed IR with elevated glucose levels (Oike et al. 2005). ANGPTL6 serum levels were found to be raised in T2D and NAFLD patients (Oike et al. 2005; Ebert et al. 2009; Ma et al. 2019). ANGPTL8/betatrophin is commonly named as HCC-associated protein, TD26, or lipasin, and is found majorly in the liver and visceral adipose tissue. Elevated ANGPTL8 levels were reported in human liver steatosis and enhanced TAG levels in plasma (von Loeffelholz et al. 2017; Wang et al. 2018). TD26 mechanistically binds with the nuclear form of SREBP1, leading to elevated lipid production and tumor cell proliferation (Wang et al. 2018). ANGPTL8, in combination with ANGPTL3, acts as an inhibitor of lipoprotein lipase (Kovrov et al. 2019).
Fibroblast growth factors (FGF families)
Armelin (1973) and Gospodarowicz (1975) were the scientists who introduced the world to fibroblast growth factors (FGFs). Up till now, four members of the family have been discovered, which undergo alternative splicing and yield seven functionally distinct receptors (i.e., FGFRs 1b, 1c, 2b, 2c, 3b, 3c, and 4) with distinct ligand binding properties. The FGF family has been found to regulate energy metabolism (Ornitz and Itoh 2022). FGFR subfamilies are responsible for the release of 18 FGFs that are capable of interacting with the tyrosine kinase with the help of various cofactors (Schumacher and Guo 2016). Canonical FGFs are also paracrine FGFs which mainly exert their functions by binding with heparin. FGF19 and its subunits, i.e., FGF19, FGF21, and FGF23, interact with α-klotho, resulting in its endocrine function (Goetz et al. 2007; Ornitz and Itoh 2015; Yanucil et al. 2022). These three members are involved in endocrine functions and thus regulate bile acid, carbohydrate, lipid metabolism, cell proliferation, differentiation, and survival (Ornitz and Itoh 2015).
The study conducted using bile duct ligation (BDL) in wild-type, and Mdr−/− mice followed by treatment with FGFR1 antagonist (AZD4547) leads to reduced FGF1 and miR-16, resulting in a protective effect against BDL-induced hepatic fibrosis, biliary proliferation, and inflammation (O’Brien et al. 2022). FGF2 is of two types; one is a low molecular weight FGF2, whereas the other is a high molecular weight form. Administration of low molecular weight FGF2 in CCl4-induced fibrotic mice led to the downregulation of Delta-like 1 via the p38 mitogen-activated protein kinase pathway, showing ameliorative effects against fibrosis (Pan et al. 2015). In another study, FGF2 treatment of mice with BDL triggers cytoglobin activation inhibits myofibroblastic human HSCs and ameliorates liver fibrosis (Sato-Matsubara et al. 2017). FGF via FGF receptor 4 (FGFr4) modulates signal transduction between Wnt16 and Dlc, activating notch signaling and leading to hepatocellular carcinoma by initiating niche formation (Lee et al. 2014). FGF19, FGF21, and FGF23, the members of FGF19, are highly expressed in the liver. These ligands are found to regulate the bile acid (BA), fatty acid, glucose, and phosphate metabolism via binding with βKlotho homologous single-pass transmembrane proteins and stimulate FGFr4 (Kurosu et al. 2007). Serum FGF19 and bile acid concentrations were found to be raised in NASH subjects, although adiponectin levels were significantly lowered (Bechmann et al. 2013). It has also been shown that decreased FGF19 and BA concentrations were associated with impaired FXR and FGFr4 signaling (Jiao et al. 2018) (Table 1). Recently, a randomized control study was performed in NAFLD patients with a score of 4 or higher, stage 1–3 fibrosis, and at least 8% liver fat content and the patients were treated with the analog of FGF19 at the dose of 3 or 6 mg for twelve weeks, which led to decreased hepatic fat contents (Harrison et al. 2018). ER stress caused elevated levels of βKlotho and FGF19 in the sera of patients with HCC (Miura et al. 2012). FGF21 and its analogs activated liver FA oxidation, significantly reducing fat buildup in the liver and improving IR. In addition, the treatments resulted in enhanced levels of adiponectin and exhibited insulin-sensitizing, anti-fibrotic, anti-inflammatory, and anti-steatosis effects (Zarei et al. 2020). CYP2E1 level may be important in FGF21 expression; a deficiency of CYP2E1 is necessary for the activation of the PPARα-FGF21 axis and is effective in the reduction of obesity (Zhang et al. 2022).
Pigment epithelium-derived factor
Pigment epithelium-derived factor (PEDF) is an endogenous glycoprotein belonging to the serine protease inhibitor family, released by the adipocytes, retinal epithelial pigment, hepatocytes, and skeletal myocytes. It also contains an extracellular matrix binding protein site (Uehara et al. 2004; Fitzgerald et al. 2012). PEDF is a regulator of angiogenesis inhibition, immunomodulation, and neurotrophic and has antioxidant activity, antivasopermeability, and anti-tumor activity (Kawaguchi et al. 2010). Studies using human primary melanocytes, as well as an in vivo model, showed higher levels of PEDF and microphthalmia-associated transcription factor expression (Fernández-Barral et al. 2014). PEDF increased the proliferation of breast cancer cells embedded in the mouse brain (Fitzgerald et al. 2012). PEDF was negatively associated with hepatic metastasis in patients with stage II (Uehara et al. 2004) (Table 1). Overexpression of PEDF stimulates NaAsO2-induced apoptosis with an increase of p53 (Zhang et al. 2019). PEDF knockout mice showed elevated expression of the genes related to the Wnt/βcatenin pathway (Protiva et al. 2015). Moreover, the levels of PEDF were significantly enhanced in T2D patients with chronic kidney diseases (Hui et al. 2014).
Hepassocin
Hepassocin (HPS) is known as a hepatocyte-derived fibrinogen-related protein (HFREP-1) and is a liver-specific gene involved in hepatic regeneration (Ou et al. 2017). HPS/liver fibrinogen-related gene-1 expressions were reduced in HCC patients and mice treated with streptozotocin (Ou et al. 2017). The results of another study show that HNF1a interacts with IL-6/IL-6R/STAT3 trajectory and upregulates the HPS promoter transcriptional factors, resulting in the upholding of homeostases such as growth and repair, like the IGFBP-1, G6Pase, and a-fibrinogen promoters (Yu et al. 2009). The study design included patients with T2D (Group I), NAFLD (Group II), and both (Group III). The serum analysis revealed a significantly higher concentration of HPS in group III patients compared to groups I and II (Abdelmoemen et al. 2019). Another study, including 199 patients with NAFLD, showed higher levels of HPS compared to the control, and similar results were also observed in mice fed with HFD. However, the HPS knock-down mice produced by using short hairpin RNAs targeting HPS showed recovery from the steatosis with decreased NAFLD activity score. Mechanistically, overexpression of HPS in HepG2 cells leads to fat accumulation by modulating the extracellular signal-regulated kinase 1/2 (ERK1/2)-dependent pathway (Wu et al. 2013). Palmitate-induced ER stress in primary hepatocytes showed dose-dependent increase of HPS via stimulation of C/EBPβ-mediated transcriptional factor (Jung et al. 2018; Watt et al. 2019).
Angiotensinogen
Hypertension is one of the most prevalent contributors to worldwide disease and socioeconomic burden with poor understanding (Brouwers et al. 2021). Various organs (i.e., kidneys, heart, liver, vessels, and immune cells) were found to be involved in the development and aggravation of hypertension under several mechanisms, such as oxidative stress and inflammation, obesity, and diabetes (Hossain et al. 2007).
ER stress is the primitive culprit in cardiovascular diseases because of its regulatory characteristics in vascular cell phenotype, dedifferentiation, calcification, and apoptotic mechanisms leading to hypertension and atherosclerosis (Furmanik and Shanahan 2017). ER stress leads to cardiovascular dysfunction and tissue damage (Furmanik and Shanahan 2017). The result of the study validates that reduction/inhibition of ER stress results in the amelioration of hypertension by protecting vascular dysfunction (Carlisle et al. 2016), cardiac impairment, and pulmonary hypertension (Spitler et al. 2013). SIRT 3 acts as a regulator of enzymatic antioxidant activity in mitochondria. The results of studies show that sirtuin 3 (SIRT 3) expression gets reduced in older age (i.e., > 65 years) with resultant hypertension and that metabolic diseases are also responsible for the declined/inactivity due to elevated NADH and acetyl-CoA intensities (Dikalova et al. 2017). SIRT 3 knockout mice experienced hypertension. Diminished SIRT 3 expressions because of hyperacetylation consequently result in oxidative stress in mitochondria (Dikalova et al. 2017). Recent studies specifically focused on mitochondrial interference for hypertension management (Miller Jr 2020). Administration of MitoQ10, a mitochondrial-specific antioxidant, controlled elevated blood pressure in rats by improving endothelial function and decreasing hypertrophy (Graham et al. 2009). In another study, hepatic loss of Gα13 leads to increased expressions of angiotensinogen (2.583-folds) (Kim et al. 2019).
Coagulation factors and proinflammatory mediators
Plasma protease C1 inhibitors (C1 INH)
C1 INH is an acute-phase protein known as a protease inhibitor primarily expressed in the liver. It belongs to the serine protease inhibitor family with a molecular weight of 105 kDa. Its physiological functions in plasma include the superintendence of various proteolytic functions such as the complement, coagulation, and fibrinolytic pathways (Davis III 1988). Pathogenesis of various diseases shares a common feature of stimulating the complement system in plasma, and the resultant proinflammatory effects produce vasoactive peptides such as C3a, C5a (anaphylaxis), and bradykinin. The deficiency of C1 INH causes a major hereditary disease named hereditary angioedema (Ivanov et al. 2019; Karnaukhova, 2022). The anticipated ratio of hereditary angioedema pervasiveness is 1 out of 50,000 persons without any distinguishable ethnic differences, while untreated patients experience severe attacks (Zuraw 2008).
Mechanistically, when FXII triggers and brings conformational changes during its interaction with negatively associated surfaces, leading to the production of activated form FXII (FXIIa). The resultant product (FXIIa) converts plasma prekallikrein to plasma kallikrein, which stimulates FXII, causing a positive feedback loop of FXII activation and producing bradykinin via breaking down the high molecular weight kininogen (Müller and Renné 2008). Bradykinin further initiates inflammatory pathways, which leads to enhanced vascular permeability, vasodilation, and chemotaxis of neutrophils. Moreover, FXIIa also activates coagulation pathways by converting FXI into FXIa, which activates the Ca+2-associated proteolytic cleavage and results in thrombin production, fibrin, and fibrin clot formation (Leeb-Lundberg et al. 2005).
C1 INH is a primitive inhibitor of various complement proteases such as C1r, C1s, and mannose-binding lectin–associated serine protease (MASP1 and 2), plasma kallikrein, and coagulation factor XIa and XIIa (Zuraw 2008). Studies also strengthen the concept that excessive bradykinin production is directly related to the genetic deficiency of C1 INH (Oschatz et al. 2011). C1 INH-deficient patients are classified as HAE type I, whereas patients with dysfunctional C1 INH protein with normal C1 INH plasma antigen levels are considered HAE type II (Bl 2008). C1 INH gains much attention in COVID-19 treatment because systemic complement stimulation and local complement-triggering effect lead to a severe acute respiratory syndrome, resulting in hyperinflammation and thrombosis (Afzali et al. 2022, van de Veerdonk et al. 2022). C1 INH and α2 M levels were reduced in patients with COVID 19, where as the levels of ITIH4 were raised (Medjeral-Thomas et al. 2021).
Recently, a significant reduction in levels of C1 INH in COVID-19 patients was reported in several studies (Demichev et al. 2021b). Hence, C1 INH gains much attention in COVID-19 treatment due to its potential inhibitory activity against SARS-CoV-2-induced hyperinflammation and hypercoagulation (Afzali et al. 2022, van de Veerdonk et al. 2022). Based on these, its therapeutic applicability to alleviate the severity of COVID-19 pathogenesis, as mediated by the reversal of SARS-CoV-2-induced proinflammatory and prothrombic events, was proposed by researchers (Thomson et al. 2020; Adesanya et al. 2021). Recent ongoing clinical trials will ultimately evaluate the use of C1 INH treatment for COVID-19 patients (Mansour et al. 2021).
Abnormal activation of the complement system leads to increased liver damage, resulting in aggravation of hepatic steatosis and ischemic reperfusion hepatic injury with fatty liver (Wlazlo et al. 2013). The results of NAFLD patients with obesity show that C1 INH levels were found to be decreased in patients with steatosis and NASH (Subudhi et al. 2022). Interestingly, the results of a recent study reported that the hepatic loss of Gα13 showed 3.2-fold increased secretion of C1 INH in HFD-fed primary hepatocytes (Kim et al. 2019). Intravitreal injection of C1 INH in streptozotocin-induced diabetic rats significantly reduced the retinal vascular permeability by inhibiting the plasma kallikrein (Clermont et al. 2011). The exact role of C1 INH in NAFLD and diabetes is still in debate and is an undiscovered area.
Xanthine oxidase/dehydrogenase
Xanthine oxidase and dehydrogenase are interchangeable forms of the same enzyme, xanthine oxidoreductase (Pacher et al. 2006). XO mainly regulates the deprivation of purines into uric acid, during which it produces two moles of superoxide and one mole of H2O2 (Furuhashi 2020). SLC2A9 gene in GLUT-9 knockout mice experienced elevated urea levels, blood pressure, dyslipidemia, and whole body fat, whereas the administration of xanthine oxidase inhibitor results in recovery of all the above-mentioned parameters (DeBosch et al. 2014). Another study showed that elevated levels of xanthine oxidase were responsible for IR and high sensitivity-C-reactive protein levels in adults (Washio et al. 2017). Increased plasma XO was reported in CDAHFD-induced mice with NASH and was involved in vascular injury in NAFLD/NASH mice (Kawachi et al. 2021). XO expression was raised and reported in primary hepatocytes of Gα13 LKO mice (Kim et al. 2019).
Mannose-binding protein C
Mannose-binding protein (MBP), also known as mannose-binding lectin, can interact with high-density sugar existing on the surface of bacteria, fungi, and parasites, stimulating the antibody-independent complement system (Ikeda et al. 1987). MBP is produced in the liver and is of two types MBP-A and MBP-C (Drickamer et al. 1986). Complement fixation activated by the serine protease, such as MBP-associated serine protease-2 (MASP-2), cleaves and downregulates the complement system (Thiel et al. 1997). The study results show that the MBL-activated complement system aggravates tissue damage, such as the thromboembolic system in COVID-19 patients (Asselta et al. 2022), showing that the MBL-activated complement system aggravates tissue damage in COVID-19 patients (Asselta et al. 2022).
Structural proteins in association with ECM
Inter-α-trypsin inhibitor heavy chain
Inter-α-inhibitors (IαI), commonly known as inter-α-trypsin inhibitors, are protein-glycosaminoglycan-protein complexes, acute phase response proteins, and have relative concentrations ranging from 150 to 500 µg/l in human plasma (Zhuo et al. 2004). IαI consist of light chains, bikunin, and heavy chains (H1-H5) combined with the chondroitin sulfate chain (Saguchi et al. 1996; Hamm et al. 2008). Two heavy chains (H1 and H2) are the predominant bikunin form, and a single bikunin and pre-α-inhibitor (H3) are present in the blood. The anti-proteolytical activity of IαI belong to the bikunin (Saguchi et al. 1996; Zhuo et al. 2004). HFD fed to triple transgenic Alzheimer's disease showed a 1.54-fold increase in the ITIH1 level and also elevated ITIH2 positively correlated with the disease progression (Wang et al. 2021). Reduced expression (70%) of ITIH2 was reported in breast cancer patients (Hamm et al. 2008). The results showed that the genetic loss of Ambp/bikunin, required for the activation of ITIH1, ITIH3, and ITIH4, leads to the mood alternation in the mice by the loss of ITIH1/ITIH3, whereas ITIH4 deficiency had no effect on the moods and behavior disorder in mice (Goulding et al. 2019). The cluster analysis of uterine LPS-primed mice fed on HFD showed reduced expressions of ITIH1 and ITIH2 (Manuel et al. 2019).
The hepatic Gα13 knockout mice showed a 2.7-fold increase in ITIH1, and ITIH1 was overexpressed in mice with diabetes and NAFLD with hyperglycemia. However, ITIH4 levels were reduced (Kim et al. 2019). The results of GWAS analysis found that ITIH3/ITIH1 genes were directly related to brain health; higher levels would cause more damage to the brain tissue (Gadd et al. 2022). Proprotein convertase subtilisin/kexin type 9 (PCSK9) has shown interaction with high-density lipoproteins in coronary artery disease, and interestingly, crosslinking mass spectrometry analysis identifies the interaction of PCSK9 with ITIH1 (hyaluronan binder) and apoA1 in immunoHDL (Burnap et al. 2021).
A recent study shows that ITIH4, after cleavage, forms a noncovalent inhibitory complex and acts as a protease that is dependent on the ITIH4 von Willebrand factor A domain. ITIH4 impedes mannan-binding lectin–associated serine protease (MASP) 1 and 2 and kallikrein. ITIH4 and MASP complex downregulate the breaking of C2 and C4 by inhibiting the contact of scissile bonds to the active binding site, leading to acting as protease inhibitors (Pihl et al. 2021). Astonishingly, results were observed in proteomics analysis of COVID-19 patients in which expressions of acute phase proteins ITIH1, ITIH3, and ITIH4 were downregulated (Demichev et al. 2021a); however, another study showed overexpressed ITIH1 and ITIH2, whereas ITIH4 was downregulated in COVID-19 patients (Geyer et al. 2021).
Recently, proteomics analysis of COVID-19 patients’ sera confirmed the association of poor prognosis with low levels of ITIH2 (Demichev et al. 2021a). Interestingly, the same family member proteins, ITIH3 and ITIH4, were found to be reduced in older patients with COVID-19 (Demichev et al. 2021b). In separate survival analyses, levels of ITIH4 increased in patients who failed to survive. This data agrees with a lower abundance of ITIH3 and ITIH4 in the non-survivors and a higher abundance of ITIH1 and ITIH2 in the survivor group (Völlmy et al. 2021). Since ITIH4 was shown to act as a protease inhibitor for mannan-binding lectin-associated serine protease-1 (MASP-1), MASP-2, and plasma kallikrein, ITIH4 may be utilized as a therapeutic target to prevent SARS-CoV-2-induced hyperinflammation, which depends on the complement and kinin-kallikrein pathways.
Summary
The consumption of excess fat and the resultant accompanying lipotoxicity, autophagy dysregulation, ER stress, and insulin resistance may cause disturbances in the secretion and modifications of the proteins and their interactions with other proteins and/or structures. This ultimately leads to cell death mechanisms. This article attempted to provide an updated overview of liver secretome biology with explanatory mechanisms with regard to metabolic liver diseases so that it may be of help in treating patients and overcoming the economic burden.
Acknowledgements
This work was supported by grants from the National Research Foundation of Korea (NRF), funded by the Korean government (MSIT) (2017K1A1A2004511).
Declarations
Conflict of interest
Authors have no conflict of interest.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
Abdelmoemen G Khodeir SA Zaki AN Kassab M Abou-Saif S Abd-Elsalam S Overexpression of hepassocin in diabetic patients with nonalcoholic fatty liver disease may facilitate increased hepatic lipid accumulation Endocr Metab Immune Disord Drug Targets 2019 19 185 188 10.2174/1871530318666180716100543 30009716
Adesanya TMA Campbell CM Cheng L Ogbogu PU Kahwash R C1 Esterase inhibition: targeting multiple systems in COVID-19 J Clin Immunol 2021 41 729 732 10.1007/s10875-021-00972-1 33474624
Afzali B Noris M Lambrecht BN Kemper C The state of complement in COVID-19 Nat Rev Immunol 2022 22 77 84 10.1038/s41577-021-00665-1 34912108
Alexander SP Christopoulos A Davenport AP Kelly E Mathie A Peters JA Veale EL Armstrong JF Faccenda E Harding SD The concise guide to phamacology 2021/22: G protein-coupled receptors Br J Pharmacol 2021 178 S27 156 10.1111/bph.15538 34529832
Amir M Czaja MJ Autophagy in nonalcoholic steatohepatitis Expert Rev Gastroenterol Hepatol 2011 5 159 166 10.1586/egh.11.4 21476911
Anstee QM Targher G Day CP Progression of NAFLD to diabetes mellitus, cardiovascular disease or cirrhosis Nat Rev Gastroenterol Hepatol 2013 10 330 344 10.1038/nrgastro.2013.41 23507799
Armelin HA Pituitary extracts and steroid hormones in the control of 3T3 cell growth Proc Natl Acad Sci USA 1973 70 9 2702 2706 10.1073/pnas.70.9.2702 4354860
Asselta R Paraboschi EM Stravalaci M Invernizzi P Bonfanti P Biondi A Pagani I Pedotti M Doni A Scavello F Reply to: Genetic determinants of mannose-binding lectin activity predispose to thromboembolic complications in critical COVID-19. Mannose-binding lectin genetics in COVID-19 Nat Immunol 2022 23 865 867 10.1038/s41590-022-01228-9 35624207
Batchvarova N Wang X Ron D Inhibition of adipogenesis by the stress-induced protein CHOP (Gadd153) EMBO J 1995 14 4654 4661 10.1002/j.1460-2075.1995.tb00147.x 7588595
Bechmann LP Kocabayoglu P Sowa JP Sydor S Best J Schlattjan M Beilfuss A Schmitt J Hannivoort RA Kilicarslan A Free fatty acids repress small heterodimer partner (SHP) activation and adiponectin counteracts bile acid-induced liver injury in superobese patients with nonalcoholic steatohepatitis Hepatology 2013 57 1394 1406 10.1002/hep.26225 23299969
Bl Z Clinical practice. Hereditary angioedema N Engl J Med 2008 359 1027 1036 10.1056/NEJMcp0803977 18768946
Brouwers S Sudano I Kokubo Y Sulaica EM Arterial hypertension Lancet 2021 398 249 261 10.1016/S0140-6736(21)00221-X 34019821
Brown W Dziegielewska K Friends and relations of the cystatin superfamily: new members and their evolution Protein Sci 1997 6 5 12 10.1002/pro.5560060102 9007972
Burnap SA Sattler K Pechlaner R Duregotti E Lu R Theofilatos K Takov K Heusch G Tsimikas S Fernández-Hernando C PCSK9 activity is potentiated through HDL binding Circ Res 2021 129 1039 1053 10.1161/CIRCRESAHA.121.319272 34601896
Cairo G Conte D Bianchi L Fraquelli M Recalcati S Reduced serum ceruloplasmin levels in hereditary haemochromatosis Br J Haematol 2001 114 226 229 10.1046/j.1365-2141.2001.02917.x 11472372
Calfon M Zeng H Urano F Till JH Hubbard SR Harding HP Clark SG Ron D IRE1 couples endoplasmic reticulum load to secretory capacity by processing the XBP-1 mRNA Nature 2002 415 92 96 10.1038/415092a 11780124
Carbone C Piro G Merz V Simionato F Santoro R Zecchetto C Tortora G Melisi D Angiopoietin-like proteins in angiogenesis, inflammation and cancer Int J Mol Sci 2018 19 431 10.3390/ijms19020431 29389861
Carlisle RE Werner KE Yum V Lu C Tat V Memon M No Y Ask K Dickhout JG Endoplasmic reticulum stress inhibition reduces hypertension through the preservation of resistance blood vessel structure and function J Hypertens 2016 34 1556 1569 10.1097/HJH.0000000000000943 27115336
Carr RM Oranu A Khungar V Nonalcoholic fatty liver disease: pathophysiology and management Gastroenterol Clin North Am 2016 45 639 652 10.1016/j.gtc.2016.07.003 27837778
Ceciliani F Pocacqua V The acute phase protein α1-acid glycoprotein: a model for altered glycosylation during diseases Curr Protein Pept Sci 2007 8 91 108 10.2174/138920307779941497 17305563
Ceciliani F Giordano A Spagnolo V The systemic reaction during inflammation: the acute-phase proteins Protein Pept Lett 2002 9 211 223 10.2174/0929866023408779 12144517
Chang T-Y Lan K-C Chiu C-Y Sheu M-L Liu S-H ANGPTL1 attenuates cancer migration, invasion, and stemness through regulating FOXO3a-mediated SOX2 expression in colorectal cancer Clin Sci 2022 136 657 673 10.1042/CS20220043
Chen HA Kuo TC Tseng CF Ma JT Yang ST Yen CJ Yang CY Sung SY Su JL Angiopoietin-like protein 1 antagonizes MET receptor activity to repress sorafenib resistance and cancer stemness in hepatocellular carcinoma Hepatology 2016 64 1637 1651 10.1002/hep.28773 27530187
Chiba T Han CY Vaisar T Shimokado K Kargi A Chen M-H Wang S Mcdonald TO O’brien KD Heinecke JW Serum amyloid A3 does not contribute to circulating SAA levels J Lipid Res 2009 50 1353 1362 10.1194/jlr.M900089-JLR200 19286646
Choi W-M Kim H-H Kim M-H Cinar R Yi H-S Eun HS Kim S-H Choi YJ Lee Y-S Kim SY Glutamate/metabotropic glutamate receptor-5 signaling in hepatic stellate cells drives endocannabinoid-mediated alcoholic steatosis Cell Metab 2019 30 877 10.1016/j.cmet.2019.08.001 31474565
Choi WM Ryu T Lee JH Shim YR Kim MH Kim HH Kim YE Yang K Kim K Choi SE Metabotropic glutamate receptor 5 in natural killer cells attenuates liver fibrosis by exerting cytotoxicity to activated stellate cells Hepatology 2021 74 2170 2185 10.1002/hep.31875 33932306
Chung KW Chung HY The effects of calorie restriction on autophagy: role on aging intervention Nutrients 2019 11 2923 10.3390/nu11122923 31810345
Cinkajzlová A Mráz M Lacinová Z Kloučková J Kaválková P Kratochvílová H Trachta P Křížová J Haluzíková D Škrha J Angiopoietin-like protein 3 and 4 in obesity, type 2 diabetes mellitus, and malnutrition: the effect of weight reduction and realimentation Nutr Diabetes 2018 8 1 11 10.1038/s41387-018-0032-2 29330446
Clermont A Chilcote TJ Kita T Liu J Riva P Sinha S Feener EP Plasma kallikrein mediates retinal vascular dysfunction and induces retinal thickening in diabetic rats Diabetes 2011 60 1590 1598 10.2337/db10-1260 21444925
Cornier M-A Dabelea D Hernandez TL Lindstrom RC Steig AJ Stob NR Van Pelt RE Wang H Eckel RH The metabolic syndrome Endocr Rev 2008 29 777 822 10.1210/er.2008-0024 18971485
Corradini E Buzzetti E Dongiovanni P Scarlini S Caleffi A Pelusi S Bernardis I Ventura P Rametta R Tenedini E Ceruloplasmin gene variants are associated with hyperferritinemia and increased liver iron in patients with NAFLD J Hepatol 2021 75 506 513 10.1016/j.jhep.2021.03.014 33774058
Crandall JP Shamoon H Goldman L Schafer AI Diabetes mellitus Goldman-cecil medicine 2020 26 Amsterdam Elsevier 1490 1510
Czaja MJ Functions of autophagy in hepatic and pancreatic physiology and disease Gastroenterology 2011 140 1895 1908 10.1053/j.gastro.2011.04.038 21530520
Davis Iii AE C1 inhibitor and hereditary angioneurotic edema Annu Rev Immunol 1988 6 595 628 10.1146/annurev.iy.06.040188.003115 3289579
Debosch BJ Kluth O Fujiwara H Schürmann A Moley K Early-onset metabolic syndrome in mice lacking the intestinal uric acid transporter SLC2A9 Nat Commun 2014 5 1 7 10.1038/ncomms5642
Debruyne EN Vanderschaeghe D Van Vlierberghe H Vanhecke A Callewaert N Delanghe JR Diagnostic value of the hemopexin N-glycan profile in hepatocellular carcinoma patients Clin Chem 2010 56 823 831 10.1373/clinchem.2009.139295 20348404
Demichev V Tober-Lau P Lemke O Nazarenko T Thibeault C Whitwell H Röhl A Freiwald A Szyrwiel L Ludwig D A time-resolved proteomic and prognostic map of COVID-19 Cell Syst 2021 12 780–794 e7 10.1016/j.cels.2021.05.005
Demichev V Tober-Lau P Lemke O Nazarenko T Thibeault C Whitwell H Rohl A Freiwald A Szyrwiel L Ludwig D Correia-Melo C Aulakh SK Helbig ET Stubbemann P Lippert LJ Gruning NM Blyuss O Vernardis S White M Messner CB Joannidis M Sonnweber T Klein SJ Pizzini A Wohlfarter Y Sahanic S Hilbe R Schaefer B Wagner S Mittermaier M Machleidt F Garcia C Ruwwe-Glosenkamp C Lingscheid T Bosquillon De Jarcy L Stegemann MS Pfeiffer M Jurgens L Denker S Zickler D Enghard P Zelezniak A Campbell A Hayward C Porteous DJ Marioni RE Uhrig A Muller-Redetzky H Zoller H Loffler-Ragg J Keller MA Tancevski I Timms JF Zaikin A Hippenstiel S Ramharter M Witzenrath M Suttorp N Lilley K Mulleder M Sander LE P-C-S GroupRalser M Kurth F A time-resolved proteomic and prognostic map of COVID-19 Cell Syst 2021 12 780–794 e7 10.1016/j.cels.2021.05.005
Denecke B Gräber S Schäfer C Heiss A Wöltje M Jahnen-Dechent W Tissue distribution and activity testing suggest a similar but not identical function of fetuin-B and fetuin-A Biochem J 2003 376 135 145 10.1042/bj20030676 12943536
Deng Y Srivastava R Howell SH Protein kinase and ribonuclease domains of IRE1 confer stress tolerance, vegetative growth, and reproductive development in Arabidopsis Proc Natl Acad Sci USA 2013 110 19633 19638 10.1073/pnas.1314749110 24145452
Diehl AM Day C Cause, pathogenesis, and treatment of nonalcoholic steatohepatitis N Engl J Med 2017 377 2063 2072 10.1056/NEJMra1503519 29166236
Dikalov SI Dikalova AE Crosstalk between mitochondrial hyperacetylation and oxidative stress in vascular dysfunction and hypertension Antioxid Redox Signal 2019 31 710 721 10.1089/ars.2018.7632 30618267
Dikalova AE Itani HA Nazarewicz RR Mcmaster WG Flynn CR Uzhachenko R Fessel JP Gamboa JL Harrison DG Dikalov SI Sirt3 impairment and SOD2 hyperacetylation in vascular oxidative stress and hypertension Circ Res 2017 121 564 574 10.1161/CIRCRESAHA.117.310933 28684630
Drickamer K Dordal MS Reynolds L Mannose-binding proteins isolated from rat liver contain carbohydrate-recognition domains linked to collagenous tails. Complete primary structures and homology with pulmonary surfactant apoprotein J Biol Chem 1986 261 6878 6887 10.1016/S0021-9258(19)62698-0 3009480
Ebert T Bachmann A Lössner U Kratzsch J Blüher M Stumvoll M Fasshauer M Serum levels of angiopoietin-related growth factor in diabetes mellitus and chronic hemodialysis Metabolism 2009 58 547 551 10.1016/j.metabol.2008.11.016 19303977
Eckel RH Alberti KG Grundy SM Zimmet PZ The metabolic syndrome Lancet 2010 375 181 183 10.1016/S0140-6736(09)61794-3 20109902
Eslam M Newsome PN Sarin SK Anstee QM Targher G Romero-Gomez M Zelber-Sagi S Wai-Sun Wong V Dufour JF Schattenberg JM Kawaguchi T Arrese M Valenti L Shiha G Tiribelli C Yki-Jarvinen H Fan JG Gronbaek H Yilmaz Y Cortez-Pinto H Oliveira CP Bedossa P Adams LA Zheng MH Fouad Y Chan WK Mendez-Sanchez N Ahn SH Castera L Bugianesi E Ratziu V George J A new definition for metabolic dysfunction-associated fatty liver disease: an international expert consensus statement J Hepatol 2020 73 202 209 10.1016/j.jhep.2020.03.039 32278004
Estes C Razavi H Loomba R Younossi Z Sanyal AJ Modeling the epidemic of nonalcoholic fatty liver disease demonstrates an exponential increase in burden of disease Hepatology 2018 67 123 133 10.1002/hep.29466 28802062
Fan J-G Kim S-U Wong VW-S New trends on obesity and NAFLD in Asia J Hepatol 2017 67 862 873 10.1016/j.jhep.2017.06.003 28642059
Fernández-Barral A Orgaz JL Baquero P Ali Z Moreno A Tiana M Gómez V Riveiro-Falkenbach E Cañadas C Zazo S Regulatory and functional connection of microphthalmia-associated transcription factor and anti-metastatic pigment epithelium derived factor in melanoma Neoplasia 2014 16 529 542 10.1016/j.neo.2014.06.001 25030625
Finkenstedt A Wolf E Höfner E Gasser BI Bösch S Bakry R Creus M Kremser C Schocke M Theurl M Hepatic but not brain iron is rapidly chelated by deferasirox in aceruloplasminemia due to a novel gene mutation J Hepatol 2010 53 1101 1107 10.1016/j.jhep.2010.04.039 20801540
Fiorito V Tolosano E Hemopexin and cancer Int J Mol Sci 2022 23 997 10.3390/ijms23020997 35055182
Fitzgerald DP Subramanian P Deshpande M Graves C Gordon I Qian Y Snitkovsky Y Liewehr DJ Steinberg SM Paltán-Ortiz JD Opposing effects of pigment epithelium-derived factor on breast cancer cell versus neuronal survival: implication for brain metastasis and metastasis-induced brain damage Cancer Res 2012 72 144 153 10.1158/0008-5472.CAN-11-1904 22215693
Friedman SL Neuschwander-Tetri BA Rinella M Sanyal AJ Mechanisms of NAFLD development and therapeutic strategies Nat Med 2018 24 908 922 10.1038/s41591-018-0104-9 29967350
Fukumoto S-i Yamauchi N Moriguchi H Hippo Y Watanabe A Shibahara J Taniguchi H Ishikawa S Ito H Yamamoto S Overexpression of the aldo-keto reductase family protein AKR1B10 is highly correlated with smokers' non-small cell lung carcinomas Clin Cancer Res 2005 11 1776 1785 10.1158/1078-0432.CCR-04-1238 15755999
Furmanik M Shanahan MC Endoplasmic reticulum stress in arterial smooth muscle cells: a novel regulator of vascular disease Curr Cardiol Rev 2017 13 94 105 10.2174/1573403X12666161014094738 27758694
Furuhashi M New insights into purine metabolism in metabolic diseases: role of xanthine oxidoreductase activity Am J Physiol Endocrinol Metab 2020 319 E827 E834 10.1152/ajpendo.00378.2020 32893671
Gadd DA Hillary RF Mccartney DL Shi L Stolicyn A Robertson NA Walker RM Mcgeachan RI Campbell A Xueyi S Integrated methylome and phenome study of the circulating proteome reveals markers pertinent to brain health Nat Commun 2022 13 1 14 10.1038/s41467-022-32319-8 34983933
Gallego O Ruiz FX Ardèvol A Domínguez M Alvarez R De Lera AR Rovira C Farrés J Fita I Parés X Structural basis for the high all-trans-retinaldehyde reductase activity of the tumor marker AKR1B10 Proc Natl Acad Sci USA 2007 104 20764 20769 10.1073/pnas.0705659105 18087047
Gao L Ge C Fang T Zhao F Chen T Yao M Li J Li H ANGPTL 2 promotes tumor metastasis in hepatocellular carcinoma J Gastroenterol Hepatol 2015 30 396 404 10.1111/jgh.12702 25090954
Garland P Durnford AJ Okemefuna AI Dunbar J Nicoll JA Galea J Boche D Bulters DO Galea I Heme–hemopexin scavenging is active in the brain and associates with outcome after subarachnoid hemorrhage Stroke 2016 47 872 876 10.1161/STROKEAHA.115.011956 26768209
Gauldie J Richards C Harnish D Lansdorp P Baumann H Interferon beta 2/B-cell stimulatory factor type 2 shares identity with monocyte-derived hepatocyte-stimulating factor and regulates the major acute phase protein response in liver cells Proc Natl Acad Sci USA 1987 84 7251 7255 10.1073/pnas.84.20.7251 2444978
Geyer PE Arend FM Doll S Louiset ML Virreira Winter S Müller-Reif JB Torun FM Weigand M Eichhorn P Bruegel M High-resolution serum proteome trajectories in COVID-19 reveal patient-specific seroconversion EMBO Mol Med 2021 13 e14167 10.15252/emmm.202114167 34232570
Goetz R Beenken A Ibrahimi OA Kalinina J Olsen SK Eliseenkova AV Xu C Neubert TA Zhang F Linhardt RJ Molecular insights into the klotho-dependent, endocrine mode of action of fibroblast growth factor 19 subfamily members Mol Cell Biol 2007 27 3417 3428 10.1128/MCB.02249-06 17339340
Goncalves CA Sesterheim P Serum amyloid A protein has been undervalued as a biomarker of COVID-19 Diabetes Metab Res Rev 2021 37 e3376 10.1002/dmrr.3376 32588938
Gospodarowicz D Purification of a fibroblast growth factor from bovine pituitary J. Biol. Chem 1975 250 7 2515 2520 10.1016/S0021-9258(19)41631-1 1168187
Goulding DR Nikolova VD Mishra L Zhuo L Kimata K Mcbride SJ Moy SS Harry GJ Garantziotis S Inter-α-inhibitor deficiency in the mouse is associated with alterations in anxiety-like behavior, exploration and social approach Genes Brain Behav 2019 18 e12505 10.1111/gbb.12505 29987918
Graham TE Yang Q Blüher M Hammarstedt A Ciaraldi TP Henry RR Wason CJ Oberbach A Jansson P-A Smith U Retinol-binding protein 4 and insulin resistance in lean, obese, and diabetic subjects N Engl J Med 2006 354 2552 2563 10.1056/NEJMoa054862 16775236
Graham D Huynh NN Hamilton CA Beattie E Smith RA Cochemé HM Murphy MP Dominiczak AF Mitochondria-targeted antioxidant MitoQ10 improves endothelial function and attenuates cardiac hypertrophy Hypertension 2009 54 322 328 10.1161/HYPERTENSIONAHA.109.130351 19581509
Graham MJ Lee RG Brandt TA Tai L-J Fu W Peralta R Yu R Hurh E Paz E Mcevoy BW Cardiovascular and metabolic effects of ANGPTL3 antisense oligonucleotides N Engl J Med 2017 377 222 232 10.1056/NEJMoa1701329 28538111
Graw JA Mayeur C Rosales I Liu Y Sabbisetti VS Riley FE Rechester O Malhotra R Warren HS Colvin RB Haptoglobin or hemopexin therapy prevents acute adverse effects of resuscitation after prolonged storage of red cells Circulation 2016 134 945 960 10.1161/CIRCULATIONAHA.115.019955 27515135
Gupta S Deepti A Deegan S Lisbona F Hetz C Samali A HSP72 protects cells from ER stress-induced apoptosis via enhancement of IRE1α-XBP1 signaling through a physical interaction PLoS Biol 2010 8 e1000410 10.1371/journal.pbio.1000410 20625543
Hackler J Heller RA Sun Q Schwarzer M Diegmann J Bachmann M Moghaddam A Schomburg L Relation of serum copper status to survival in COVID-19 Nutrients 2021 10.3390/nu13061898 34203015
Haider DG Schindler K Prager G Bohdjalian A Luger A Wolzt M Ludvik B Serum retinol-binding protein 4 is reduced after weight loss in morbidly obese subjects J Clin Endocrinol Metab 2007 92 1168 1171 10.1210/jc.2006-1839 17164313
Hamm A Veeck J Bektas N Wild PJ Hartmann A Heindrichs U Kristiansen G Werbowetski-Ogilvie T Del Maestro R Knuechel R Frequent expression loss of Inter-alpha-trypsin inhibitor heavy chain (ITIH) genes in multiple human solid tumors: a systematic expression analysis BMC Cancer 2008 8 1 15 10.1186/1471-2407-8-25 18173856
Hammoutene A Rautou P-E Role of liver sinusoidal endothelial cells in non-alcoholic fatty liver disease J Hepatol 2019 70 1278 1291 10.1016/j.jhep.2019.02.012 30797053
Harrison SA Rinella ME Abdelmalek MF Trotter JF Paredes AH Arnold HL Kugelmas M Bashir MR Jaros MJ Ling L NGM282 for treatment of non-alcoholic steatohepatitis: a multicentre, randomised, double-blind, placebo-controlled, phase 2 trial Lancet 2018 391 1174 1185 10.1016/S0140-6736(18)30474-4 29519502
Heringlake S Hofdmann M Fiebeler A Manns MP Schmiegel W Tannapfel A Identification and expression analysis of the aldo–ketoreductase1-B10 gene in primary malignant liver tumours J Hepatol 2010 52 220 227 10.1016/j.jhep.2009.11.005 20036025
Hetz C Bernasconi P Fisher J Lee A-H Bassik MC Antonsson B Brandt GS Iwakoshi NN Schinzel A Glimcher LH Proapoptotic BAX and BAK modulate the unfolded protein response by a direct interaction with IRE1α Science 2006 312 572 576 10.1126/science.1123480 16645094
Hetz C Zhang K Kaufman RJ Mechanisms, regulation and functions of the unfolded protein response Nat Rev Mol Cell Biol 2020 21 421 438 10.1038/s41580-020-0250-z 32457508
Hochberg I Hochberg Z Expanding the definition of hypothalamic obesity Obes Rev 2010 11 709 721 10.1111/j.1467-789X.2010.00727.x 20233310
Holt H Wild S Wood P Zhang J Darekar A Dewbury K Poole R Holt R Phillips D Byrne C Non-esterified fatty acid concentrations are independently associated with hepatic steatosis in obese subjects Diabetologia 2006 49 141 148 10.1007/s00125-005-0070-x 16323001
Hossain P Kawar B El Nahas M Obesity and diabetes in the developing world: a growing challenge N Engl J Med 2007 356 213 215 10.1056/NEJMp068177 17229948
Hou J Zhang J Cui P Zhou Y Liu C Wu X Ji Y Wang S Cheng B Ye H Shu L Zhang K Wang D Xu J Shu Q Colonna M Fang X TREM2 sustains macrophage-hepatocyte metabolic coordination in nonalcoholic fatty liver disease and sepsis J Clin Invest 2021 10.1172/JCI135197 34263740
Huang Y-H Tiao M-M Huang L-T Chuang J-H Kuo K-C Yang Y-L Wang F-S Activation of Mir-29a in activated hepatic stellate cells modulates its profibrogenic phenotype through inhibition of histone deacetylases 4 PLoS ONE 2015 10 e0136453 10.1371/journal.pone.0136453 26305546
Hughes D Mallucci GR The unfolded protein response in neurodegenerative disorders–therapeutic modulation of the PERK pathway FEBS J 2019 286 342 355 10.1111/febs.14422 29476642
Hui E Yeung C-Y Lee PC Woo Y-C Fong CH Chow W-S Xu A Lam KS Elevated circulating pigment epithelium-derived factor predicts the progression of diabetic nephropathy in patients with type 2 diabetes J Clin Endocrinol Metab 2014 99 E2169 E2177 10.1210/jc.2014-2235 25166721
Ikeda K Sannoh T Kawasaki N Kawasaki T Yamashina I Serum lectin with known structure activates complement through the classical pathway J Biol Chem 1987 262 7451 7454 10.1016/S0021-9258(18)47587-4 3584121
Ivanov I Matafonov A Sun M-f Mohammed BM Cheng Q Dickeson SK Kundu S Verhamme IM Gruber A Mccrae K A mechanism for hereditary angioedema with normal C1 inhibitor: an inhibitory regulatory role for the factor XII heavy chain Blood 2019 133 1152 1163 10.1182/blood-2018-06-860270 30591525
Jana AK Greenwood AB Hansmann UHE Presence of a SARS-CoV-2 protein enhances amyloid formation of serum amyloid A J Phys Chem B 2021 125 9155 9167 10.1021/acs.jpcb.1c04871 34370466
Jiao N Baker SS Chapa-Rodriguez A Liu W Nugent CA Tsompana M Mastrandrea L Buck MJ Baker RD Genco RJ Suppressed hepatic bile acid signalling despite elevated production of primary and secondary bile acids in NAFLD Gut 2018 67 1881 1891 10.1136/gutjnl-2017-314307 28774887
Jim-No K Tanimizu M Hyodo I Kurimoto F Yamashita T Plasma level of basic fibroblast growth factor increases with progression of chronic liver disease J Gastroenterol 1997 32 119 121 10.1007/BF01213308 9058307
Jin J Evaluating the effects of PCBs in non-alcoholic fatty liver disease & diabetes and the role of AhR in regulating the hepatic proteome and lipid metabolism 2020 Louisville Department of Pharmacology and Toxicology. University of Louisville 3549
Jung TW Chung YH Kim H-C Abd El-Aty A Jeong JH Hyperlipidemia-induced hepassocin in the liver contributes to insulin resistance in skeletal muscle Mol Cell Endocrinol 2018 470 26 33 10.1016/j.mce.2017.10.014 29111387
Kanno M Kawaguchi K Honda M Horii R Takatori H Shimakami T Kitamura K Arai K Yamashita T Sakai Y Serum aldo–keto reductase family 1 member B10 predicts advanced liver fibrosis and fatal complications of nonalcoholic steatohepatitis J Gastroenterol 2019 54 549 557 10.1007/s00535-019-01551-3 30707282
Karnaukhova E C1-inhibitor: structure, functional diversity and therapeutic development Curr Med Chem 2022 29 467 488 10.2174/0929867328666210804085636 34348603
Katsiki N Mikhailidis DP Mantzoros CS Non-alcoholic fatty liver disease and dyslipidemia: an update Metabolism 2016 65 1109 1123 10.1016/j.metabol.2016.05.003 27237577
Kawachi Y Fujishima Y Nishizawa H Nakamura T Akari S Murase T Saito T Miyazaki Y Nagao H Fukuda S Increased plasma XOR activity induced by NAFLD/NASH and its possible involvement in vascular neointimal proliferation JCI Insight 2021 10.1172/jci.insight.144762 34494551
Kawaguchi T Yamagishi S-I Sata M Structure-function relationships of PEDF Curr Mol Med 2010 10 302 311 10.2174/156652410791065255 20236052
Kim YC Guan K-L mTOR: a pharmacologic target for autophagy regulation J Clin Invest 2015 125 25 32 10.1172/JCI73939 25654547
Kim YS Kim SG Endoplasmic reticulum stress and autophagy dysregulation in alcoholic and non-alcoholic liver diseases Clin Mol Hepatol 2020 26 715 10.3350/cmh.2020.0173 32951410
Kim JS Nitta T Mohuczy D O’malley KA Moldawer LL Dunn WA Jr Behrns KE Impaired autophagy: a mechanism of mitochondrial dysfunction in anoxic rat hepatocytes Hepatology 2008 47 1725 1736 10.1002/hep.22187 18311843
Kim KH Jeong YT Oh H Kim SH Cho JM Kim Y-N Kim SS Kim DH Hur KY Kim HK Autophagy deficiency leads to protection from obesity and insulin resistance by inducing Fgf21 as a mitokine Nat Med 2013 19 83 92 10.1038/nm.3014 23202295
Kim KM Han CY Kim JY Cho SS Kim YS Koo JH Lee JM Lim SC Kang KW Kim J-S Gα12 overexpression induced by miR-16 dysregulation contributes to liver fibrosis by promoting autophagy in hepatic stellate cells J Hepatol 2018 68 493 504 10.1016/j.jhep.2017.10.011 29080810
Kim TH Koo JH Heo MJ Han CY Kim Y-I Park S-Y Cho IJ Lee CH Choi CS Lee JW Overproduction of inter-α-trypsin inhibitor heavy chain 1 after loss of Gα13 in liver exacerbates systemic insulin resistance in mice Sci Transl Med 2019 11 eaan4735 10.1126/scitranslmed.aan4735 31597753
Kim Y Nam H Han C Joo M Jang K Jun D Kim S LXRα activation inhibits autophagy and lipophagy in hepatocytes by dysregulating ATG4B and Rab-8B, reducing mitochondrial fuel oxidation Hepatology 2020 73 1307 1326 10.1002/hep.31423
Kim YS Nam HJ Han CY Joo MS Jang K Jun DW Kim SG Liver X receptor alpha activation inhibits autophagy and lipophagy in hepatocytes by dysregulating autophagy-related 4B cysteine peptidase and Rab-8B, reducing mitochondrial fuel oxidation Hepatology 2021 73 1307 1326 10.1002/hep.31423 32557804
Koishi R Ando Y Ono M Shimamura M Yasumo H Fujiwara T Horikoshi H Furukawa H Angptl3 regulates lipid metabolism in mice Nat Genet 2002 30 151 157 10.1038/ng814 11788823
Kono S Aceruloplasminemia Curr Drug Targets 2012 13 1190 1199 10.2174/138945012802002320 22515740
Koo JH Lee HJ Kim W Kim SG Endoplasmic reticulum stress in hepatic stellate cells promotes liver fibrosis via PERK-mediated degradation of HNRNPA1 and up-regulation of SMAD2 Gastroenterology 2016 150 181–193 e8 10.1053/j.gastro.2015.09.039
Koo JH Kim TH Park SY Joo MS Han CY Choi CS Kim SG Gα13 ablation reprograms myofibers to oxidative phenotype and enhances whole-body metabolism J Clin Invest 2017 127 3845 3860 10.1172/JCI92067 28920922
Kovrov O Kristensen KK Larsson E Ploug M Olivecrona G On the mechanism of angiopoietin-like protein 8 for control of lipoprotein lipase activity J Lipid Res 2019 60 783 793 10.1194/jlr.M088807 30686789
Kumar A Sundaram K Mu J Dryden GW Sriwastva MK Lei C Zhang L Qiu X Xu F Yan J High-fat diet-induced upregulation of exosomal phosphatidylcholine contributes to insulin resistance Nat Commun 2021 12 1 21 10.1038/s41467-020-20500-w 33397941
Kuo T-C Tan C-T Chang Y-W Hong C-C Lee W-J Chen M-W Jeng Y-M Chiou J Yu P Chen P-S Angiopoietin-like protein 1 suppresses SLUG to inhibit cancer cell motility J Clin Investig 2013 123 1082 1095 10.1172/JCI64044 23434592
Kurosu H Choi M Ogawa Y Dickson AS Goetz R Eliseenkova AV Mohammadi M Rosenblatt KP Kliewer SA Kuro-O M Tissue-specific expression of βKlotho and fibroblast growth factor (FGF) receptor isoforms determines metabolic activity of FGF19 and FGF21 J Biol Chem 2007 282 26687 26695 10.1074/jbc.M704165200 17623664
Kwanten WJ Martinet W Michielsen PP Francque SM Role of autophagy in the pathophysiology of nonalcoholic fatty liver disease: a controversial issue World J Gastroenterol 2014 20 7325 10.3748/wjg.v20.i23.7325 24966603
Le MH Devaki P Ha NB Jun DW Te HS Cheung RC Nguyen MH Prevalence of non-alcoholic fatty liver disease and risk factors for advanced fibrosis and mortality in the United States PLoS ONE 2017 12 e0173499 10.1371/journal.pone.0173499 28346543
Lee K Tirasophon W Shen X Michalak M Prywes R Okada T Yoshida H Mori K Kaufman RJ IRE1-mediated unconventional mRNA splicing and S2P-mediated ATF6 cleavage merge to regulate XBP1 in signaling the unfolded protein response Genes Dev 2002 16 452 466 10.1101/gad.964702 11850408
Lee Y Manegold JE Kim AD Pouget C Stachura DL Clements WK Traver D FGF signalling specifies haematopoietic stem cells through its regulation of somitic Notch signalling Nat Commun 2014 5 1 13 10.1038/ncomms6583
Leeb-Lundberg LF Marceau F Müller-Esterl W Pettibone DJ Zuraw BL International union of pharmacology. XLV. Classification of the kinin receptor family: from molecular mechanisms to pathophysiological consequences Pharmacol Rev 2005 57 27 77 10.1124/pr.57.1.2 15734727
Lefere S Puengel T Hundertmark J Penners C Frank AK Guillot A De Muynck K Heymann F Adarbes V Defrêne E Differential effects of selective-and pan-PPAR agonists on experimental steatohepatitis and hepatic macrophages J Hepatol 2020 73 757 770 10.1016/j.jhep.2020.04.025 32360434
Li Y Teng C Angiopoietin-like proteins 3, 4 and 8: regulating lipid metabolism and providing new hope for metabolic syndrome J Drug Target 2014 22 679 687 10.3109/1061186X.2014.928715 24960069
Li B Nolte LA Ju J-S Ho Han D Coleman T Holloszy JO Semenkovich CF Skeletal muscle respiratory uncoupling prevents diet-induced obesity and insulin resistance in mice Nat Med 2000 6 1115 1120 10.1038/80450 11017142
Li D Zhang H Zhong Y Hepatic GDF15 is regulated by CHOP of the unfolded protein response and alleviates NAFLD progression in obese mice Biochem Biophys Res Commun 2018 498 388 394 10.1016/j.bbrc.2017.08.096 28847729
Lichtenstein L Mattijssen F De Wit NJ Georgiadi A Hooiveld GJ Van Der Meer R He Y Qi L Köster A Tamsma JT Angptl4 protects against severe proinflammatory effects of saturated fat by inhibiting fatty acid uptake into mesenteric lymph node macrophages Cell Metab 2010 12 580 592 10.1016/j.cmet.2010.11.002 21109191
Lin T Maita D Thundivalappil SR Riley FE Hambsch J Van Marter LJ Christou HA Berra L Fagan S Christiani DC Hemopexin in severe inflammation and infection: mouse models and human diseases Crit Care 2015 19 1 8 10.1186/s13054-015-0885-x 25560635
Little JL Wheeler FB Fels DR Koumenis C Kridel SJ Inhibition of fatty acid synthase induces endoplasmic reticulum stress in tumor cells Cancer Res 2007 67 1262 1269 10.1158/0008-5472.CAN-06-1794 17283163
Liu DJ Peloso GM Yu H Butterworth AS Wang X Mahajan A Saleheen D Emdin C Alam D Alves AC Exome-wide association study of plasma lipids in> 300,000 individuals Nat Genet 2017 49 1758 1766 10.1038/ng.3977 29083408
Liu Z Zhang Y Graham S Wang X Cai D Huang M Pique-Regi R Dong XC Chen YE Willer C Liu W Causal relationships between NAFLD, T2D and obesity have implications for disease subphenotyping J Hepatol 2020 73 263 276 10.1016/j.jhep.2020.03.006 32165250
Liu C Zhou B Meng M Zhao W Wang D Yuan Y Zheng Y Qiu J Li Y Li G FOXA3 induction under endoplasmic reticulum stress contributes to non-alcoholic fatty liver disease J Hepatol 2021 75 150 162 10.1016/j.jhep.2021.01.042 33548387
Liu C-H Zheng S Wang S Wu D Jiang W Zeng Q Wei Y Zhang Y Tang H Urine proteome in distinguishing hepatic steatosis in patients with metabolic-associated fatty liver disease Diagnostics 2022 12 1412 10.3390/diagnostics12061412 35741222
Lonardo A Lugari S Ballestri S Nascimbeni F Baldelli E Maurantonio M A round trip from nonalcoholic fatty liver disease to diabetes: molecular targets to the rescue? Acta Diabetol 2019 56 385 396 10.1007/s00592-018-1266-0 30519965
Loréal O Turlin B Pigeon C Moisan A Ropert M Morice P Gandon Y Jouanolle A-M Vérin M Hider RC Aceruloplasminemia: new clinical, pathophysiological and therapeutic insights J Hepatol 2002 36 851 856 10.1016/S0168-8278(02)00042-9 12044538
Luo D-X Huang MC Ma J Gao Z Liao D-f Cao D Aldo-keto reductase family 1, member B10 is secreted through a lysosome-mediated non-classical pathway Biochem J 2011 438 71 80 10.1042/BJ20110111 21585341
Ma J Luo DX Huang C Shen Y Bu Y Markwell S Gao J Liu J Zu X Cao Z AKR1B10 overexpression in breast cancer: association with tumor size, lymph node metastasis and patient survival and its potential as a novel serum marker Int J Cancer 2012 131 E862 E871 10.1002/ijc.27618 22539036
Ma Y Wang Y Wang X Yang L Lang J Zhao D 1884-P: Serum ANGPTL2, ANGPTL3, and ANGPTL4 as potential biomarkers for diagnosis of nonalcoholic fatty liver disease Diabetes 2019 10.2337/db19-1884-P 31451517
Machado M Marques-Vidal P Cortez-Pinto H Hepatic histology in obese patients undergoing bariatric surgery J Hepatol 2006 45 600 606 10.1016/j.jhep.2006.06.013 16899321
Makri E Goulas A Polyzos SA Epidemiology, pathogenesis, diagnosis and emerging treatment of nonalcoholic fatty liver disease Arch Med Res 2021 52 25 37 10.1016/j.arcmed.2020.11.010 33334622
Mansour E Bueno FF De Lima-Junior JC Palma A Monfort-Pires M Bombassaro B Araujo EP Bernardes AF Ulaf RG Nunes TA Ribeiro LC Falcao ALE Santos TM Trabasso P Dertkigil RP Dertkigil SS Maia RP Benaglia T Moretti ML Velloso LA Evaluation of the efficacy and safety of icatibant and C1 esterase/kallikrein inhibitor in severe COVID-19: study protocol for a three-armed randomized controlled trial Trials 2021 22 71 10.1186/s13063-021-05027-9 33472675
Manuel CR Latuga MS Ashby CR Jr Reznik SE Immune tolerance attenuates gut dysbiosis, dysregulated uterine gene expression and high-fat diet potentiated preterm birth in mice Am J Obstet Gynecol 2019 220 596.e1 596.e28 10.1016/j.ajog.2019.02.028 30790568
Martinet W Agostinis P Vanhoecke B Dewaele M De Meyer GR Autophagy in disease: a double-edged sword with therapeutic potential Clin Sci 2009 116 697 712 10.1042/CS20080508
Mccullough KD Martindale JL Klotz L-O Aw T-Y Holbrook NJ Gadd153 sensitizes cells to endoplasmic reticulum stress by down-regulating Bcl2 and perturbing the cellular redox state Mol Cell Biol 2001 21 1249 1259 10.1128/MCB.21.4.1249-1259.2001 11158311
Medjeral-Thomas NR Troldborg A Hansen AG Pihl R Clarke CL Peters JE Thomas DC Willicombe M Palarasah Y Botto M Protease inhibitor plasma concentrations associate with COVID-19 infection Oxf Open Immunol 2021 2 iqab014 10.1093/oxfimm/iqab014 34458849
Meex RC Hoy AJ Morris A Brown RD Lo JC Burke M Goode RJ Kingwell BA Kraakman MJ Febbraio MA Fetuin B is a secreted hepatocyte factor linking steatosis to impaired glucose metabolism Cell Metab 2015 22 1078 1089 10.1016/j.cmet.2015.09.023 26603189
Miller FJ Jr Hypertension and mitochondrial oxidative stress revisited: sirtuin 3, the improved “antioxidant” Circ Res 2020 126 453 455 10.1161/CIRCRESAHA.120.316567 32078456
Miura S Mitsuhashi N Shimizu H Kimura F Yoshidome H Otsuka M Kato A Shida T Okamura D Miyazaki M Fibroblast growth factor 19 expression correlates with tumor progression and poorer prognosis of hepatocellular carcinoma BMC Cancer 2012 12 1 15 10.1186/1471-2407-12-56 22212211
Müller F Renné T Novel roles for factor XII-driven plasma contact activation system Curr Opin Hematol 2008 15 516 521 10.1097/MOH.0b013e328309ec85 18695377
Mundi MS Velapati S Patel J Kellogg TA Abu Dayyeh BK Hurt RT Evolution of NAFLD and its management Nutr Clin Pract 2020 35 72 84 10.1002/ncp.10449 31840865
Musso G Olivetti C Cassader M Gambino R Obstructive sleep apnea–hypopnea syndrome and nonalcoholic fatty liver disease: emerging evidence and mechanisms Semin Liver Dis 2012 32 049 064 10.1055/s-0032-1306426
Niederau C Fischer R Purschel A Stremmel W Haussinger D Strohmeyer G Long-term survival in patients with hereditary hemochromatosis Gastroenterology 1996 110 1107 1119 10.1053/gast.1996.v110.pm8613000 8613000
Nobili V Alkhouri N Alisi A Ottino S Lopez R Manco M Feldstein AE Retinol-binding protein 4: a promising circulating marker of liver damage in pediatric nonalcoholic fatty liver disease Clin Gastroenterol Hepatol 2009 7 575 579 10.1016/j.cgh.2008.12.031 19268270
Nobili V Siotto M Bedogni G Ravà L Pietrobattista A Panera N Alisi A Squitti R Levels of serum ceruloplasmin associate with pediatric nonalcoholic fatty liver disease J Pediatr Gastroenterol Nutr 2013 56 370 375 10.1097/MPG.0b013e31827aced4 23154483
Norseen J Hosooka T Hammarstedt A Yore MM Kant S Aryal P Kiernan UA Phillips DA Maruyama H Kraus BJ Retinol-binding protein 4 inhibits insulin signaling in adipocytes by inducing proinflammatory cytokines in macrophages through a c-Jun N-terminal kinase-and toll-like receptor 4-dependent and retinol-independent mechanism Mol Cell Biol 2012 32 2010 2019 10.1128/MCB.06193-11 22431523
O’brien A Zhou T White T Medford A Chen L Kyritsi K Wu N Childs J Stiles D Ceci L FGF1 signaling modulates biliary injury and liver fibrosis in the Mdr2−/− mouse model of primary sclerosing cholangitis Hepatol Commun 2022 10.1002/hep4.1909 35271760
Observatory WOGO India National Obesity Prevalence. World Obesity Federation Global Obesity Observatory (2021a)
Observatory WOGO Pakistan National Obesity Prevalence. World Obesity Federation Global Obesity Observatory (2021b)
Oike Y Yasunaga K Ito Y Matsumoto S-i Maekawa H Morisada T Arai F Nakagata N Takeya M Masuho Y Angiopoietin-related growth factor (AGF) promotes epidermal proliferation, remodeling, and regeneration Proc Natl Acad Sci USA 2003 100 9494 9499 10.1073/pnas.1531901100 12871997
Oike Y Akao M Yasunaga K Yamauchi T Morisada T Ito Y Urano T Kimura Y Kubota Y Maekawa H Angiopoietin-related growth factor antagonizes obesity and insulin resistance Nat Med 2005 11 400 408 10.1038/nm1214 15778720
Okashah N Wright SC Kawakami K Mathiasen S Zhou J Lu S Javitch JA Inoue A Bouvier M Lambert NA Agonist-induced formation of unproductive receptor-G12 complexes Proc Natl Acad Sci USA 2020 117 21723 21730 10.1073/pnas.2003787117 32817560
Ooi GJ Meikle PJ Huynh K Earnest A Roberts SK Kemp W Parker BL Brown W Burton P Watt MJ Hepatic lipidomic remodeling in severe obesity manifests with steatosis and does not evolve with non-alcoholic steatohepatitis J Hepatol 2021 75 524 535 10.1016/j.jhep.2021.04.013 33887358
Ornitz DM Itoh N The fibroblast growth factor signaling pathway Wiley Interdiscip Rev Dev Biol 2015 4 215 266 10.1002/wdev.176 25772309
Ornitz DM Itoh N New developments in the biology of fibroblast growth factors Wires Mech Dis 2022 14 e1549 10.1002/wsbm.1549 35142107
Oschatz C Maas C Lecher B Jansen T Björkqvist J Tradler T Sedlmeier R Burfeind P Cichon S Hammerschmidt S Mast cells increase vascular permeability by heparin-initiated bradykinin formation in vivo Immunity 2011 34 258 268 10.1016/j.immuni.2011.02.008 21349432
Oslowski CM Urano F Measuring ER stress and the unfolded protein response using mammalian tissue culture system Methods Enzymol 2011 490 71 92 10.1016/B978-0-12-385114-7.00004-0 21266244
Ou H-Y Wu H-T Lin C-H Du Y-F Hu C-Y Hung H-C Wu P Li H-Y Wang S-H Chang C-J The hepatic protection effects of hepassocin in hyperglycemic crisis J Clin Endocrinol Metab 2017 102 2407 2415 10.1210/jc.2016-3287 28402540
Pacher P Nivorozhkin A Szabó C Therapeutic effects of xanthine oxidase inhibitors: renaissance half a century after the discovery of allopurinol Pharmacol Rev 2006 58 87 114 10.1124/pr.58.1.6 16507884
Pal D Dasgupta S Kundu R Maitra S Das G Mukhopadhyay S Ray S Majumdar SS Bhattacharya S Fetuin-A acts as an endogenous ligand of TLR4 to promote lipid-induced insulin resistance Nat Med 2012 18 1279 1285 10.1038/nm.2851 22842477
Pan RL Xiang LX Wang P Liu XY Nie L Huang W Shao JZ Low-molecular-weight fibroblast growth factor 2 attenuates hepatic fibrosis by epigenetic down-regulation of Delta-like1 Hepatology 2015 61 1708 1720 10.1002/hep.27649 25501710
Pan X-F Wang L Pan A Epidemiology and determinants of obesity in China Lancet Diabetes Endocrinol 2021 9 373 392 10.1016/S2213-8587(21)00045-0 34022156
Parker R Kim S Im G Nahas J Dhesi B Vergis N Sinha A Ghezzi A Rink M Mccune A Obesity in acute alcoholic hepatitis increases morbidity and mortality EBioMedicine 2019 45 511 518 10.1016/j.ebiom.2019.03.046 31278069
Pelucchi S Ravasi G Piperno A Ceruloplasmin variants might have different effects in different iron overload disorders J Hepatol 2021 75 1003 1004 10.1016/j.jhep.2021.05.005 34023349
Peter A Kovarova M Staiger H Machann J Schick F Königsrainer A Koenigsrainer I Schleicher E Fritsche A Häring H-U The hepatokines fetuin-A and fetuin-B are upregulated in the state of hepatic steatosis and may differently impact on glucose homeostasis in humans Am J Physiol Endocrinol Metab 2018 314 E266 E273 10.1152/ajpendo.00262.2017 29138227
Petito-Da-Silva TI Souza-Mello V Barbosa-Da-Silva S Empaglifozin mitigates NAFLD in high-fat-fed mice by alleviating insulin resistance, lipogenesis and ER stress Mol Cell Endocrinol 2019 498 110539 10.1016/j.mce.2019.110539 31419466
Piccinin E Villani G Moschetta A Metabolic aspects in NAFLD, NASH and hepatocellular carcinoma: the role of PGC1 coactivators Nat Rev Gastroenterol Hepatol 2019 16 160 174 10.1038/s41575-018-0089-3 30518830
Pihl R Jensen RK Poulsen EC Jensen L Hansen AG Thøgersen IB Dobó J Gál P Andersen GR Enghild JJ ITIH4 acts as a protease inhibitor by a novel inhibitory mechanism Sci Adv 2021 7 eaba7381 10.1126/sciadv.aba7381 33523981
Polyzos SA Kountouras J Mantzoros CS Obesity and nonalcoholic fatty liver disease: from pathophysiology to therapeutics Metabolism 2019 92 82 97 10.1016/j.metabol.2018.11.014 30502373
Powell EE Wong VW-S Rinella M Non-alcoholic fatty liver disease Lancet 2021 397 2212 2224 10.1016/s0140-6736(20)32511-3 33894145
Protiva P Gong J Sreekumar B Torres R Zhang X Belinsky GS Cornwell M Crawford SE Iwakiri Y Chung C Pigment epithelium-derived factor (PEDF) inhibits Wnt/β-catenin signaling in the liver Cell Mol Gastroenterol Hepatol 2015 1 535–549 e14 10.1016/j.jcmgh.2015.06.006
Prystupa A Dąbrowska A Sak JJ Tarach J Toruń-Jurkowska A Lachowska-Kotowska P Dzida G Concentrations of fetuin-A, osteoprotegerin and α-Klotho in patients with alcoholic liver cirrhosis Exp Ther Med 2016 12 3464 3470 10.3892/etm.2016.3754 27882180
Rada P González-Rodríguez Á García-Monzón C Valverde ÁM Understanding lipotoxicity in NAFLD pathogenesis: is CD36 a key driver? Cell Death Dis 2020 11 1 15 10.1038/s41419-020-03003-w 31911576
Rey E Meléndez-Rodríguez F Marañón P Gil-Valle M Carrasco AG Torres-Capelli M Chávez S Del Pozo-Maroto E Rodríguez De Cía J Aragonés J Hypoxia-inducible factor 2α drives hepatosteatosis through the fatty acid translocase CD36 Liver Int 2020 40 2553 2567 10.1111/liv.14519 32432822
Roeser H Lee G Nacht S Cartwright G The role of ceruloplasmin in iron metabolism J Clin Investig 1970 49 2408 2417 10.1172/JCI106460 5480864
Romeo S Pennacchio LA Fu Y Boerwinkle E Tybjaerg-Hansen A Hobbs HH Cohen JC Population-based resequencing of ANGPTL4 uncovers variations that reduce triglycerides and increase HDL Nat Genet 2007 39 513 516 10.1038/ng1984 17322881
Romeo S Yin W Kozlitina J Pennacchio LA Boerwinkle E Hobbs HH Cohen JC Rare loss-of-function mutations in ANGPTL family members contribute to plasma triglyceride levels in humans J Clin Investig 2009 119 70 79 10.1172/JCI37118 19075393
Roth CL Hypothalamic obesity in craniopharyngioma patients: disturbed energy homeostasis related to extent of hypothalamic damage and its implication for obesity intervention J Clin Med 2015 4 1774 1797 10.3390/jcm4091774 26371051
Ryu S Chang Y Jung H-S Yun KE Kwon M-J Choi Y Kim C-W Cho J Suh B-S Cho YK Relationship of sitting time and physical activity with non-alcoholic fatty liver disease J Hepatol 2015 63 1229 1237 10.1016/j.jhep.2015.07.010 26385766
Saguchi K-i Tobe T Hashimoto K Nagasaki Y Oda E Nakano Y Miura N-H Tomita M Isolation and characterization of the human inter-α-trypsin inhibitor family heavy chain-related protein (IHRP) gene (ITIHL1) J Biochem 1996 119 898 905 10.1093/oxfordjournals.jbchem.a021327 8797089
Samsoondar JP Burke AC Sutherland BG Telford DE Sawyez CG Edwards JY Pinkosky SL Newton RS Huff MW Prevention of diet-induced metabolic dysregulation, inflammation, and atherosclerosis in Ldlr−/− mice by treatment with the ATP-citrate lyase inhibitor bempedoic acid Arterioscler Thromb Vasc Biol 2017 37 647 656 10.1161/ATVBAHA.116.308963 28153881
Sato-Matsubara M Matsubara T Daikoku A Okina Y Longato L Rombouts K Adachi J Tomonaga T Ikeda K Yoshizato K Fibroblast growth factor 2 (FGF2) regulates cytoglobin expression and activation of human hepatic stellate cells via JNK signaling J Biol Chem 2017 292 18961 18972 10.1074/jbc.M117.793794 28916723
Schumacher JD Guo GL Regulation of hepatic stellate cells and fibrogenesis by fibroblast growth factors Biomed Res Int 2016 10.1155/2016/8323747 27975063
Serbource-Goguel N Corbic M Erlinger S Durand G Agneray J Feger J Measurement of serum α1-acid glycoprotein and cei-antitrypsin desialylation in liver disease Hepatology 1983 3 356 359 10.1002/hep.1840030313 6601607
Singh R Kaushik S Wang Y Xiang Y Novak I Komatsu M Tanaka K Cuervo AM Czaja MJ Autophagy regulates lipid metabolism Nature 2009 458 1131 1135 10.1038/nature07976 19339967
Singh AK Chaube B Zhang X Sun J Citrin KM Canfrán-Duque A Aryal B Rotllan N Varela L Lee RG Hepatocyte-specific suppression of ANGPTL4 improves obesity-associated diabetes and mitigates atherosclerosis in mice J Clin Investig 2021 10.1172/JCI140989 34255741
Spitler KM Matsumoto T Webb RC Suppression of endoplasmic reticulum stress improves endothelium-dependent contractile responses in aorta of the spontaneously hypertensive rat Am J Physiol Heart Circ Physiol 2013 305 H344 H353 10.1152/ajpheart.00952.2012 23709602
Spitler KM Shetty SK Cushing EM Sylvers-Davie KL Davies BS Chronic high-fat feeding and prolonged fasting in liver-specific ANGPTL4 knockout mice Am J Physiol Heart Circ Physiol 2021 321 E464 E478 10.1152/ajpendo.00144.2021
Su Q Kim SY Adewale F Zhou Y Aldler C Ni M Wei Y Burczynski ME Atwal GS Sleeman MW Single-cell RNA transcriptome landscape of hepatocytes and non-parenchymal cells in healthy and NAFLD mouse liver iScience 2021 24 103233 10.1016/j.isci.2021.103233 34755088
Subudhi S Drescher HK Dichtel LE Bartsch LM Chung RT Hutter MM Gee DW Meireles OR Witkowski ER Gelrud L Distinct hepatic gene-expression patterns of NAFLD in patients with obesity Hepatol Commun 2022 6 77 89 10.1002/hep4.1789 34558849
Suzuki N Hajicek N Kozasa T Regulation and physiological functions of G12/13-mediated signaling pathways Neurosignals 2009 17 55 70 10.1159/000186690 19212140
Suzuki Y Takadate T Mizuma M Shima H Suzuki T Tachibana T Shimura M Hata T Iseki M Kawaguchi K Stromal expression of hemopexin is associated with lymph-node metastasis in pancreatic ductal adenocarcinoma PLoS ONE 2020 15 e0235904 10.1371/journal.pone.0235904 32663208
Tabata M Kadomatsu T Fukuhara S Miyata K Ito Y Endo M Urano T Zhu HJ Tsukano H Tazume H Angiopoietin-like protein 2 promotes chronic adipose tissue inflammation and obesity-related systemic insulin resistance Cell Metab 2009 10 178 188 10.1016/j.cmet.2009.08.003 19723494
Tak J Kim YS Kim TH Park GC Hwang S Kim SG Gα12 overexpression in hepatocytes by ER stress exacerbates acute liver injury via ROCK1-mediated miR-15a and ALOX12 dysregulation Theranostics 2022 12 1570 1588 10.7150/thno.67722 35198058
Tanaka S Hikita H Tatsumi T Sakamori R Nozaki Y Sakane S Shiode Y Nakabori T Saito Y Hiramatsu N Rubicon inhibits autophagy and accelerates hepatocyte apoptosis and lipid accumulation in nonalcoholic fatty liver disease in mice Hepatology 2016 64 1994 2014 10.1002/hep.28820 27637015
Targher G Corey KE Byrne CD Roden M The complex link between NAFLD and type 2 diabetes mellitus: mechanisms and treatments Nat Rev Gastroenterol Hepatol 2021 18 599 612 10.1038/s41575-021-00448-y 33972770
Thiel S Vorup-Jensen T Stover CM Schwaeble W Laursen SB Poulsen K Willis AC Eggleton P Hansen S Holmskov U A second serine protease associated with mannan-binding lectin that activates complement Nature 1997 386 506 510 10.1038/386506a0 9087411
Thomson TM Toscano-Guerra E Casis E Paciucci R C1 esterase inhibitor and the contact system in COVID-19 Br J Haematol 2020 190 520 524 10.1111/bjh.16938 32531085
Trepo E Romeo S Zucman-Rossi J Nahon P PNPLA3 gene in liver diseases J Hepatol 2016 65 399 412 10.1016/j.jhep.2016.03.011 27038645
Tsai T-H Chen E Li L Saha P Lee H-J Huang L-S Shelness GS Chan L Chang BH-J The constitutive lipid droplet protein PLIN2 regulates autophagy in liver Autophagy 2017 13 1130 1144 10.1080/15548627.2017.1319544 28548876
Uehara H Miyamoto M Kato K Ebihara Y Kaneko H Hashimoto H Murakami Y Hase R Takahashi R Mega S Expression of pigment epithelium-derived factor decreases liver metastasis and correlates with favorable prognosis for patients with ductal pancreatic adenocarcinoma Cancer Res 2004 64 3533 3537 10.1158/0008-5472.CAN-03-3725 15150108
Vachette P Dainese E Vasyliev VB Di Muro P Beltramini M Svergun DI De Filippis V Salvato B A key structural role for active site type 3 copper ions in human ceruloplasmin J Biol Chem 2002 277 40823 40831 10.1074/jbc.M207188200 12177070
Van De Veerdonk FL Giamarellos-Bourboulis E Pickkers P Derde L Leavis H Van Crevel R Engel JJ Wiersinga WJ Vlaar AP Shankar-Hari M A guide to immunotherapy for COVID-19 Nat Med 2022 28 39 50 10.1038/s41591-021-01643-9 35064248
Vinchi F Costa Da Silva M Ingoglia G Petrillo S Brinkman N Zuercher A Cerwenka A Tolosano E Muckenthaler MU Hemopexin therapy reverts heme-induced proinflammatory phenotypic switching of macrophages in a mouse model of sickle cell disease Blood 2016 127 473 486 10.1182/blood-2015-08-663245 26675351
Völlmy F Van Den Toorn H Zenezini Chiozzi R Zucchetti O Papi A Volta CA Marracino L Vieceli Dalla Sega F Fortini F Demichev V Tober-Lau P Campo G Contoli M Ralser M Kurth F Spadaro S Rizzo P Heck AJ A serum proteome signature to predict mortality in severe COVID-19 patients Life Sci Alliance 2021 10.26508/lsa.202101099 34226277
Von Loeffelholz C Pfeiffer AF Lock JF Lieske S Döcke S Murahovschi V Kriebel J De Las Heras Gala T Grallert H Rudovich N ANGPTL8 (betatrophin) is expressed in visceral adipose tissue and relates to human hepatic steatosis in two independent clinical collectives Horm Metab Res 2017 49 343 349 10.1055/s-0043-102950 28351093
Wang M Kaufman RJ Protein misfolding in the endoplasmic reticulum as a conduit to human disease Nature 2016 529 326 335 10.1038/nature17041 26791723
Wang C Tong Y Wen Y Hepatocellular carcinoma-associated protein TD26 interacts and enhances SREBP1 activity to promote tumor cell proliferation and growth Hepatology 2018 68 1833 1850 10.1002/hep.30030 29663480
Wang X Chen X Zhang H Pang J Lin J Xu X Yang L Ma J Ling W Chen Y Circulating retinol-binding protein 4 is associated with the development and regression of non-alcoholic fatty liver disease Diabetes Metab 2020 46 119 128 10.1016/j.diabet.2019.04.009 31078721
Wang M Lv J Huang X Wisniewski T Zhang W High-fat diet-induced atherosclerosis promotes neurodegeneration in the triple transgenic (3× Tg) mouse model of Alzheimer’s disease associated with chronic platelet activation Alzheimers Res Ther 2021 13 1 16 10.1186/s13195-021-00890-9 33397495
Washio KW Kusunoki Y Murase T Nakamura T Osugi K Ohigashi M Sukenaga T Ochi F Matsuo T Katsuno T Xanthine oxidoreductase activity is correlated with insulin resistance and subclinical inflammation in young humans Metabolism 2017 70 51 56 10.1016/j.metabol.2017.01.031 28403945
Watt MJ Miotto PM De Nardo W Montgomery MK The liver as an endocrine organ: linking NAFLD and insulin resistance Endocr Rev 2019 40 1367 1393 10.1210/er.2019-00034 31098621
Wei X Nie S Liu H Sun J Liu J Li J Li S Wang S Han S Wang J Angiopoietin-like protein 2 facilitates non-small cell lung cancer progression by promoting the polarization of M2 tumor-associated macrophages Am J Cancer Res 2017 7 2220 29218246
Wible DJ Chao H-P Tang DG Bratton SB ATG5 cancer mutations and alternative mRNA splicing reveal a conjugation switch that regulates ATG12–ATG5-ATG16L1 complex assembly and autophagy Cell Discov 2019 5 1 19 10.1038/s41421-019-0110-1 30603102
Wlazlo N Van Greevenbroek MM Ferreira I Jansen EH Feskens EJ Van Der Kallen CJ Schalkwijk CG Bravenboer B Stehouwer CD Activated complement factor 3 is associated with liver fat and liver enzymes: the CODAM study Eur J Clin Invest 2013 43 679 688 10.1111/eci.12093 23586841
Wolf PL Griffiths J Ceruloplasmin: methods and clinical use CRC Crit Rev Clin Lab Sci 1982 17 229 245 10.3109/10408368209107037
Wu H Jia W Bao Y Lu J Zhu J Wang R Chen Y Xiang K Serum retinol binding protein 4 and nonalcoholic fatty liver disease in patients with type 2 diabetes mellitus Diabetes Res Clin Pract 2008 79 185 190 10.1016/j.diabres.2007.08.016 17904683
Wu H-T Lu F-H Ou H-Y Su Y-C Hung H-C Wu J-S Yang Y-C Wu C-L Chang C-J The role of hepassocin in the development of non-alcoholic fatty liver disease J Hepatol 2013 59 1065 1072 10.1016/j.jhep.2013.06.004 23792031
Xu H Tian Y Tang D Zou S Liu G Song J Zhang G Du X Huang W He B Lin W Jin L Huang W Yang J Fu X An endoplasmic reticulum stress-MicroRNA-26a feedback circuit in NAFLD Hepatology 2021 73 1327 1345 10.1002/hep.31428 32567701
Yang Q Graham TE Mody N Preitner F Peroni OD Zabolotny JM Kotani K Quadro L Kahn BB Serum retinol binding protein 4 contributes to insulin resistance in obesity and type 2 diabetes Nature 2005 436 356 362 10.1038/nature03711 16034410
Yang L Sun R Wang Y Fu Y Zhang Y Zheng Z Ji Z Zhao D Expression of ANGPTL2 and its impact on papillary thyroid cancer Cancer Cell Int 2019 19 1 10 10.1186/s12935-019-0908-9 30622437
Yang YM Kuen D-S Chung Y Kurose H Kim SG Gα 12/13 signaling in metabolic diseases Exp Mol Med 2020 52 896 910 10.1038/s12276-020-0454-5 32576930
Yanucil C Kentrup D Campos I Czaya B Heitman K Westbrook D Osis G Grabner A Wende AR Vallejo J Soluble α-klotho and heparin modulate the pathologic cardiac actions of fibroblast growth factor 23 in chronic kidney disease Kidney Int 2022 10.1016/j.kint.2022.03.028 35513125
Yao J Wu D Zhang C Yan T Zhao Y Shen H Xue K Huang X Wang Z Qiu Y Macrophage IRX3 promotes diet-induced obesity and metabolic inflammation Nat Immunol 2021 22 1268 1279 10.1038/s41590-021-01023-y 34556885
Yaribeygi H Farrokhi FR Butler AE Sahebkar A Insulin resistance: review of the underlying molecular mechanisms J Cell Physiol 2019 234 8152 8161 10.1002/jcp.27603 30317615
Ye X Li C Zu X Lin M Liu Q Liu J Xu G Chen Z Xu Y Liu L A large-scale multicenter study validates aldo-keto reductase family 1 member B10 as a prevalent serum marker for detection of hepatocellular carcinoma Hepatology 2019 69 2489 2501 10.1002/hep.30519 30672601
Yoon H Shaw JL Haigis MC Greka A Lipid metabolism in sickness and in health: emerging regulators of lipotoxicity Mol Cell 2021 81 3708 3730 10.1016/j.molcel.2021.08.027 34547235
Younossi Z Karrar A Pierobon M Younoszai Z Felix S Jeffers T Stepanova M Jeiran K Hodge A Monge F Nano LC-mass spectrometry based proteomic profiling associated with significant fibrosis and collagen deposition in non-alcoholic fatty liver disease (NAFLD) J Hepatol 2017 1 S612 10.1016/S0168-8278(17)31664-1
Younossi ZM Golabi P De Avila L Paik JM Srishord M Fukui N Qiu Y Burns L Afendy A Nader F The global epidemiology of NAFLD and NASH in patients with type 2 diabetes: a systematic review and meta-analysis J Hepatol 2019 71 793 801 10.1016/j.jhep.2019.06.021 31279902
Yu H-T Yu M Li C-Y Zhan Y-Q Xu W-X Li Y-H Li W Wang Z-D Ge C-H Yang X-M Specific expression and regulation of hepassocin in the liver and down-regulation of the correlation of HNF1α with decreased levels of hepassocin in human hepatocellular carcinoma J Biol Chem 2009 284 13335 13347 10.1074/jbc.M806393200 19304666
Yueh M-F He F Chen C Vu C Tripathi A Knight R Karin M Chen S Tukey RH Triclosan leads to dysregulation of the metabolic regulator FGF21 exacerbating high fat diet-induced nonalcoholic fatty liver disease Proc Natl Acad Sci USA 2020 117 31259 31266 10.1073/pnas.2017129117 33229553
Yun J Kim YS Heo MJ Kim MJ Moon A Kim SG ERα inhibits mesenchymal and amoeboidal movement of liver cancer cell via Gα12 Int J Cancer 2022 10.1002/ijc.33929 35020952
Zarei M Pizarro-Delgado J Barroso E Palomer X Vázquez-Carrera M Targeting FGF21 for the treatment of nonalcoholic steatohepatitis Trends Pharmacol Sci 2020 41 199 208 10.1016/j.tips.2019.12.005 31980251
Zhang W Cui X Gao Y Sun L Wang J Yang Y Liu X Li Y Guo X Sun D Role of pigment epithelium-derived factor (PEDF) on arsenic-induced neuronal apoptosis Chemosphere 2019 215 925 931 10.1016/j.chemosphere.2018.10.100 30408888
Zhang T Liu J Shen S Tong Q Ma X Lin L SIRT3 promotes lipophagy and chaperon-mediated autophagy to protect hepatocytes against lipotoxicity Cell Death Differ 2020 27 329 344 10.1038/s41418-019-0356-z 31160717
Zhang D Huang WJ Lan MQ Gan FR Liu YY Sun L Chen JL Sun YF Tao CM Association between serum amyloid A levels and predicting disase severity in COVID-19 patients: a systematic review and meta-analysis Eur Rev Med Pharmacol Sci 2021 25 4627 4638 10.26355/eurrev_202107_26255 34286504
Zhang Y Yan T Wang T Liu X Hamada K Sun D Sun Y Yang Y Wang J Takahashi S Crosstalk between CYP2E1 and PPARα substrates and agonists modulate adipose browning and obesity Acta Pharm Sin B 2022 12 2224 2238 10.1016/j.apsb.2022.02.004 35646522
Zhong G Kirkwood J Won K-J Tjota N Jeong H Isoherranen N Characterization of vitamin A metabolome in human livers with and without nonalcoholic fatty liver disease J Pharmacol Exp Ther 2019 370 92 103 10.1124/jpet.119.258517 31043436
Zhou W Yang J Zhu J Wang Y Wu Y Xu L Yang Y Fetuin B aggravates liver X receptor-mediated hepatic steatosis through AMPK in HepG2 cells and mice Am J Transl Res 2019 11 1498 30972177
Zhou B Lu D Wang A Cui J Zhang L Li J Fan L Wei W Liu J Sun G Endoplasmic reticulum stress promotes sorafenib resistance via miR-188-5p/hnRNPA2B1-mediated upregulation of PKM2 in hepatocellular carcinoma Mol Ther Nucleic Acids 2021 26 1051 1065 10.1016/j.omtn.2021.09.014 34786210
Zhuo L Hascall VC Kimata K Inter-α-trypsin inhibitor, a covalent protein-glycosaminoglycan-protein complex J Biol Chem 2004 279 38079 38082 10.1074/jbc.R300039200 15151994
Zinellu A Paliogiannis P Carru C Mangoni AA Serum amyloid A concentrations, COVID-19 severity and mortality: an updated systematic review and meta-analysis Int J Infect Dis 2021 105 668 674 10.1016/j.ijid.2021.03.025 33737133
Zuraw BL Hereditary angioedema N Engl J Med 2008 359 1027 1036 10.1056/NEJMcp0803977 18768946
| 36441472 | PMC9703441 | NO-CC CODE | 2022-11-29 23:21:42 | no | Arch Pharm Res. 2022 Nov 28; 45(12):938-963 | utf-8 | Arch Pharm Res | 2,022 | 10.1007/s12272-022-01419-w | oa_other |
==== Front
J Bionic Eng
J Bionic Eng
Journal of Bionic Engineering
1672-6529
2543-2141
Springer Nature Singapore Singapore
292
10.1007/s42235-022-00292-z
Research Article
Identification of Pulmonary Hypertension Animal Models Using a New Evolutionary Machine Learning Framework Based on Blood Routine Indicators
Hu Jiao [email protected]
1
Lv Shushu [email protected]
2
Zhou Tao [email protected]
3
http://orcid.org/0000-0002-7714-9693
Chen Huiling [email protected]
1
Xiao Lei [email protected]
1
Huang Xiaoying [email protected]
4
Wang Liangxing [email protected]
4
Wu Peiliang [email protected]
4
1 grid.412899.f 0000 0000 9117 1462 Department of Computer Science and Artificial Intelligence, Wenzhou University, Wenzhou, 325035 People’s Republic of China
2 grid.414373.6 0000 0004 1758 1243 Department of Dermatology, Beijing Tongren Hospital, Capital Medical University, Beijing, 100730 People’s Republic of China
3 grid.268099.c 0000 0001 0348 3990 The First Clinical College, Wenzhou Medical University, Wenzhou, 325000 People’s Republic of China
4 grid.414906.e 0000 0004 1808 0918 Department of Pulmonary and Critical Care Medicine, The First Affiliated Hospital of Wenzhou Medical University, Wenzhou, 325000 People’s Republic of China
28 11 2022
120
5 6 2022
17 10 2022
19 10 2022
© Jilin University 2022, Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Pulmonary Hypertension (PH) is a global health problem that affects about 1% of the global population. Animal models of PH play a vital role in unraveling the pathophysiological mechanisms of the disease. The present study proposes a Kernel Extreme Learning Machine (KELM) model based on an improved Whale Optimization Algorithm (WOA) for predicting PH mouse models. The experimental results showed that the selected blood indicators, including Haemoglobin (HGB), Hematocrit (HCT), Mean, Platelet Volume (MPV), Platelet distribution width (PDW), and Platelet–Large Cell Ratio (P-LCR), were essential for identifying PH mouse models using the feature selection method proposed in this paper. Remarkably, the method achieved 100.0% accuracy and 100.0% specificity in classification, demonstrating that our method has great potential to be used for evaluating and identifying mouse PH models.
Keywords
Feature selection
Pulmonary hypertension
Whale optimization algorithm
Extreme learning machine
==== Body
pmcIntroduction
The pulmonary artery is a major blood vessel that runs from the heart to the lungs. As a global health problem, Pulmonary Hypertension (PH) is estimated to affect about 1% of the global population[1]. PH is best defined by the concomitant presence of mean pulmonary arterial pressure (mPAP) > 20 mmHg, pulmonary arterial wedge pressure (PAWP) ≤ 15 mmHg, and pulmonary vascular resistance (PVR) ≥ 3 Wood, emphasizing the need for right heart catheterization with mandatory measurement of cardiac output and accurate measurement of PAWP[2]. Affecting people of all ages and all races, PH can be divided into five categories: pulmonary arterial hypertension; PH due to heart disease; PH due to lung disease; PH due to blockage of blood vessels, and PH due to other causes[3].
Transthoracic echocardiography is the most frequently employed non-invasive approach for evaluating PH. The degree of tricuspid regurgitation, myocardial performance index, presence of pericardial effusion, pulmonary vascular resistance, cardiac index, and right atrial pressure are all echocardiographic characteristics that can be employed to forecast a patient’s survival with PH [4]. However, right-sided heart catheterization is relatively more accurate for diagnosing PH[5]. PH is a disease characterized by progressive remodeling of the distal pulmonary arteries, with all three layers of the arterial wall involved. In PH, typical arterial abnormalities include intimal and neointimal fibrosis, adventitial fibrosis with varying degrees of perivascular inflammation, and medial hyperplasia of pulmonary artery smooth muscle cells. adventitial fibrosis with varying degrees of perivascular inflammation. The pathological hallmark of severe PH is complex plexiform lesions consisting of endothelial over expansion, some of which display cancer-like features of monoclonal proliferation [5, 6]. In patients with PH, pulmonary vascular resistance can lead to Right Ventricle (RV) dilation, RV dysfunction, and RV failure, resulting in increased RV wall stress[7]. To maintain right ventricular output, the RV can accommodate a slow increase in pressure load by increasing contractility and wall thickness. Although the RV dilates to maintain stroke volume when pulmonary vascular resistance continues to rise, the RV cannot remodel indefinitely. The RV eventually decouples from the pressure load, resulting in RV failure[8].
PH is usually progressive without intervention, leading to right heart failure and death[6]. PH may impair heart function, resulting in breathing difficulties with effort, fatigue, chest discomfort, swelling of the legs or abdomen, and fainting or dizzy feeling [3, 5]. Therefore, PH requires aggressive treatment. Currently, the recognized basic measures and routine PH treatment include exercise rehabilitation, oxygen therapy, anticoagulation, calcium channel blockers, diuretics, and early electric cardioversion. Pregnancy should generally be avoided, and early termination is recommended for patients with PH. Genetic counseling and psychosocial support should also be considered [6]. In patients with PH, restricted exercise is necessary. Patients should be encouraged to exercise, but only when physical conditions allow. Recent randomized controlled trials have shown that supervised rehabilitation improves exercise capacity and quality of life[9]. Excitingly, some new targeted therapy methods have also been recently proposed. Chen and his colleagues discovered that nicotinamide phosphoribosyl-transferase contributes to in pulmonary vascular remodeling, and that inhibiting it might be a viable therapy for PH [10].
Despite improved survival of patients with PH in recent years, PH still cannot be cured, and its mechanisms have not been fully elucidated [11, 12]. In the research of PH, establishing a mouse model of PH is essential. Right heart catheterization is the most accurate measure of whether PH is established in mice. However, right heart catheterization, an invasive examination that requires anesthesia and intubation of the mouse, is a blind intubation method that can easily lead to failure of pressure measurement, bleeding, and even death of the mouse. Therefore, it is crucial to determine whether a model is established non-invasively. In recent years, machine learning methods have become increasingly widely used in the medical field. Predictive models based on machine learning can be used for medical decision-making and resource allocation and help medical personnel make medical model predictions [13, 14]. In this study, we explored whether a machine learning method could be created to test the successful establishment of a mouse model of PH.
Machine learning methods have been used for the medical diagnosis of a wide range of diseases. For example, Polat et al. [15] developed a support vector machine classification algorithm for diagnosing chronic kidney disease that uses a greedy search-based classifier and a best-first search-based wrapper to evaluate a subset. Abbad et al. [16] proposed using machine learning algorithms to detect and diagnose thyroid disorders that adopt efficient classifiers. Pashaei et al. [17] proposed a chimpanzee optimization-based feature selection method for biomedical data classification wrapper. Alsaeedi et al. [18] introduced a feature selection method based on the new caledonian crow learning algorithm to identify whether a population is infected with COVID-19. Lamba et al. [19] proposed a hybrid speech signal-based Parkinson’s disease diagnosis system for early diagnosing Parkinson's disease by combining several feature selection methods and classification algorithms. Hu et al. [20] used a machine learning method based on an improved binary mutation quantum grey wolf optimizer combined with fuzzy k-nearest neighbor to predict the trend of serum albumin levels. Faisal et al. [21] proposed an ad hoc feature selection method to distinguish between Alzheimer's disease, MCI and health control patients to reduce the model complexity considered when using machine learning methods. Hu et al. [22] proposed a prediction framework based on a kernel extreme learning machine combined with improved binary Harris hawk optimization to classify the severity of COVID-19. This paper combines optimization algorithms and machine learning methods to diagnose PH.
The remaining sections of this paper are organized as follows: Sect. 2 focuses on extracting relevant data sets. The methods used for feature selection are described in Sect. 3. In Sect. 4, the experimental results are analyzed. The final section discusses the experimental results.
Materials and Methods
Laboratory Animals
107 healthy specific pathogen-free male C57BL/6 mice aged 12–14 weeks, weighing 20–25 g (animal certificate number: SYXK (zhe) 2020–0014), were purchased from the Academy of Medical Sciences of Zhejiang Province and reared at the Experimental Animal Center of Wenzhou Medical University. All animal experiments were carried out following the Institutional Animal Care guidelines and were approved ethically by the Administration Committee of Experimental Animals, Laboratory Animal Center, Wenzhou Medical University.
Reagents and Instruments
Animal oxygen chamber (Changsha Huaxi Electronic Technology Co., Ltd., Hunan, China), Power Lab4/30 multi-channel physiological instrument (ML866, Australia Ed Instruments International Trading Co., Ltd.), powerlab multi-channel bio-signal recording system, pressure transducer probe and tee, 10 mL and 1 mL syringes, 0.9% saline, heparin sodium saline (10 U/mL), 20% urethane, III-0 silk thread, NIKON SMZ745T dissecting microscope, blood cell analyzer, mouse fixation plate, medical Tape, straight forceps, curved forceps, ophthalmic scissors, arterial clips, medical cotton balls, etc.
Methods
A total of 107 mice were randomly divided into the normoxic control group (50 mice) and the hypoxic pulmonary hypertension (HPH) group (57 mice). Mice in the HPH group were kept in an oxygen chamber with an oxygen concentration of (10 ± 1)% and a CO2 concentration of (2 ± 1)% for 24 h and were reared for 21 days to induce PH. The mice in the normoxic control group were kept in an oxygen chamber with an oxygen concentration of 20.9% and a CO2 concentration of 0.03% for 24 h and were raised for 24 h. Other conditions, such as food and water, were the same. After 21 days, 0.2–0.3 ml of blood was drawn from the abdominal aorta for blood routine measurements. A blood cell analyzer analyzed blood samples to obtain blood routine data, which were then used in machine learning experiments. The study flow chart is presented in Fig. 1. Table 1 lists the 25 blood routine indicators (features) measured in this study. At the same time, all the mice were connected to the multi-channel physiology instrument through the right heart catheterization to measure the right ventricular systolic pressure and the pressure of the pulmonary artery to ensure that the mice in the HPH group successfully became PH mice.Fig. 1 Flow chart of the study
Table 1 List of the features used in this study and their abbreviations
Features Abbreviation
F1 White blood cell WBC
F2 Red blood cell RBC
F3 Haemoglobin HGB
F4 Mean corpuscular volume MCV
F5 Blood platelet PLT
F6 Neutrophil percentage NEU%
F7 Lymphocyte percentage LYM%
F8 Monocyte percentage MON%
F9 Eosinophils percentage EOS%
F10 Basophils percentage BAS%
F11 Neutrophil count NEU
F12 Lymphocyte count LYM
F13 Monocyte count MON
F14 Eosinophils count EOS
F15 Basophils count BAS
F16 Hematocrit HCT
F17 Mean corpuscular hemoglobin MCH
F18 Mean corpuscular hemoglobin concentration MCHC
F19 Red blood cell distribution width coefficient of variation RDW-CV
F20 Red blood cell distribution width-size distribution RDW-SD
F21 Mean platelet volume MPV
F22 Plateletcrit PCT
F23 Platelet distribution width PDW
F24 Platelet–large cell ratio P-LCR
F25 Platelet–large cell count P-LCC
Statistical Analysis
Comparisons between the normoxic and HPH groups were evaluated using an independent sample t test. Measurement data are expressed as mean (X¯) ± standard deviation (SD). P < 0.05 was considered a statistical significance. Analyses were performed using SPSS, version 21. Results are presented in Table 2.Table 2 Blood routine indicators in the normoxic group and HPH group
Index Normoxic group (n = 50) HPH group (n = 57) p value
F1 WBC 3.516 ± 1.526 7.585 ± 4.129 0.000
F2 RBC 7.271 ± 0.998 8.865 ± 0.51001 0.000
F3 HGB 124.140 ± 18.732 164.930 ± 12.068 0.000
F4 MCV 42.256 ± 0.992 46.698 ± 2.926 0.000
F5 PLT 280.260 ± 114.212 267.280 ± 75.172 0.496
F6 NEU% 22.176 ± 7.006 34.379 ± 5.036 0.000
F7 LYM% 62.362 ± 9.353 49.416 ± 5.726 0.000
F8 MON% 13.580 ± 2.799 13.623 ± 2.659 0.936
F9 EOS% 1.134 ± 0.602 1.781 ± 0.969 0.000
F10 BAS% 0.748 ± 0.318 0.802 ± 0.485 0.506
F11 NEU 0.770 ± 0.365 2.542 ± 1.317 0.000
F12 LYM 2.207 ± 1.081 3.812 ± 2.269 0.000
F13 MON 0.474 ± 0.212 1.021 ± 0.587 0.000
F14 EOS 0.039 ± 0.027 0.144 ± 0.132 0.000
F15 BAS 0.026 ± 0.014 0.066 ± 0.065 0.000
F16 HCT 30.760 ± 4.339 41.426 ± 3.813 0.000
F17 MCH 16.990 ± 0.747 18.591 ± 0.483 0.000
F18 MCHC 402.200 ± 17.458 399.460 ± 23.769 0.494
F19 RDW-CV 8.552 ± 0.272 9.593 ± 1.417 0.000
F20 RDW-SD 24.314 ± 0.575 26.263 ± 1.029 0.000
F21 MPV 12.688 ± 1.494 14.009 ± 1.304 0.000
F22 PCT 0.3479 ± 0.132 0.3741 ± 0.116 0.278
F23 PDW 14.622 ± 2.353 16.281 ± 0.194 0.000
F24 P-LCR 0.359 ± 0.097 0.4743 ± 0.077 0.000
F25 P-LCC 93.120 ± 30.563 127.210 ± 49.672 0.000
Presented Method
This section introduces the related knowledge of the whale optimization algorithm and the rough set theory. Then, the whale-optimized gene selection algorithm based on a rough set is further elaborated.
Mathematical Model of WOA
Many new optimization algorithms have been proposed in recent years, such as Harris hawks optimization (HHO) [23], hunger games search (HGS) [24], colony predation algorithm (CPA) [25], Runge Kutta optimizer (RUN) [26], weighted mean of vectors (INFO) [27], and slime mould algorithm (SMA) [28]. There are many fields in which they have achieved remarkable success, such as optimization of machine learning model [29], scheduling problems [30–32], medical diagnosis [33, 34], fault diagnosis [35], solar cell parameter identification [36], multi-objective problems [37, 38], combination optimization problems [39], and global optimization [40, 41].
Apart from the above, the principles of the Whale Optimization Algorithm (WOA) are derived from the modeling of humpback whale hunting behavior. Humpback whale hunting methods can usually be summarized into two behaviors. The first is sprint feeding, where when a humpback whale finds prey, it will dash directly towards the prey and open its mouth to swallow it directly. The second way is to rise upwards in a spiral-like position at a distance of 15 m from the surface, spitting out bubbles of different sizes, as they arise in order for all the bubbles to reach the surface simultaneously. The humpback whales are surrounded by these bubbles and swallow the prey when they reach the surface. This section details the basic whale algorithm based on the mathematical models of these two predatory behaviors.
Spiral Position Update
The mathematical model for the spiral period is shown in Eqs. (1) and (2):1 X(t+1)→=D′→·ebl·cos(2πl)+X∗t,→
2 D′→=|X∗t→-Xt→|,
where vector X(t)→ is the current position of the whale, vector X(t+1)→ is the position of the whale after iteration, and vector X∗t→ is the best position. b is a constant that controls the shape of the logarithmic spiral. l is a random number between [− 1,1]. When l reaches 1, the whale is farthest from the optimal position; when l reaches − 1, the whale is closest to the optimal position. '∙ ' denotes point multiplication.
Reduce the Encirclement
3 X(t+1)→=X∗t→-A→·D→,
4 D′→=|C→·X∗t→-Xt→|,
5 A→=2a·r1→-a,
6 C→=2r2→,
7 a=2-2tTmax,
where r1→, r2→, are random numbers between (0, 1), vector A→ and vector C→ are used to control the position of the whale in the process of position update. a is an important variable in the WOA, because a affects vector A→ by indirectly controlling the mode of travel of whale. As shown in Eq. (7), a linearly decreases from 2 to 0. Tmax is the maximum number of iterations.
The above sections combined are the previously described bubble net attacks. A random number is used to select the method of the bubble net attack, and the probability of this random number is usually set to p. The probability of p is usually set to 0.5.
Depending on the parameter p, the bubble net attack can be made in different ways. They are as follows:8 Xt+1→=X∗t→-A→·D→,rand<pD′→·ebl·cos2πl+X∗t→,rand≥p
Random Search
The random search process evolves into a mathematical model as shown in Eq. (9):9 X(t+1)→=Xrand→-A→·Drand→
10 Drand→=|C→·Xrand→-X(t)→|
Xrand→ is a randomly selected individual from a population of whales.
Which to use, the bubble spiral attack way or the random search way, is determined by the vector A→ of Eq. (3.5). When A→ ≥ 1, the position of the WOA is updated in the random search way, and when A→ < l, the position of the WOA is updated in the bubble spiral attack way.
The simplified pseudo-code of WOA is listed in Algorithm 1.
Whale Optimization Algorithm with Hybrid Conversion Mechanism (HCWOA)
In this section, a hybrid conversion mechanism inspired by TRIZ (Teoriya Resheniya Izobreatatelskikh Zadatch) creative solution is introduced and combined with the basic WOA to maintain population diversity in the search process to explore more efficient gene interactions, resulting in new individuals.
TRIZ [42] is a knowledge-based, human-oriented systematic methodology for inventive problem solving. The TRIZ theory itself is based on the practice of decomposing systems into subsystems, distinguishing between useful and harmful functions, and these decompositions are problem and context-dependent and inherently stochastic. Furthermore, creatively solving problems is central to the innovation process. Although there are several theories and methods, the standard procedure for dealing with them is to use randomized trial and error. The findings of TRIZ and evolutionary algorithms support the idea that creativity can be systematically understood and developed.
One of the TRIZ tools is the 40 invention principles, which consist of a set of generic solutions that solve technical contradictions in many fields. The principles are organized according to the contradictions they solve, which makes it easy to deal with problems. In recent years, the TRIZ theory has been used to solve different problems in various industries. For example, Paolo et al. [43] proposed an innovation management framework to rely on partners with TRIZ skills to coordinate customized innovation processes. Vladimir et al. [44] analyzed several fashion inventions and apparel production techniques from a TRIZ perspective. Liu et al. [45] explored the impact of TRIZ learning on graduate students and showed that participants with TRIZ learning experience produced more novel design solutions and demonstrated the positive impact of TRIZ learning on bio-inspired designs. Khadija et al. [46] used a TRIZ contradiction matrix to address the main issues that arose during TRIZ matrix development. Christian et al. [47] improved the sustainability of different devices to a great extent by reducing environmental impact through TRIZ strategies.
In this paper, a Hybrid Conversion (HC) mechanism derived from TRIZ theory is used to help expand the advantages of each stage of WOA by addressing the shortcomings of the three different stages: spiral position update, narrowing envelope, and random search. The first operation of HC is crossover behavior. In whale populations, individuals with little difference in fitness tend to cluster together, and various types of crossover variation will be in this small population as the population evolves. However, this situation can make the population evolve to a limited extent and easily fall into local optima in the global optimization of complex problems. To alleviate this situation, we propose a crossover behavior. First, the whole whale population is divided into K groups based on time fitness values, and then individuals in each group are fused with individuals in other groups to form a new population. The specific process is shown in Fig. 2.Fig. 2 Schematic illustration of crossover operation
The second operation in HC is local optimization, as shown in Fig. 3. This operation is divided into three main operators: dynamic partitioning, mutation, and transition. These three operators are described in detail below.Fig. 3 Schematic illustration of individual changes
Dynamic Partitioning Operator
In this operator, the optimal individual in the population is invariant, and the other individuals are decomposed into smaller blocks. The specific size S of the block is generated dynamically by the algorithm. If the individual dimension is not divisible by S, then the last piece is used as a reminder of the size of the block. In particular, if the remainder is 1, then it is merged with the previous block. This setting is mainly used to evaluate individual dimensions to find the more important features for classification during the feature selection process.
Mutation Operator
This operator is roughly the same as the commonly used mutation operator. The specific operation is to subdivide each individual block into two groups. One of the two groups will remain unchanged, and the other will perform a mutation operation. The specific mutation is to generate a random number between 0 and 1 randomly. If this random number is less than 0.5, a Gaussian mutation operation is performed, and if it is greater than 0.5, a Cauchy mutation operation is performed, and then the two groups are reunited into one block. The mutation operation during the feature selection operation means that the original 0 is mutated to 1, and 1 is mutated to 0. For example, 0010101 to 1101010.
Transition Operator
This is an inter-block operation. In this operator, two random block indices are randomly generated with random numbers between 0 and K. K indicates the total number of blocks each individual is divided into. For the two selected blocks, the positions of the two blocks are swapped in the individual. The blocks are then re-linked into a new individual.
Classification Based on KELM
Extreme Learning Machine (ELM) is a special variant of the fast single hidden layer feedforward neural algorithm. Huang et al. [48] introduced regularization schemes and kernel functions into ELM to obtain the kernel extreme learning machine (KELM). KELM can improve the prediction performance of the model while retaining the advantages of ELM. Among them, ELM is a single hidden layer feedforward neural network whose learning objective function F(x) can be represented by the matrix:11 Fx=hx×β=H×β=L,
where x is the input vector, h(x) and H are the output of the hidden layer node, β is the output weight, and L is the desired output.
Network training can be turned into a problem solved by a linear system. β can be determined from β=H∗·L, where H∗ is the generalized inverse matrix of H. In addition, to enhance the stability of the neural network, the regularization coefficient C and the unit matrix I are introduced, and the least-squares solution of the output weights is as follows:12 β=HTHHT+IC-1L.
The kernel function is introduced into ELM, and the kernel matrix is13 ΩELM=HHT=hxihxj=Kxi,xj,
where xi and xj are the test input vectors, then Eq. (3.11) can be expressed as14 Fx=Kx,x1;⋯;Kx,xnIC+ΩELM-1L,
where (x1,x2,⋯,xn) is the given training sample, n is the number of samples, and K() is the kernel function.
In this paper, the routine blood data of pulmonary hypertension were classified, and the training set and test set were generated by tenfold cross-validation. The regularization coefficient C and the kernel function parameter S were, respectively, obtained after HCWOA optimization, and the kernel function was selected as the RBF Gaussian kernel function.
To better classify pulmonary hypertension blood routine data, HCWOA and KELM are combined for feature selection. The method can provide interpretable results for classification and help physicians in medical diagnosis. The flow chart of the method in this paper is as follows and the flowchart is shown in Fig. 4.Fig. 4 Flowchart of the HCWOA–KELM
Step 1: Initialize the input parameters of HCWOA, including the population size, the boundary of the search space, the maximum number of iterations, and the dynamic block S.
Step 2: Initialize a random binary whale population.
Step 3: Obtain a subset of features (1: features are selected, 0: features are not selected) according to the whales' locations.
Step 4: Use the selected feature subset to calculate the fitness of each whale as follows:15 Fitness=α·E+β·Rd,
where E denotes the classification error rate of KELM, d denotes the number of selected feature subsets, and R denotes the total number of selected features. In addition, α and β are two weights that measure the importance of the classification error rate and the size of the selected feature subset. we set α = 0.99 and β = 1 − α.
Step 5: Select the whale with the smallest fitness as the optimal position.
Step 6: Update the control parameters a, A and C according to Eqs. (5–7)
Step 7: Update the whale population location.
Step 8: Perform the hybrid conversion mechanism for the whales other than the optimal position to obtain a new population.
Step 9: Select the optimal one to reconstitute the new population according to the greedy idea.
Step 10: Judge whether the end condition is reached, if not, repeat steps 4–9.
Step 11: Return to the optimal position.
Experiment and Results
Validation of Function Optimization
In this paper, to test the performance of the proposed HCWOA, 28 benchmark functions were used as test sets, including 4 unimodal functions (F1–F4), 6 multimodal functions (F5–F10), 4 fixed modal functions (F10–F15), 6 hybrid functions (F15–F20) and 8 composition functions (F21–F28). Table 3 lists the descriptions of the 28 benchmark functions. These benchmark functions represent a variety of the most complex mathematical optimization problems. Therefore, they are often used to evaluate the comprehensive ability of algorithms. The average performance of all compared algorithms was further compared statistically using the Freidman test, and the average rank was given according to comparison results.Table 3 Descriptions of test functions
F1 f1x=∑i=1nxi2 30 [− 100, 100] 0
F2 f2x=∑i=1nxi+∏i=1nxi 30 [− 10, 10] 0
F3 f3x=∑i=1n(∑j-1ixj)2 30 [− 100, 100] 0
F4 f4x=maxixi,1≤i≤n 30 [− 100, 100] 0
F5 f8x=∑i=1n-xi·sin(|xi|) 30 [− 500, 500] − 418.982
F6 f9x=∑i=1nxi2-10cos2πxi+10 30 [− 5.12, 5.12] 0
F7 f10x=-20exp-0.21n∑i=1nxi-exp1n∑i=1n2πxi+e 30 [− 32, 32] 0
F8 f11x=14000∑i=1nxi2-∏i=1ncosxii+1 30 [− 600, 600] 0
F9 f12x=πn{10sinay1+∑i=1n-1(yi-1)21+10sin2πyi+1+(yn-1)2+∑i=1nμ(xi,10,100,4)} 30 [− 50,50] 0
F10 f13x=0.1{sin23πxi+∑i=1n(xi-1)21+sin23πxi+1+[1+sin22πxn]∑i=1nμ(xi,5,100,4)} 30 [− 50, 50] 0
F11 f15x=∑i=111ai-x1(bi2+bix2)bi2+bix3+x42 4 [− 5, 5] 0.00030
F12 f21x=-∑i=15X-aiX-aiT+ci-1 4 [0, 10] − 10.1532
F13 f22x=-∑i=17X-aiX-aiT+ci-1 4 [0, 10] − 10.4028
F14 f23x=-∑i=110X-aiX-aiT+ci-1 4 [0, 10] − 10.5363
F15 Shifted and Rotated Katsuura Function Hybrid 1200
F16 Shifted and Rotated HappyCat Function Hybrid 1300
F17 Shifted and Rotated HGBat Function Hybrid 1400
F18 Shifted and Rotated Expanded Griewank’s plus Rosenbrock’s Function Hybrid 1500
F19 Shifted and Rotated Expanded Scaffer’s F6 Function Hybrid 1600
F20 Hybrid Function 1 (N = 3) Hybrid 1700
F21 Composition Function 1 (N = 5) Composition 2300
F22 Composition Function 2 (N = 3) Composition 2400
F23 Composition Function 3 (N = 3) Composition 2500
F24 Composition Function 4 (N = 5) Composition 2600
F25 Composition Function 5 (N = 5) Composition 2700
F26 Composition Function 6 (N = 5) Composition 2800
F27 Composition Function 7 (N = 3) Composition 2900
F28 Composition Function 8 (N = 3) Composition 3000
The comparison algorithm adopted includes original WOA, common meta-heuristic algorithm: Bat Algorithm (BA) [49], Firefly Algorithm (FA), Moth-Flame Optimization (MFO) [50], Differential Evolution (DE) [51], and improved advanced algorithm: differential Evolution algorithm based on Chaotic Local Search (DECLS) [52], Chaotic and Gaussian Particle Swarm Optimization (CGPSO) [53], Particle Swarm Optimization with Aging Leader and Challengers (ALCPSO) [54].
These tests were conducted out using a Microsoft Windows Server 2012 R2 data center version of windows with 128 GB RAM and an Intel (R) Xeon (R) E5-2650 v4 (2.20 GHz) CPU, with code written in MATLAB R2014b.
All techniques were evaluated under the identical conditions to ensure a fair assessment. The population size was set to 30 and the maximum number of evaluations was set to 1500. To reduce other effects, all algorithms were tested on the benchmark functions for 30 times. In this paper, the numerical results of these methods were selected based on the average (Avg.) and standard deviation (Std.) of the optimal function values. Avg. is used to evaluate the global search ability, and Std. is used to evaluate the algorithm's robustness. In addition, the optimal results for each problem are marked in bold to display the optimal results clearly. The Wilcoxon signed-rank test [55], a nonparametric statistical test at the 0.05 significance level, was used to determine whether the improvement was statistically significant. The symbols” + / = /” indicate that the proposed method is better than, equal to and worse than the other competitors, respectively.
Table 4 shows the optimization results of HCWOA and 8 competitors on 28 benchmark functions. The bolded values in the table indicate the best optimization on that test function. It can be seen that HCWOA achieves optimal results on 24 functions. In particular, HCWOA can directly find the optimal values on the unimodal functions. On the multimodal functions (F5–F10), HCWOA reaches the optimum on some functions. Although HCWOA does not optimize as well as FA and CGPSO on F5, F9 and F10, there is only a small difference between the results obtained by HCWOA and the minimum values. In addition, for the fixed modal functions, although HCWOA's optimization result on F17 is worse than BA, HCWOA's optimization results are optimal on the other functions. HCWOA always has the best optimization results for other test functions. This is because the operation for populations and individuals in HCWOA can help avoid local optima.Table 4 Comparison of optimal values between HCWOA and famous MAs
Item HCWOA WOA BA FA MFO DE DECLS CGPSO ALCPSO
F1 Avg 0.00E+00 1.37E−74 2.18E+00 1.52E+04 1.69E+03 4.58E−04 1.41E−03 1.03E−03 1.09E−05
Std 0.00E+00 4.98E−74 1.01E+00 1.50E+03 5.30E+03 2.25E−04 2.71E−03 2.32E−03 1.27E−05
F2 Avg 0.00E+00 9.34E−51 8.05E+01 6.39E+01 3.91E+01 2.20E−03 1.16E−02 1.22E−02 2.87E−03
Std 0.00E+00 2.89E−50 3.55E+02 7.97E+00 1.98E+01 4.53E−04 1.38E−02 1.21E−02 1.06E−02
F3 Avg 0.00E+00 4.56E+04 2.62E+01 2.76E+04 1.97E+04 3.11E+04 1.62E−01 4.25E+00 2.62E+03
Std 0.00E+00 1.54E+04 1.30E+01 3.71E+03 1.22E+04 5.06E+03 2.16E−01 4.66E+00 1.18E+03
F4 Avg 0.00E+00 4.43E+01 1.12E+01 4.83E+01 6.91E+01 1.32E+01 4.02E−03 4.17E−03 1.29E+01
Std 0.00E+00 2.86E+01 5.87E+00 1.94E+00 8.18E+00 1.74E+00 3.59E−03 4.75E−03 3.18E+00
F5 Avg −1.07E+04 −1.10E+04 −7.42E+03 − 3.55E+03 − 8.73E+03 − 9.76E+03 −1.26E+04 −2.30E+04 −1.03E+04
Std 1.43E+03 1.57E+03 7.06E+02 3.17E+02 9.73E+02 5.03E+02 7.90E−04 3.89E+03 6.32E+02
F6 Avg 0.00E+00 1.89E−15 2.85E+02 2.60E+02 1.70E+02 8.78E+01 3.13E−04 2.05E−04 6.79E+01
Std 0.00E+00 1.04E−14 3.04E+01 1.53E++01 3.61E+01 8.36E+00 7.60E−04 3.38E−04 2.17E+01
F7 Avg 8.88E−16 3.85E−15 4.25E+00 1.69E+01 1.71E+01 5.84E−03 4.52E−03 3.20E−03 1.39E+00
Std 0.00E+00 3.24E−15 3.00E+00 4.43E−01 5.58E+00 1.21E−03 4.68E−03 3.01E−03 9.48E−01
F8 Avg 0.00E+00 4.04E−03 2.98E+00 1.35E+02 1.90E+01 6.80E−03 3.26E−03 1.50E−03 2.59E−02
Std 0.00E+00 2.21E−02 6.04E+00 1.59E+01 4.36E+01 8.28E−03 6.26E−03 3.28E−03 4.15E−02
F9 Avg 2.02E−03 4.65E−02 1.29E+01 7.67E+06 8.24E+00 4.92E−05 8.86E−06 4.61E−06 5.33E−01
Std 1.17E−03 1.25E−01 7.78E+00 3.64E+06 6.36E+00 2.75E−05 2.09E−05 1.01E−05 4.83E−01
F10 Avg 6.45E−02 5.30E−01 7.45E−01 3.60E+07 9.29E+02 2.84E−04 1.43E−04 5.44E–05 4.48E−02
Std 6.25E−02 2.72E−01 2.74E−01 1.41E+07 4.65E+03 1.22E−04 3.16E−04 1.57E−04 6.11E−02
F11 Avg 5.24E−04 6.26E−04 6.89E−03 2.40E−03 1.03E−03 9.70E−04 8.04E−04 1.41E−03 1.76E−03
Std 3.93E−04 3.47E−04 8.99E−03 9.04E−04 4.50E−04 1.42E−03 2.97E−04 3.72E−04 4.16E−03
F12 Avg −8.01E+00 −7.16E+00 −4.76E+00 −5.11E+00 −5.97E+00 −9.97E+00 −1.02E+01 −1.02E+01 −5.48E+00
Std 2.88E+00 3.27E+00 2.18E+00 1.20E+00 3.37E+00 9.22E−01 4.06E−05 9.70E−06 3.26E+00
F13 Avg −6.94E+00 −6.70E+00 −5.02E+00 −5.67E+00 −7.19E+00 −1.02E+01 −1.04E+01 −1.04E+01 −6.26E+00
Std 3.63E+00 3.19E+00 2.31E+00 1.30E+00 3.54E+00 9.62E−01 7.28E−05 9.35E−05 3.74E+00
F14 Avg −7.19E+00 −7.13E+00 −6.23E+00 −5.42E++00 −8.02E+00 −1.03E+01 −1.05E+01 −1.05E+01 −6.19E++00
Std 3.26E+00 3.59E+00 2.06E+00 1.07E+00 3.40E+00 1.22E+00 7.30E−05 5.05E−05 3.67E+00
F15 Avg 2.39E+05 7.86E+08 6.15E+07 3.94E+09 5.47E+08 3.69E+07 1.24E+08 1.62E+08 4.14E+08
Std 2.53E+05 4.07E+08 3.38E+07 8.65E+08 5.07E+08 1.42E+07 4.18E+07 5.37E+07 8.56E+08
F16 Avg 5.56E+03 4.34E+07 3.71E+05 9.20E+08 5.34E+07 1.95E+03 1.02E+04 6.98E+06 3.32E+06
Std 2.93E+03 7.54E+07 1.03E+05 2.34E+08 1.63E+08 3.98E+02 3.30E+03 2.43E+06 4.95E+06
F17 Avg 2.85E+05 7.60E+05 9.42E+04 1.34E+06 2.20E+05 7.64E+04 2.72E+05 1.17E+05 4.47E+04
Std 2.11E+05 5.24E+05 8.06E+04 4.98E+05 4.08E+05 7.83E+04 3.08E+05 7.23E+04 2.11E+05
F18 Avg 3.62E+04 1.73E+05 1.47E+05 1.99E+08 3.86E+04 4.91E+03 4.42E+04 1.12E+06 4.96E+03
Std 7.36E+04 1.80E+05 5.68E++04 8.47E+07 4.02E+04 2.31E+03 2.09E+04 5.71E+05 3.30E+03
F19 Avg 2.55E+03 3.91E+03 3.61E+03 3.68E+03 3.06E+03 2.89E+03 3.21E+03 3.10E+03 2.91E+03
Std 3.66E+02 4.99E+02 4.79E+02 1.09E+02 4.19E+02 1.92E+02 2.06E+02 3.77E+02 2.93E+02
F20 Avg 2.17E+03 2.84E+03 2.89E+03 2.89E+03 2.43E+03 2.39E++03 2.55E+03 2.53E+03 2.48E+03
Std 2.70E+02 2.87E+02 3.15E+02 1.25E+02 2.17E+02 1.12E++02 1.16E+02 2.25E+02 2.15E+02
F21 Avg 2.50E++03 3.48E+03 3.73E+03 3.15E+03 2.97E+03 2.96E+03 2.53E+03 2.50E+03 3.11E+03
Std 0.00E+00 2.50E+02 4.00E+02 2.41E+01 3.04E+01 1.19E+01 1.22E+02 2.93E−02 1.45E+02
F22 Avg 2.60E+03 2.74E+03 3.18E+03 3.77E+03 3.51E+03 3.48E+03 2.60E+03 2.60E+03 3.25E+03
Std 0.00E+00 4.16E+02 7.12E+02 2.78E+01 4.64E+01 8.72E+00 1.50E−01 1.49E−01 3.57E+02
F23 Avg 2.70E+03 2.74E+03 3.02E+03 4.65E+03 3.63E+03 3.02E+03 2.70E+03 2.70E+03 2.99E+03
Std 0.00E+00 1.79E+02 7.56E+01 2.38E+02 9.65E+02 4.20E+01 1.09E+00 9.92E−01 6.05E+01
F24 Avg 2.80E+03 3.43E+03 5.88E+03 7.74E+03 6.51E+03 6.31E+03 2.80E++03 2.80E+03 5.22E+03
Std 0.00E+00 1.92E+03 3.77E+03 1.84E+02 4.80E+02 1.78E+02 1.43E+00 1.35E+00 1.35E+03
F25 Avg 2.90E+03 4.10E+03 3.98E+03 4.16E+03 3.60E+03 3.48E+03 2.99E+03 3.10E+03 3.87E+03
Std 0.00E+00 2.34E+02 2.24E+02 1.54E+02 8.22E+01 3.35E+01 2.34E+02 7.68E+02 1.74E+02
F26 Avg 3.00E+03 3.54E+03 3.82E+03 4.55E+03 5.22E+03 4.56E+03 3.00E+03 3.00E+03 3.42E+03
Std 0.00E+00 8.75E+02 1.14E+03 2.84E+02 3.87E+02 6.85E+02 2.23E+00 1.16E+00 5.00E+02
F27 Avg 3.10E+03 4.69E+03 4.71E+03 5.00E+03 4.07E+03 4.18E+03 3.10E+03 3.14E+03 3.84E+03
Std 0.00E+00 5.90E+02 3.48E+02 2.06E+02 2.24E+02 1.84E+02 2.85E+00 1.77E+02 2.24E+02
F28 Avg 3.20E+03 2.07E+07 4.17E+06 2.60E+08 2.19E+06 2.38E+06 2.58E+05 2.89E+05 8.55E+04
Std 0.00E+00 3.07E+07 2.61E+06 8.21E+07 3.71E+06 9.75E+05 2.52E+05 2.84E+05 1.03E+05
± / = ~ 18/0/10 24/3/1 27/0/1 21/0/7 16/9/3 18/8/2 19/8/1 17/6/5
Avg 2.535 5.536 6.786 8.464 6.393 4.321 3.124 3.321 4.429
Rank 1 6 8 9 7 4 2 3 5
The bolded values in the table indicate the best optimization on that test function
Table 5 shows the p value of the Wilcoxon signed-rank test. p value less than 0.05 indicates statistical significance. Values less than 0.05 are bolded in the table, indicating that HCWOA is significantly better than the comparison method. As can be seen from the table, HCWOA significantly outperforms BA, FA, and DE on the 28 benchmark functions. Although on F5–F6 and F11–F14, the optimization results of HCWOA and WOA are not significantly different, it can be seen that more than 95% of the data in the table are less than 0.05. Therefore, it can be said that HCWOA has better optimization results compared with the other 8 algorithms.Table 5 P value of the Wilcoxon signed-rank test between HCWOA and other MAs
WOA BA FA MFO DE DECLS CGPSO ALCPSO
F1 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173
F2 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173
F3 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173
F4 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173
F5 0.49000000 0.00000173 0.00000173 0.00004070 0.00385000 0.00000173 0.00000173 0.26200000
F6 1.00000000 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173
F7 0.00009600 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173
F8 1.00000000 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173
F9 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000472
F10 0.00000192 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.15800000
F11 0.22800000 0.00000472 0.00000173 0.00017400 0.11000000 0.01560000 0.00000521 0.05710000
F12 0.42800000 0.00004860 0.00022200 0.15200000 0.00018900 0.00000173 0.00000173 0.02840000
F13 0.92600000 0.01750000 0.06260000 0.57100000 0.00001970 0.00000173 0.00000173 0.70300000
F14 0.95900000 0.17100000 0.01850000 0.23600000 0.00006320 0.00000173 0.00000173 0.25300000
F15 0.00008920 0.00000173 0.00000192 0.02840000 0.00000173 0.00002840 0.00196000 0.00000173
F16 0.00000522 0.00003110 0.00000173 0.00468000 0.00000173 0.00000173 0.00103000 0.00000173
F17 0.00018900 0.00002370 0.00000212 0.05190000 0.00018900 0.31800000 0.00089400 0.00003110
F18 0.00001240 0.00003110 0.00000173 0.21300000 0.00000212 0.00170000 0.00000173 0.00000260
F19 0.00000388 0.00000769 0.00000212 0.12000000 0.74900000 0.00083000 0.02700000 0.00057000
F20 0.00002590 0.00007510 0.00000575 0.50300000 0.14700000 0.37000000 0.44000000 0.00020500
F21 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173
F22 0.25000000 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173
F23 0.25000000 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173
F24 0.25000000 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173
F25 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173
F26 0.00390000 0.00000173 0.00000173 0.00000172 0.00000173 0.00000173 0.00000173 0.00000173
F27 0.00000256 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173
F28 0.00000379 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173 0.00000173
Bold value indicating that HCWOA is significantly better than the comparison method
Figure 5 shows the convergence effect of HCWOA on unimodal, multimodal, fixed modal, hybrid, and composition functions. The convergence curves were obtained from the evaluation process by selecting 50 points in sequence and plotting them as smooth curves. It can be seen from the figure that HCWOA has not only the smallest convergence but also the fastest convergence on the unimodal and multimodal functions. On the fixed modal function, although the convergence speed of HCWOA is not the fastest, the convergence value is the smallest. For F21 and F27, although the convergence speed of HCWOA is similar to those of DECLS and CGPSO, the convergence value of HCWOA is the smallest. As can be seen from the convergence figure, the population diversity and convergence accuracy of HCWOA are greatly improved compared to the basic WOA.Fig. 5 Convergence curves of HCWOA and other MAs on 9 functions
Feature Selection Experiment in the PH Data Set
To verify the effectiveness of the proposed HCWOA–KELM method, bHCWOA–KELM was compared with a range of machine learning methods, including BP, CART, Random Forest, AdaBoostM1, ELM and KELM traditional classification algorithms. Among them, the BP algorithm, CART, RandomF and AdaBoostM1 used self-built classifiers in MATLAB. The ELM algorithm can be found at http://www.ntu.edu.sg/eee/icis/cv/egbhuang.htm. KELM is based on ELM with the addition of kernel functions. For a fair comparison, all experiments were performed in the same simulated environment. The number of hidden neurons was 10 for the BP algorithm and 20 for the ELM algorithm. The regularization factor C and the kernel function parameter S for KELM were set to 2 and 4, respectively. The settings for the remaining classifiers were specified by convention. To get a fair and impartial outcome, the classification performance was assessed employing tenfold cross-validation (CV) methodology. In addition, to evaluate the performance of bHCWOA–KELM, we used four popular metrics, i.e., specificity, sensitivity, classification accuracy (ACC), and MCC.
The classifier's effectiveness is verified using four mutual rules based on the confusion matrix. The complete definitions of these metrics are provided in [56, 57]. Here, we present their formulations to avoid discussions beyond the scope of this study:16 Accuracy=TP+TNTP+TN+FP+FN,
17 Specificity=TNFP+TN,
18 Sensitivity=TPTP+FN,
19 MCC=TP×TN-FP×FN(TP+FP)×(TP+FN)×(TN+FP)×(TN+FN).
The specific experimental results are as follows.
To verify its effectiveness in feature selection, the proposed bHCWOA–KELM algorithm was compared with commonly used classification methods without feature selection. The comparison results of the seven classifiers are shown in Fig. 6. It can be seen from the figure that, although the differences between bHCWOA–KELM and AdaBoost in terms of specificity and classification accuracy are small, bHCWOA–KELM is much better than the other algorithms in terms of sensitivity and MCC. The experimental results show that bHCWOA–KELM can handle the classification of high-pressure pulmonary artery blood routine data, suggesting that it can provide some help for experts and doctors in real life.Fig. 6 Comparison effect of bHCWOA on 6 classifiers
From the above, it is clear that bHCWOA–KELM is significantly more effective in feature selection than the common classification methods. To further evaluate the effectiveness of this method on the routine high-pressure pulmonary artery blood data set, bHCWOA–KELM was compared with the commonly used bMFO, BPSO, BSSA, BBA, and bWOA in combination with KELM. The convergence curves of these six algorithms are shown in Fig. 7. The convergence values are the fitness set in Sect. 3 of this paper. As shown in the figure, the effectiveness of this algorithm is measured in terms of both convergence value and convergence speed. From the figure, it can be seen that bHCWOA–KELM has not only the fastest convergence speed but also the smallest convergence value. These results demonstrate that bHCWOA–KELM can effectively establish a proper balance between classification accuracy and feature subset size when dealing with the feature selection problem.Fig. 7 Convergence evolution trends of the six methods
Finally, to further help physicians in the analysis of the classification results, the specific selected features from the tenfold cross-validation results were analyzed. The specifically selected features in each experiment are shown in Fig. 8. As can be seen from the figure, the most frequently selected features are HGB, HCT, MPV, PDW and p-LCR. Further analysis shows that these five features are the most important for the final classification results. Therefore, bHCWOA–KELM can provide accurate classification results and help physicians analyze which features are the key factors in routine blood data.Fig. 8 Selected features by the bHCWOA-KELM during the10 times tenfold CV procedure
Discussion
The Analysis Method and Parameter Selection
The optimization properties of the HC operator based on TRIZ theory, HCWOA can obtain better global optimal solutions with better feature selection ability, as shown by the experimental results. The dynamic segmentation operator, variation, and transposition operations can effectively partition and optimize the structure of the search space. This can improve the performance of the whole population from the individual whales by gene exchange and global optimization. The three stages of WOA, namely, spiral position update, narrowing envelope, and random search, helped to expand each stage’s advantages by borrowing the hybrid transformation mechanism distilled from TRIZ theory. This is because TRIZ theory itself is based on the practice of decomposing systems into subsystems, distinguishing useful and harmful functions, and these decompositions depend on the problem and the environment and are somewhat stochastic in nature.
In addition, to verify the practical application capability of HCWOA, we apply its discretization and KELM in combination with the PH data set. From the experimental results, it can be seen that such a combination not only enhances the performance of KELM but also shows excellent performance in the feature selection problem. In addition, bHCWOA–KELM can accurately select the features in the PH data set that plays a deterministic role in the classification effect.
Physiological Significance
In this study, machine learning was used to identify whether or not a mouse model of PH was successfully established by analyzing blood samples. Several key features were identified, including HGB, HCT, MPV, PDW and P-LCR.
It is well-known that normal HGB is a tetramer composed of two α-like polypeptide subunits and two β-like polypeptide subunits[58]. HGB's primary purpose is to carry oxygen from the lungs to peripheral tissues and carbon dioxide from peripheral tissues to the lungs for excretion, therefore, controlling the body’s blood oxygen balance [59]. Therefore, in an oxygen-deficient environment, HGB is bound to change. Anna Hauser et al. reported increased HGB content in hypoxic environments[60]. Steven Deem et al. study found that oxy-HGB rapidly oxidizes NO to nitrate and methemoglobin, resulting in a greatly weakened duration and magnitude of the vasodilation effect of NO in the pulmonary circulation, thus manifesting as increased pulmonary vascular resistance[61, 62]. Meanwhile, cell-free hemoglobin as a pro-inflammatory oxidant [63]was found to cause not only acute lung injury when cell-free hemoglobin is present in the trachea of mice, but also airspace inflammation and alveolar capillary barrier damage[64, 65]. Based on these studies, it can be seen that hypoxia can lead to increased HGB, while elevated HGB damages alveoli and increases pulmonary vascular resistance. This study found that the HGB of mice in the HPH group was 1.33 times higher than that in the control group (p = 0.000), suggesting that HGB may be a promising predictor for evaluating models of PH in mice.
HCT refers to the volume fraction of red blood cells in the blood and serves as one of the most important indicators of a patient's blood status. The content of HCT is of great significance to diagnosing many diseases, such as inflammation, anemia, polycythemia, and blood stimulants [66]. The content of HCT depends on the level of erythropoietin (EPO) produced by the kidneys. The partial pressure of oxygen can regulate EPO. Under hypoxic conditions, EPO will increase, resulting in increased HCT [67]. In addition, many studies have shown that, within a certain range, the ability of the circulatory system to deliver oxygen increases with the elevation of HCT [68]. Numerous epidemiological data suggest that there is a close relationship between HCT and cardiovascular disease [69]. Similar to these studies, Stauffer E and Beall CM et al. found that in healthy individuals, prolonged hypoxic conditions lead to elevated HCT [70, 71]. In line with these findings, our research confirmed that the HCT level of mice in the HPH group was higher than that of the control group (p = 0.000) (1.35 times), and the HCT level may be effective in assessing PH models in mice.
MPV is defined as the mean volume of platelets in the blood [72]. Under physiological conditions, MPV is inversely proportional to the number of platelets to maintain normal platelet function, which means that an increase in MPV will simultaneously lead to a decrease in platelets [73]. Studies have found that in the presence of hypoxia, the destruction of platelets is increased, and at the same time, MPV is increased [74]. MPV may also be a predictive marker of cardiovascular events, closely related to thrombosis and inflammation [75, 76]. Tromsø et al. found that increased MPV was a predictor of venous thromboembolism (VTE) [77]. The related meta-analysis found that the MPV of VTE patients was significantly higher than that of the control group [78]. Steiropoulos et al. found that the MPV of COPD patients was significantly higher than that of healthy people [79]. Numerous studies have also shown that severe hypertensive disease and related target organ damage are associated with elevated MPV [80–82]. Similar to the findings of their study, we also found that the MPV of mice in the HPH group was higher than that in the control group, about 1.10 times that of the control group. Therefore, MPV might have a significant predictive value in distinguishing mice with PH.
Similar to MPV, PDW is also a volume parameter, meaning the distribution of platelet size, which is the relative height of the 20% platelet size distribution histogram [83, 84]. Increased PDW means uneven platelet volume, suggesting platelet activation [85]. PDW was positively correlated with mean pulmonary artery pressure [86]. Some possible mechanisms are: activated platelets play a key role in pulmonary vascular remodeling and thrombosis [87, 88]. Sonali Jindal et al. also found a positive correlation between PDW and microvascular complications [89]. Multiple studies have shown that PDW in patients with idiopathic pulmonary hypertension is significantly higher than that in controls [90]. Consistent with these findings, our experimental results also showed that the PDW of mice in the HPH group was 1.11 times that of the control group (p = 0.000).
P-LCR, an indicator used to quantify large platelets, is the proportion of platelets larger than 12 fl in the total number of platelets [91]. Large platelets are hypothesized to be more active [92, 93]. Numerous studies have shown that large platelets contain more particles and receptors and have higher hemostasis, thrombosis, and pro-inflammatory abilities than small platelets [94]. In relevant animal experiments, it was found that after the consumption of platelets, the newly generated platelets were significantly larger than the previous platelets [95]. The same is true after chemotherapy in humans [96]. In our experiment, we found that the P-LCR of mice in the HPH group was higher than that in the control group, indicating it could potentially serve as a valuable predictor (p = 0.000).
However, because our experiments may have many limitations, these results require more rigorous interpretation and in-depth thinking. First, our experimental results partially depend on our experimental apparatus, resulting in possible deviations in the results. Second, the number of factors we selected was still too small. In further research, we should expand the research scope to blood gas analysis, coagulation mechanism, inflammatory factors, etc. Third, the sample size we studied was not large enough, and the next step should be to increase the sample size and improve the accuracy of the study. Based on the fact that the proposed algorithm has such an excellent performance, it can be applied to other fields as well, such as recommender system [97–100], human activity recognition [101], text clustering [102], medical image augmentation [103, 104], microgrids planning [105], named entity recognition [106], COVID-19 classification [107–109], and object tracking [110].
Conclusion
To find a machine learning method to predict whether a mice PH model is successfully established, a model called HCWOA–KELM is proposed in this paper. To improve the effectiveness and stability of the model, a TRIZ-based hybrid transformation mechanism was added to the WOA. A total of 28 benchmark functions were used to test the global optimality of the improved WOA. The experimental results show that the proposed algorithm performs better global optimization on various test functions than the other eight optimization methods. Similarly, a comparison of bHCWOA–KELM with the other six classifiers and five feature selection methods demonstrates that the proposed model has higher classification accuracy and stability. Finally, the five most selected features in tenfold cross-validation experiments proved crucial for PH prediction.
Acknowledgements
This research was supported by grants from the National Natural Science Foundation of China (82003831, 62076185 and U1809209); the Project of Health Commission of Zhejiang Province (2020KY177); the Wenzhou Technology Foundation (Y2020002); the Natural Science Foundation of Zhejiang Province (LZ22F020005), and project funded by the First Affiliated Hospital of Wenzhou Medical University Youth Excellence Project (QNYC114).
Availability of Data and Materials
The data involved in this study are all public data, which can be downloaded through public channels.
Declarations
Conflict of Interest
The authors declare that there is no conflict of interests regarding the publication of article.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Jiao Hu, Shushu Lv and Tao Zhou contributed equally to this work.
==== Refs
References
1. Mandras SA Mehta HS Vaidya A Pulmonary hypertension: A brief guide for clinicians Mayo Clinic Proceedings 2020 95 1978 1988 10.1016/j.mayocp.2020.04.039 32861339
2. Simonneau, G., Montani, D., Celermajer, D. S., Denton, C. P., Gatzoulis, M. A., Krowka, M., Williams, P. G. & Souza, R. (2019). Haemodynamic definitions and updated clinical classification of pulmonary hypertension. European Respiratory Journal, 53, 01913–2018.
3. Walter K Pulmonary hypertension JAMA 2021 326 1116 10.1001/jama.2021.11054 34546299
4. Grapsa J Pereira Nunes MC Tan TC Cabrita IZ Coulter T Smith BC Dawson D Gibbs JS Nihoyannopoulos P Echocardiographic and hemodynamic predictors of survival in precapillary pulmonary hypertension: seven-year follow-up Circulation: Cardiovascular Imaging 2015 8 e002107 26063743
5. McLaughlin VV Archer SL Badesch DB Barst RJ Farber HW Lindner JR Mathier MA McGoon MD Park MH Rosenson RS Rubin LJ Tapson VF Varga J ACCF/AHA 2009 expert consensus document on pulmonary hypertension a report of the American College of Cardiology Foundation Task Force on Expert Consensus Documents and the American Heart Association developed in collaboration with the American College of Chest Physicians; American Thoracic Society, Inc.; and the Pulmonary Hypertension Association Journal of the American College of Cardiology 2009 53 1573 1619 10.1016/j.jacc.2009.01.004 19389575
6. Humbert M Lau EM Montani D Jaïs X Sitbon O Simonneau G Advances in therapeutic interventions for patients with pulmonary arterial hypertension Circulation 2014 130 2189 2208 10.1161/CIRCULATIONAHA.114.006974 25602947
7. Rubin LJ Primary pulmonary hypertension Chest 1993 104 236 250 10.1378/chest.104.1.236 8325077
8. Westerhof BE Saouti N van der Laarse WJ Westerhof N Vonk Noordegraaf A Treatment strategies for the right heart in pulmonary hypertension Cardiovascular Research 2017 113 1465 1473 10.1093/cvr/cvx148 28957540
9. Chan L Chin LMK Kennedy M Woolstenhulme JG Nathan SD Weinstein AA Connors G Weir NA Drinkard B Lamberti J Keyser RE Benefits of intensive treadmill exercise training on cardiorespiratory function and quality of life in patients with pulmonary hypertension Chest 2013 143 333 343 10.1378/chest.12-0993 22922554
10. Chen J Sysol JR Singla S Zhao S Yamamura A Valdez-Jasso D Abbasi T Shioura KM Sahni S Reddy V Sridhar A Gao H Torres J Camp SM Tang H Ye SQ Comhair S Dweik R Hassoun P Yuan JX Garcia JGN Machado RF Nicotinamide phosphoribosyltransferase promotes pulmonary vascular remodeling and is a therapeutic target in pulmonary arterial hypertension Circulation 2017 135 1532 1546 10.1161/CIRCULATIONAHA.116.024557 28202489
11. Zelt JGE Chaudhary KR Cadete VJ Mielniczuk LM Stewart DJ Medical therapy for Heart failure associated with Pulmonary Hypertension Circulation Research 2019 124 1551 1567 10.1161/CIRCRESAHA.118.313650 31120820
12. Lahm T Douglas IS Archer SL Bogaard HJ Chesler NC Haddad F Hemnes AR Kawut SM Kline JA Kolb TM Mathai SC Mercier O Michelakis ED Naeije R Tuder RM Ventetuolo CE Vieillard-Baron A Voelkel NF Vonk-Noordegraaf A Hassoun PM Assessment of right ventricular function in the research setting: Knowledge gaps and pathways forward. An official American thoracic society research statement American Journal of Respiratory and Critical Care Medicine 2018 198 e15 e43 10.1164/rccm.201806-1160ST 30109950
13. Ayyoubzadeh SM Ayyoubzadeh SM Zahedi H Ahmadi M S, R. N. K. Predicting COVID-19 incidence through analysis of google trends data in Iran: Data mining and deep learning pilot study JMIR Public Health and Surveillance 2020 6 e18828 10.2196/18828 32234709
14. Pourhomayoun M Shakibi M Predicting mortality risk in patients with COVID-19 using machine learning to help medical decision-making Smart Health 2021 20 100178 10.1016/j.smhl.2020.100178 33521226
15. Polat H Mehr HD Cetin A Diagnosis of chronic kidney disease based on support vector machine by feature selection methods Journal of medical systems 2017 10.1007/s10916-017-0703-x
16. Abbad Ur Rehman H Lin C-Y Mushtaq Z Su S-F Performance analysis of machine learning algorithms for thyroid disease Arabian Journal for Science and Engineering 2021 46 9437 9449 10.1007/s13369-020-05206-x
17. Pashaei E Pashaei E An efficient binary chimp optimization algorithm for feature selection in biomedical data classification Neural Computing & Applications 2022 34 6427 6451 10.1007/s00521-021-06775-0
18. Alsaeedi MAK Kurnaz S Feature selection for diagnose coronavirus (COVID-19) disease by neural network and Caledonian crow learning algorithm Applied Nanoscience 2022 10.1007/s13204-021-02159-x
19. Lamba R Gulati T Alharbi HF Jain A A hybrid system for parkinson's disease diagnosis using machine learning techniques International Journal of Speech Technology 2022 25 583 593 10.1007/s10772-021-09837-9
20. Hu J Liu Y Heidari AA Bano Y Ibrohimov A Liang G Chen H Chen X Zaguia A Turabieh H An effective model for predicting serum albumin level in hemodialysis patients Computers in Biology and Medicine 2022 140 105054 10.1016/j.compbiomed.2021.105054
21. Faisal FUR Khatri U Kwon G-R Diagnosis of Alzheimer's disease using combined feature selection method Journal of Korea Multimedia Society 2021 24 667 675
22. Hu J Han Z Heidari AA Shou Y Ye H Wang L Huang X Chen HL Chen Y Wu P Detection of COVID-19 severity using blood gas analysis parameters and Harris hawks optimized extreme learning machine Computers in Biology and Medicine 2022 142 105166 10.1016/j.compbiomed.2021.105166 35077935
23. Heidari AA Mirjalili S Faris H Aljarah I Mafarja M Chen H Harris Hawks optimization: Algorithm and applications Future Generation Computer Systems-the International Journal of Escience 2019 97 849 872 10.1016/j.future.2019.02.028
24. Yang Y Chen HL Heidari AA Gandomi AH Hunger games search: Visions, conception, implementation, deep analysis, perspectives, and towards performance shifts Expert Systems with Applications 2021 177 114864 10.1016/j.eswa.2021.114864
25. Tu JZ Chen HL Wang MJ Gandomi AH The colony predation algorithm Journal of Bionic Engineering 2021 18 674 710 10.1007/s42235-021-0050-y
26. Ahmadianfar I Asghar Heidari A Gandomi AH Chu X Chen H RUN beyond the metaphor: An efficient optimization algorithm based on Runge Kutta method Expert Systems with Applications 2021 181 115079 10.1016/j.eswa.2021.115079
27. Ahmadianfar I Asghar Heidari A Noshadian S Chen H Gandomi AH INFO: An efficient optimization algorithm based on weighted mean of vectors Expert Systems with Applications 2022 195 116516 10.1016/j.eswa.2022.116516
28. Li SM Chen H Wang M Heidari AA Mirjalili S Slime mould algorithm: A new method for stochastic optimization Future Generation Computer Systems 2020 111 300 323 10.1016/j.future.2020.03.055
29. Chen HL Yang B Wang JS Wang G Li HZ Liu BW Towards an optimal support vector machine classifier using a parallel particle swarm optimization strategy Applied Mathematics and Computation 2014 239 180 197 10.1016/j.amc.2014.04.039
30. Han X Han Y Chen Q Li J Sang H Liu Y Pan Q Nojima Y Distributed flow shop scheduling with sequence-dependent setup times using an improved iterated greedy algorithm Complex System Modeling and Simulation 2021 1 198 217 10.23919/CSMS.2021.0018
31. Gao D Wang G-G Pedrycz W Solving fuzzy job-shop scheduling problem using DE algorithm improved by a selection mechanism IEEE Transactions on Fuzzy Systems 2020 28 3265 3275 10.1109/TFUZZ.2020.3003506
32. Wang G Gao D Pedrycz W Solving multi-objective fuzzy job-shop scheduling problem by a hybrid adaptive differential evolution algorithm IEEE Transactions on Industrial Informatics 2022 10.1109/TII.2022.3165636
33. Li Q Chen HL Huang H Zhao XH Cai ZN Tong CF Liu WB Tian X An enhanced grey wolf optimization based feature selection wrapped kernel extreme learning machine for medical diagnosis Computational and Mathematical Methods in Medicine 2017 2017 9512741 10.1155/2017/9512741 28246543
34. Cai ZN Gu JH Wen CY Zhao D Huang CY Huang H Tong CF Li J Chen HL An intelligent Parkinson’s disease diagnostic system based on a chaotic bacterial foraging optimization enhanced fuzzy KNN approach Computational and Mathematical Methods in Medicine 2018 2018 2396952 10.1155/2018/2396952 30034509
35. Yu H Yuan K Li W Zhao N Chen W Huang C Chen H Wang M Improved butterfly optimizer-configured extreme learning machine for fault diagnosis Complexity 2021 2021 6315010
36. Ye XJ Liu W Li H Wang MJ Chi C Liang GX Chen HL Huang H Modified whale optimization algorithm for solar cell and PV module parameter identification Complexity 2021 2021 8878686 10.1155/2021/8878686
37. Deng W Zhang XX Zhou YQ Liu Y Zhou XB Chen HL Zhao HM An enhanced fast non-dominated solution sorting genetic algorithm for multi-objective problems Information Sciences 2022 585 441 453 10.1016/j.ins.2021.11.052
38. Hua YC Liu QQ Hao KG Jin YC A survey of evolutionary algorithms for multi-objective optimization problems with irregular pareto fronts IEEE/CAA Journal of Automatica Sinica 2021 8 303 318 10.1109/JAS.2021.1003817
39. Zhao F Di S Cao J Tang J A novel cooperative multi-stage hyper-heuristic for combination optimization problems Complex System Modeling and Simulation 2021 1 91 108 10.23919/CSMS.2021.0010
40. Wang G Gui W Liang G Zhao X Wang M Mafarja M Turabieh H Xin J Chen H Ma X Spiral motion enhanced elite whale optimizer for global tasks Complexity 2021 2021 8130378 10.1155/2021/8130378
41. Chen CC Wang XC Yu HL Zhao NN Wang MJ Chen HL An enhanced comprehensive learning particle swarm optimizer with the elite-based dominance scheme Complexity 2020 2020 4968063 10.1155/2020/4968063
42. Altshuller, G. S. (1998). 40 Principles: TRIZ Keys to technical innovation.
43. Carrara P Morandi W Campagiorni V Tandoi E An innovation management framework to enhance TRIZ diffusion in smes Acta Technica Napocensis Series-Applied Mathematics Mechanics and Engineering 2021 64 455 464
44. Petrov V Solodkina H TRIZ: Innovation opportunities for the fashion industry Acta Technica Napocensis Series-Applied Mathematics Mechanics and Engineering 2020 63 15 22
45. Liu W Tan R Peng Q Li H Li Z Yang B Impact of TRIZ learning on performance in biologically inspired design International Journal of Engineering Education 2020 36 974 987
46. Hmina K El Amine M Lasri L Sallaou M Preferences-based approach for TRIZ contradiction matrix exploitation in preliminary design Fme Transactions 2020 48 588 599 10.5937/fme2003588H
47. Spreafico C Quantifying the advantages of TRIZ in sustainability through life cycle assessment Journal of Cleaner Production 2021 303 126955 10.1016/j.jclepro.2021.126955
48. Huang GB Zhou H Ding X Zhang R Extreme learning machine for regression and multiclass classification IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics) 2012 42 513 529 10.1109/TSMCB.2011.2168604
49. Yang XS A new metaheuristic bat-inspired algorithm 2010 Springer
50. Mirjalili S Moth-flame optimization algorithm: A novel nature-inspired heuristic paradigm Knowledge-Based Systems 2015 89 228 249 10.1016/j.knosys.2015.07.006
51. Storn R Price K Differential evolution—A simple and efficient heuristic for global optimization over continuous spaces Journal of Global Optimization 1997 11 341 359 10.1023/A:1008202821328
52. Jia D Zheng G Khurram Khan M An effective memetic differential evolution algorithm based on chaotic local search Information Sciences 2011 181 3175 3187 10.1016/j.ins.2011.03.018
53. Buchari MA Mardiyanto S Hendradjaya B Implementation of chaotic gaussian particle swarm optimization for optimize learning-to-rank software defect prediction model construction Journal of Physics: Conference Series 2018 978 012079
54. Chen W Zhang J Lin Y Chen N Zhan Z Chung HS Li Y Shi Y Particle swarm optimization with an aging leader and challengers IEEE Transactions on Evolutionary Computation 2013 17 241 258 10.1109/TEVC.2011.2173577
55. Garcia S Fernandez A Luengo J Herrera F Advanced nonparametric tests for multiple comparisons in the design of experiments in computational intelligence and data mining: Experimental analysis of power Information Sciences 2010 180 2044 2064 10.1016/j.ins.2009.12.010
56. Chen H Li S Heidari AA Wang P Li J Yang Y Wang M Huang C Efficient multi-population outpost fruit fly-driven optimizers: Framework and advances in support vector machines Expert Systems with Applications 2020 142 112999 10.1016/j.eswa.2019.112999
57. Hu J Liu Y Heidari AA Bano Y Ibrohimov A Liang G Chen H Chen X Zaguia A Turabieh H An effective model for predicting serum albumin levels in hemodialysis patients Computers in Biology and Medicine 2021 2021 105054
58. Sankaran VG Xu J Orkin SH Advances in the understanding of haemoglobin switching British Journal of Haematology 2010 149 181 194 10.1111/j.1365-2141.2010.08105.x 20201948
59. Thom CS Dickson CF Gell DA Weiss MJ Hemoglobin variants: Biochemical properties and clinical correlates Cold Spring Harbor Perspectives in Medicine 2013 3 11858 10.1101/cshperspect.a011858
60. Hauser A Schmitt L Troesch S Saugy JJ Cejuela-Anta R Faiss R Robinson N Wehrlin JP Millet GP Similar hemoglobin mass response in hypobaric and normobaric hypoxia in Athletes Medicine and Science in Sports and Exercise 2016 48 734 741 10.1249/MSS.0000000000000808 26540262
61. Deem S Kim JU Manjula BN Acharya AS Kerr ME Patel RP Gladwin MT Swenson ER Effects of S-nitrosation and cross-linking of hemoglobin on hypoxic pulmonary vasoconstriction in isolated rat lungs Circulation Research 2002 91 626 632 10.1161/01.RES.0000036268.47509.02 12364391
62. Donadee C Raat NJ Kanias T Tejero J Lee JS Kelley EE Zhao X Liu C Reynolds H Azarov I Frizzell S Meyer EM Donnenberg AD Qu L Triulzi D Kim-Shapiro DB Gladwin MT Nitric oxide scavenging by red blood cell microparticles and cell-free hemoglobin as a mechanism for the red cell storage lesion Circulation 2011 124 465 476 10.1161/CIRCULATIONAHA.110.008698 21747051
63. Shaver CM Wickersham N McNeil JB Nagata H Miller A Landstreet SR Kuck JL Diamond JM Lederer DJ Kawut SM Palmer SM Wille KM Weinacker A Lama VN Crespo MM Orens JB Shah PD Hage CA Cantu E 3rd Porteous MK Dhillon G McDyer J Bastarache JA Christie JD Ware LB Cell-free hemoglobin promotes primary graft dysfunction through oxidative lung endothelial injury JCI Insight 2018 10.1172/jci.insight.98546
64. Shaver CM Upchurch CP Janz DR Grove BS Putz ND Wickersham NE Dikalov SI Ware LB Bastarache JA Cell-free hemoglobin: A novel mediator of acute lung injury American Journal of Physiology: Lung Cellular and Molecular Physiology 2016 310 L532 L541 26773065
65. Detterich JA Kato RM Rabai M Meiselman HJ Coates TD Wood JC Chronic transfusion therapy improves but does not normalize systemic and pulmonary vasculopathy in sickle cell disease Blood 2015 126 703 710 10.1182/blood-2014-12-614370 26036801
66. Fernandez AJ Flaxman NA Common laboratory tests, values, and interpretations Special Care in Dentistry 1985 5 264 269 10.1111/j.1754-4505.1985.tb00591.x 3867169
67. Reinhart WH The optimum hematocrit Clinical Hemorheology and Microcirculation 2016 64 575 585 10.3233/CH-168032 27767984
68. Crowell JW Smith EE Determinant of the optimal hematocrit Journal of Applied Physiology 1967 22 501 504 10.1152/jappl.1967.22.3.501 6020234
69. Ernst E Haematocrit and cardiovascular risk Journal of Internal Medicine 1995 237 527 528 10.1111/j.1365-2796.1995.tb00880.x 7782722
70. Stauffer E Loyrion E Hancco I Waltz X Ulliel-Roche M Oberholzer L Robach P Pichon A Brugniaux JV Bouzat P Doutreleau S Connes P Verges S Blood viscosity and its determinants in the highest city in the world Journal of Physiology 2020 598 4121 4130 10.1113/JP279694 32445208
71. Beall CM Two routes to functional adaptation: Tibetan and Andean high-altitude natives Proceedings of the National Academy of Sciences of the United States of America 2007 104 8655 8660 10.1073/pnas.0701985104 17494744
72. Bath PM Butterworth RJ Platelet size: Measurement, physiology and vascular disease Blood Coagulation & Fibrinolysis 1996 7 157 161 10.1097/00001721-199603000-00011 8735807
73. Thompson CB Jakubowski JA The pathophysiology and clinical relevance of platelet heterogeneity Blood 1988 72 1 8 10.1182/blood.V72.1.1.1 3291975
74. Gladwin AM Martin JF The control of megakaryocyte ploidy and platelet production: Biology and pathology International Journal of Cell Cloning 1990 8 291 298 10.1002/stem.5530080414 2205666
75. Kaya MG Yarlioglues M Gunebakmaz O Gunturk E Inanc T Dogan A Kalay N Topsakal R Platelet activation and inflammatory response in patients with non-dipper hypertension Atherosclerosis 2010 209 278 282 10.1016/j.atherosclerosis.2009.09.010 19782364
76. Korniluk A Koper-Lenkiewicz OM Kamińska J Kemona H Dymicka-Piekarska V Mean Platelet Volume (MPV): New perspectives for an old marker in the course and prognosis of inflammatory conditions Mediators of Inflammation 2019 2019 9213074 10.1155/2019/9213074 31148950
77. Braekkan SK Mathiesen EB Njølstad I Wilsgaard T Størmer J Hansen JB Mean platelet volume is a risk factor for venous thromboembolism: The Tromsø Study, Tromsø, Norway Journal of Thrombosis and Haemostasis 2010 8 157 162 10.1111/j.1538-7836.2009.03498.x 19496920
78. Kovács S Csiki Z Zsóri KS Bereczky Z Shemirani AH Characteristics of platelet count and size and diagnostic accuracy of mean platelet volume in patients with venous thromboembolism. A systematic review and meta-analysis Platelets 2019 30 139 147 10.1080/09537104.2017.1414175 29252063
79. Steiropoulos P Papanas N Nena E Xanthoudaki M Goula T Froudarakis M Pita E Maltezos E Bouros D Mean platelet volume and platelet distribution width in patients with chronic obstructive pulmonary disease: The role of comorbidities Angiology 2013 64 535 539 10.1177/0003319712461436 23052724
80. Nadar SK Blann AD Kamath S Beevers DG Lip GY Platelet indexes in relation to target organ damage in high-risk hypertensive patients: A substudy of the Anglo-Scandinavian Cardiac Outcomes Trial (ASCOT) Journal of the American College Cardiology 2004 44 415 422 10.1016/j.jacc.2004.03.067
81. Boos CJ Beevers GD Lip GY Assessment of platelet activation indices using the ADVIATM 120 amongst 'high-risk' patients with hypertension Annals of Medicine 2007 39 72 78 10.1080/07853890601040063 17364453
82. Inanc T Kaya MG Yarlioglues M Ardic I Ozdogru I Dogan A Kalay N Gunturk E Gunebakmaz O Gul I Topsakal R The mean platelet volume in patients with non-dipper hypertension compared to dippers and normotensives Blood Pressure 2010 19 81 85 10.3109/08037050903516284 20367545
83. Handtke S Thiele T Large and small platelets-(When) do they differ? Journal of Thrombosis and Haemostasis 2020 18 1256 1267 10.1111/jth.14788 32108994
84. Sachdev R Tiwari AK Goel S Raina V Sethi M Establishing biological reference intervals for novel platelet parameters (immature platelet fraction, high immature platelet fraction, platelet distribution width, platelet large cell ratio, platelet-X, plateletcrit, and platelet distribution width) and their correlations among each other Indian Journal of Pathology and Microbiology 2014 57 231 235 10.4103/0377-4929.134676 24943755
85. Arslan D Cimen D Guvenc O Kaya F Sert A Oran B Platelet distribution width and mean platelet volume in children with pulmonary arterial hypertension secondary to congenital heart disease with left-to-right shunt: New indices of severity? Pediatric Cardiology 2013 34 1013 1016 10.1007/s00246-012-0600-5 23254645
86. Li T Jin J Meng Z Zhang W Li Y Yu X Du X Liu M Zhang Q Gao Y Song K Wang X Fan Y Wang Y 2020 The different associations between platelet distribution width and hypertension subtypes in males and females Bioscience Reports 10.1042/BSR20201747
87. Bekler A Ozkan MT Tenekecioglu E Gazi E Yener AU Temiz A Altun B Barutcu A Erbag G Binnetoglu E Increased platelet distribution width is associated with severity of coronary artery disease in patients with acute coronary syndrome Angiology 2015 66 638 643 10.1177/0003319714545779 25112777
88. Vagdatli E Gounari E Lazaridou E Katsibourlia E Tsikopoulou F Labrianou I Platelet distribution width: A simple, practical and specific marker of activation of coagulation Hippokratia 2010 14 28 32 20411056
89. Jindal S Gupta S Gupta R Kakkar A Singh HV Gupta K Singh S Platelet indices in diabetes mellitus: Indicators of diabetic microvascular complications Hematology 2011 16 86 89 10.1179/102453311X12902908412110 21418738
90. Zheng YG Yang T Xiong CM He JG Liu ZH Gu Q Zhao ZH Ni XH Platelet distribution width and mean platelet volume in idiopathic pulmonary arterial hypertension Heart Lung and Circulation 2015 24 566 572 10.1016/j.hlc.2014.11.025
91. Kaito K Otsubo H Usui N Yoshida M Tanno J Kurihara E Matsumoto K Hirata R Domitsu K Kobayashi M Platelet size deviation width, platelet large cell ratio, and mean platelet volume have sufficient sensitivity and specificity in the diagnosis of immune thrombocytopenia British Journal of Haematology 2005 128 698 702 10.1111/j.1365-2141.2004.05357.x 15725092
92. Karpatkin S Heterogeneity of human platelets. I. Metabolic and kinetic evidence suggestive of young and old platelets Journal of Clinical Investigation 1969 48 1073 1082 10.1172/JCI106063 5771188
93. Karpatkin S Strick N Heterogeneity of human platelets. V. Differences in glycolytic and related enzymes with possible relation to platelet age Journal of Clinical Investigation 1972 51 1235 1243 10.1172/JCI106918 4262550
94. Colkesen Y Muderrisoglu H The role of mean platelet volume in predicting thrombotic events Clinical Chemistry and Laboratory Medicine 2012 50 631 634 10.1515/cclm.2011.806 22112054
95. Detwiler TC Odell TT Jr Mac DT Platelet size, ATP content, and clot retraction in relation to platelet age American Journal of Physiology 1962 203 107 110 10.1152/ajplegacy.1962.203.1.107 13885826
96. Balduini CL Noris P Spedini P Belletti S Zambelli A Da Prada GA Relationship between size and thiazole orange fluorescence of platelets in patients undergoing high-dose chemotherapy British Journal of Haematology 1999 106 202 207 10.1046/j.1365-2141.1999.01475.x 10444188
97. Wang D Liang Y Xu D Feng X Guan RJK-BS A content-based recommender system for computer science publications Knowledge-Based Systems 2018 157 1 9 10.1016/j.knosys.2018.05.001
98. Li J Chen C Chen H Tong C Towards context-aware social recommendation via individual trust Knowledge-Based Systems 2017 127 58 66 10.1016/j.knosys.2017.02.032
99. Li J Lin J A probability distribution detection based hybrid ensemble QoS prediction approach Information Sciences 2020 519 289 305 10.1016/j.ins.2020.01.046
100. Li J Zheng X-L Chen S-T Song W-W Chen D-R An efficient and reliable approach for quality-of-service-aware service composition Information Sciences 2014 269 238 254 10.1016/j.ins.2013.12.015
101. Qiu S Zhao H Jiang N Wang Z Liu L An Y Zhao H Miao X Liu R Fortino G Multi-sensor information fusion based on machine learning for real applications in human activity recognition: State-of-the-art and research challenges Information Fusion 2022 80 241 265 10.1016/j.inffus.2021.11.006
102. Guan R Zhang H Liang Y Giunchiglia F Huang L Feng X Deep feature-based text clustering and its explanation IEEE Transactions on Knowledge and Data Engineering 2020 10.1109/TKDE.2020.3001558
103. Guan Q Chen Y Wei Z Heidari AA Hu H Yang X-H Zheng J Zhou Q Chen H Chen F Medical image augmentation for lesion detection using a texture-constrained multichannel progressive GAN Computers in Biology and Medicine 2022 145 105444 10.1016/j.compbiomed.2022.105444 35421795
104. Chen Y Yang X-H Wei Z Heidari AA Zheng N Li Z Chen H Hu H Zhou Q Guan Q Generative adversarial networks in medical image augmentation: A review Computers in Biology and Medicine 2022 144 105382 10.1016/j.compbiomed.2022.105382 35276550
105. Cao X Sun X Xu Z Zeng B Guan X Hydrogen-based networked microgrids planning through two-stage stochastic programming with mixed-integer conic recourse IEEE Transactions on Automation Science and Engineering 2021 10.1109/TASE.2021.3130179
106. Yang Z Chen H Zhang J Chang Y Context-aware attentive multi-level feature fusion for named entity recognition IEEE Transactions on Neural Networks and Learning Systems 2022 10.1109/TNNLS.2022.3178522
107. Wang S-H Nayak DR Guttery DS Zhang X Zhang Y-D COVID-19 classification by CCSHNet with deep fusion using transfer learning and discriminant correlation analysis Information Fusion 2021 68 131 148 10.1016/j.inffus.2020.11.005 33519321
108. Wang S-H Govindaraj VV Górriz JM Zhang X Zhang Y-D Covid-19 classification by FGCNet with deep feature fusion from graph convolutional network and convolutional neural network Information Fusion 2021 67 208 229 10.1016/j.inffus.2020.10.004 33052196
109. Yang L Wang S-H Zhang Y-D EDNC: Ensemble deep neural network for COVID-19 recognition Tomography 2022 8 869 890 10.3390/tomography8020071 35314648
110. Zhang XQ Hu WM Xie NH Bao HJ Maybank S A robust tracking system for low frame rate video International Journal of Computer Vision 2015 115 279 304 10.1007/s11263-015-0819-8
| 36466726 | PMC9703443 | NO-CC CODE | 2022-11-29 23:21:09 | no | J Bionic Eng. 2022 Nov 28;:1-20 | utf-8 | J Bionic Eng | 2,022 | 10.1007/s42235-022-00292-z | oa_other |
==== Front
Air Qual Atmos Health
Air Qual Atmos Health
Air Quality, Atmosphere, & Health
1873-9318
1873-9326
Springer Netherlands Dordrecht
1286
10.1007/s11869-022-01286-w
Article
Airborne transmission of biological agents within the indoor built environment: a multidisciplinary review
Argyropoulos Christos D. 1
Skoulou Vasiliki 2
Efthimiou Georgios 3
http://orcid.org/0000-0003-3250-998X
Michopoulos Apostolos K. [email protected]
4
1 grid.440838.3 0000 0001 0642 7601 School of Medicine, European University Cyprus, 6 Diogenes Street, 2404 Egkomi, Nicosia Cyprus
2 grid.9481.4 0000 0004 0412 8669 B3 Challenge Group, Chemical Engineering, School of Engineering, University of Hull, Cottingham Road, Hull, HU6 7RX UK
3 grid.9481.4 0000 0004 0412 8669 Centre for Biomedicine, Hull York Medical School, University of Hull, Cottingham Road, Hull, HU6 7RX UK
4 grid.6603.3 0000000121167908 Energy & Environmental Design of Buildings Research Laboratory, University of Cyprus, P.O. Box 20537, 1678 Nicosia, Cyprus
28 11 2022
157
2 7 2021
17 11 2022
© The Author(s), under exclusive licence to Springer Nature B.V. 2022, Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
The nature and airborne dispersion of the underestimated biological agents, monitoring, analysis and transmission among the human occupants into building environment is a major challenge of today. Those agents play a crucial role in ensuring comfortable, healthy and risk-free conditions into indoor working and leaving spaces. It is known that ventilation systems influence strongly the transmission of indoor air pollutants, with scarce information although to have been reported for biological agents until 2019. The biological agents’ source release and the trajectory of airborne transmission are both important in terms of optimising the design of the heating, ventilation and air conditioning systems of the future. In addition, modelling via computational fluid dynamics (CFD) will become a more valuable tool in foreseeing risks and tackle hazards when pollutants and biological agents released into closed spaces. Promising results on the prediction of their dispersion routes and concentration levels, as well as the selection of the appropriate ventilation strategy, provide crucial information on risk minimisation of the airborne transmission among humans. Under this context, the present multidisciplinary review considers four interrelated aspects of the dispersion of biological agents in closed spaces, (a) the nature and airborne transmission route of the examined agents, (b) the biological origin and health effects of the major microbial pathogens on the human respiratory system, (c) the role of heating, ventilation and air-conditioning systems in the airborne transmission and (d) the associated computer modelling approaches. This adopted methodology allows the discussion of the existing findings, on-going research, identification of the main research gaps and future directions from a multidisciplinary point of view which will be helpful for substantial innovations in the field.
Keywords
Indoor air quality
Building ventilation
Airborne transmission
Bioaerosols
CFD models
Droplets
==== Body
pmcIntroduction
Release, circulation and dispersion of chemical (harmful pollutants) and biological agents within confined indoor spaces can be easily inhaled. For that reason, it is considered a serious threat for public health and therefore there is a continuous effort for preventing or controlling their release (Jones 1999). Such agents may include from toxic chemicals, pathogenic microorganisms (e.g. fungal and bacterial spores) and microbe-bearing air particles, such as droplets to various types of solids such as dust (Ghosh et al. 2015). Those are responsible for chemical poisoning or serious respiratory infections via the spread of infectious biological agents at hospitals, long-term care facilities (Vogazianos et al. 2021), schools and office areas (Taylor et al. 2012). The main route of human infection by biological agents is usually via the human respiratory system. This takes place by inhalation of tiny particles or droplets, commonly referred as particulates, however, in the case of pathogens those can be also contracted from touching infected surfaces, such as door handles, taps and furniture (Madigan 2009; Prat and Lacoma 2016).
The shape, size and formation–dispersion mechanisms of these particulates when especially are in a liquid place (droplets) as well as their physicochemical properties affect significantly their potential to cause respiratory diseases. Those characteristics determine biological agents transmission patterns and how easy it is to be inserted into the human body via inhalation and further penetrating into the tissues of the lower respiratory system. Usually, only the micron-sized particulates can reach our lungs and the alveoli, leading to serious respiratory diseases (Bansal et al. 2018; Jones 1999). More information on which particle/droplet sizes are deposited in which part of the human respiratory tract, depending on the nature of the particulates, can be found in the following sections.
New respiratory pathogens have emerged during the last couple of years, with the most notorious being SARS-CoV-2, a novel coronavirus which is responsible for the infectious disease COVID-19 that has caused more than a million of deaths worldwide in 2019–2020 according to Rothan and Byrareddy (2020). Another coronavirus, MERS (Middle East respiratory syndrome), also caused many deaths in the Middle East in 2017 (Hageman 2020).
In addition, traditional pulmonary infectious agents, such as influenza virus (causing common flu), Streptococcus pneumoniae (causing pneumonia), Mycobacterium tuberculosis (causing tuberculosis) and Aspergillus fumigatus (causing lung aspergillosis) are still considered major health hazards (Hunter 2016; Latgé and Chamilos 2019; Murray et al., 2013; Pleschka 2013).
In order to rapidly detect and identify these infectious biological agents in the air or on surfaces, an arsenal of sophisticated new technologies is necessary to be developed. Those technologies will provide real-time accurate information about the presence of particulates in an indoor environment. Several such approaches have been developed (Huffman et al. 2020; Nasir et al. 2019; Usachev et al. 2013); however, most of them are still at low technology readiness level, an experimental level, and they are not routinely applied.
In addition, Heating, Ventilation and Air-Conditioning (HVAC) systems can be employed to control the transmission of harmful particles (solids or droplets). Different types of HVAC methods can reduce the spread of such agents in buildings or even eliminate the threat posed by pathogenic infectious microorganisms (Li et al. 2007); Shajahan et al. 2019).
Also, factors like the ventilation rate and heating/cooling settings of such systems can significantly influence the indoor transmission of hazardous agents (Li et al. 2007; Zhang et al. 2020a, b, c, d).
Moreover, computer modelling approaches have been used for predicting transmission patterns of chemical and biological agents in confined indoor areas. The most predominant methods are multi-zone and CFD modelling that are often used in combination for obtaining more robust results (Wang and Chen 2008a). Numerous such studies have been carried out in key close space areas such as hospitals and offices and have helped in designing new effective sanitation approaches (Chen et al. 2011; Emmerich et al. 2013; Karakitsios et al. 2020; Lim et al. 2011).
The aim of this multidisciplinary review is to examine the critical issue of harmful particles control, with the emphasis drawn on biological agents, within indoor environments from four different angles (physical, biological, HVAC and computer modelling), highlighting key research gaps in each area and suggest solutions that could lead to substantially improved indoor health strategies in the near future.
This manuscript is organised in seven sections. The ‘Introduction’ section presents a short introduction to the reviewed topic. The ‘Methods–literature review approach’ section includes the classification of the present review and methodology of the collection and analysis of the relevant research works in the field. In the ‘Droplet formation mechanisms’ section, the effect of the physicochemical nature (chemical characteristics, size and shape) of particulates such as dust and droplets of water, the most characteristic formation mechanisms of droplets and aerosols and their dispersion into indoor space environment are discussed. The ‘Aerosols and bioaerosols’ section includes the most characteristic microbiological agents that are carried within aerosols with an account on the methods that are currently used for their detection and identification. In the ‘The role of heating, ventilation and air-conditioning systems’ section, the role of heating, ventilation and air-conditioning (HVAC) systems in association with the alternative ventilation patterns regarding the dispersion of pollutants and biological agents into indoor spaces is presented. The ‘Computer modelling of particles and biological agents’ airborne transmission into indoor built environment’ section exhibits the available computational modelling techniques for the prediction of biological agents’ airborne transmission routes. Finally, in the ‘Conclusions—future directions’ section, the major findings, remarks and recommendations for future research are presented.
Methods—literature review approach
The present review is classified as a semi-systematic review, designed for the topic of dispersion of biological agents and pollutants in indoor air environments. This type of literature review studies is suitable for works of multidisciplinary group of researchers within diverse disciplines of engineering and other sciences as described in (Snyder 2019). The adopted literature review strategy focused on how the research in the field of the indoor air pollutants of biological origin, the latest often underestimated, has progressed and developed over time. The authors attempt to identify the potentially relevant research aspects which are important for the corresponding topic and synthesise these instead of measuring effect size, by using meta-narratives.
The importance of contribution of the present work is (a) mapping the recent trends of biological agents and pollutant dispersion in the indoor air research, (b) synthesize the current status of knowledge from different perspectives of a variety of disciplines and (c) create an updated agenda for further multidisciplinary research on the topic of indoor air pollution from biological agents, the main focus of this study, in which the current literature is scarce.
The research methodology used in the present semi-systematic review is composed of three primary and independent steps:Step 1: Database selection. Scopus, Google Scholar, PubMed, Web of Science and database platforms were used to retrieve the relevant literature related to the scope of the study.
Step 2: Searching Keywords. Due to the multidisciplinary context of this work, four different keyword families were used to identify the relevant articles per section. In the ‘Droplet formation mechanisms’ section, the words “particles”, “particulates”, “size”, “shape”, “indoor air pollution”, transmission”, “dispersion droplets”, “formation”, “technology”, “mechanism”, “suspension”, “re-suspension”, “particle size distribution”, “atomisation” and “coalescence”, as well as any combination among them, was used. The research works found were further narrowed down to the engineering aspects of the particles and droplets formation and airborne dispersion in the field of indoor air quality. In the ‘Aerosols and bioaerosols’ section, the names of the microbial agents and the relevant methods were used as keywords, in addition to biomedical terms such as “bio-aerosols”, “dust”, “pollen”, “transmission”, “air microbiology”, “microbial identification”, “airborne disease”, “respiratory disease”, “lung infection”, “infectious dose” and “immunity”, used to identify the relevant articles. The terms “ventilation”, “natural ventilation”, “personal ventilation”, “mixed ventilation”, “underfloor ventilation”, “mechanical ventilation”, “air distribution” and combined with the Boolean operators “OR” and “AND” with the associated terms “airborne transmission”, “thermal plume”, “droplet”, “contaminant removal efficiency” “heating”, “cooling” and “bioaerosol” were adopted in the ‘The role of heating, ventilation and air-conditioning systems’ section. In the ‘Computer modelling of particles and biological agents’ airborne transmission into indoor built environment’ section, regarding turbulence, modelling techniques terms such as “Reynolds-Averaged Navier–Stokes (RANS)”, “Unsteady Reynolds-Averaged Navier–Stokes (URANS)”, “Detached Eddy Simulation (DES)”, “Reynolds stress models (RSM)” and “Large Eddy Simulation (LES)” were used, in addition, to terms such as “indoor dispersion”, “dilution” “multiphase flows”, “Eulerian–Lagrangian techniques”, “Eulerian-Eulerian techniques”, multizone models”, “CFD—Physiologically Based Pharmacokinetic (PBPK)” or “CFD—Physiologically Based Toxicokinetic (PBTK)”. Furthermore, the combination of the aforementioned terms/ keywords from the ‘Droplet formation mechanisms’ and ‘‘Aerosols and bioaerosols’ sections along with “CFD” was also used to identify relevant papers.
Step 3: Article screening and reviewing. Articles were preliminary analysed through title, keywords, abstract and conclusions. This analysis was later on followed by an extensive reviewing of the articles selected from the screening process. The available material is certainly too much to be reviewed in a single paper. For this reason, regarding the modelling papers, the authors give special attention to what they consider the better established or more promising modelling approaches, such as single- and multi-zone models, CFD, coupling of CFD and multi-zone models, CFD-PKTE or CFD-PTBK models. No disrespect is therefore implied for studies with other models. It should be noted that extensive use has been made of the published literature on the field and of previous reviews.
Droplet formation mechanisms
The challenging nature of biological agents’ transmission in indoor environment
The importance of indoor air quality (IAQ) and spreading of pollutants and biological agents into indoor air, ranges from new types of chemicals and particulates released to infectious droplets spreading several kinds of diseases, and those are well known threats for the societies (Brundage et al. 1988; Cooke 1991; Jones 1999; Mutuku et al. 2020a). At the opening of the twentieth century (1918–2019), the outbreak of Spanish flu (H1N1) caused more than 1 billion infections and was then considered as the most lethal flu pandemic. Recently, Ni et al. (2020) reported that people spend approximately 90% of their time indoors with minimum time for outdoor activities. It is then obvious that staying long periods of time in a contaminated indoor environment increases the risk of respiratory diseases triggered due to the poor IAQ.
The nature, characteristics, behaviour and release mode of different pollutants and more importantly biological agents in indoor environment are still some of the areas which cause confusion among the researchers. This might be happening for reasons expanding from, for instance, the volatilisation or release of new types of chemicals emerging from new types of processes such as construction materials (Salthammer 2020) to recently developed unknown types of respiratory diseases. From all respiratory diseases, the severe acute respiratory ones are deemed to be the most important due to the nature of the disease spreading and infection via the ‘invisible’ airborne routes.
Nowadays, there is a good understanding of pollutants’ nature and their impact on human health. The way also the modern types of indoor air purification systems and processes are operating to more efficiently trap and separate indoor air pollutants, as well as their spreading mode (Luengas et al. 2015) is better understood. For the most common, old-generation indoor polluting agents such as chemicals ranging from asbestos, tar droplets of tobacco products, carbon monoxide (CO), volatile organic compounds (VOCs) to dust, coal and pollen particulates, there is a much thorough and better understanding of their transmission to humans when these released into indoor air. The same good level of understanding exists of their associated health problems, causes and effects for those well-known pollutants which are studied for more than two decades (Domingo et al. 2020; Jones 1999; Monn 2001).
How the recently appeared droplets of infectious diseases occur, it is still though unclear to the global scientific community, as well as how they spread into indoor air and infect human occupants. Two very characteristic examples are the infectious severe acute respiratory syndrome (SARS) or SARS-CoV-2 variant or subvariant respiratory system diseases. Such types of biological agents are dispersed and, most importantly, among infected to non-infected individuals, resulting in alarming public health problems.
Lately, there is also an increasing concern of companion animal-to-human transmission risk (Yin et al. 2020) and other animals infected by coronaviruses (Carducci et al., 2020). There is also a lively discussion around transmission of such diseases by contaminated droplets of human saliva, along with a discussion on the origin and nature of the new infectious diseases which proved to lead to epidemic crisis, such the one caused by SARS-CoV-2.
Today, the general understanding is that the infectious saliva droplets are transmitted in indoor spaces via two prevailing modes: (1) the direct and (2) indirect mode of transmission between the occupants of a confined indoor space environment (Dhand and Li 2020; Galvin et al. 2020). The alarming and yet urgent need for better understanding of the above-mentioned transmission routes have led the scientific community to classify and further investigate such biological agents transmitting modes, focusing especially on the most risky ones to be released in indoor environments.
The importance of not only better understanding, but also hindering the transmission of such airborne, either biological agents or hazardous chemicals inhaled, and targeting the human respiratory system, can be showcased by the SARS outbreak which first appeared in 2002–2003, (Morawska 2006) causing 774 deaths worldwide [www.nhs.uk/conditions/sars/, last accessed on 14.10.2022] (Lauxmann et al. 2020; Razzini et al. 2020). SARS-CoV-2 has recently been declared a pandemic by the World Health Association (WHO) and during the 21 months of 2020–2021 (January 2019–November 2021) killed more than 6,586,200 patients around the globe [https://www.worldometers.info/coronavirus/, last accessed on 26.10.2022].
According to Zhang et al. (2020a, b, c, d), the lower respiratory infections remain the primary cause of patients mortality worldwide, accounting for 650,000 deaths each year. This fact makes the issue of shading light and better understanding the pollutants and biological agents’ transmission through the droplet formation during inhalation and retention in the human tracheobronchial system, an area of research which necessitates further investigation as a matter of urgency. On the other hand, the chemical pollutants’ transport and deposition in the respiratory system have been studied excessively (Lauxmann et al. 2020; Mittal et al. 2020; Rothan and Byrareddy 2020) and as a result the main focus of this study will mainly be on the biological agents nature, spreading and transmission.
The human respiratory system
The anatomy and physiology of the human respiratory system both play an important role in either short (~ 2 m) or long distance transmission (> 2 m) of the airborne infectious diseases. During the accidental release of pollutants and/or biological agents in a sick building environment (Jones 1999) or unintentional release by a patient of a contaminant and inhalation of droplets from other healthy adults, there is a direct relevance of the human respiratory system’s role and especially the lungs’ operation (Bansal et al. 2018).
The human respiratory system is very complex and is constituted from many compartments of different shapes and sizes. It has the ability to absorb the indoor air’s droplets or solid particulates by inhalation (Steiner et al. 2020). When a person talks, coughs and sneezes spreads a cloud of tiny saliva droplets (aerosol) in a very short period, of a couple of hundreds of milliseconds (200 ms) (Bourouiba et al. 2014; Scharfman et al. 2016). Sneezes especially, which in fact are described as violent exhalation incidents, have received much less attention in the scientific literature and it is a field which needs further investigation. A sneeze leads to an extremely short (in the order of 150 ms) incident of aerosols formation and spreads at extremely high speed in the order of 35 m/s (Scharfman et al. 2016). The occurrence of such events is very similar to that of the well know liquid atomisation process of the liquid fuels (Vadivukkarasan et al. 2020). It is also important to note that aerosols of infectious respiratory diseases like SARS-CoV-2 survives for at least 3 h (Netz 2020), while similar viruses might survive for days. When those droplets land on open surfaces substantially increasing the risk of indirect transmission to humans via touches. As aerosol is defined the suspension of fine solid particles or liquid droplets in a gaseous medium. Both droplets and particulates, commonly known in engineering science as particles can be potentially carried away by indoor air flows, in either short or long distances. How far those aerosol droplets or any other infected solid nanoparticles can be transported depends mainly on their size, which only in the case of solids is a stable characteristic. This is much more complicated for the case of different transport mechanisms of droplets of infectious diseases and particulates taking place simultaneously. For example, in an air-conditioned environment convective mass transfer (enhanced by the air currents) is taking place when a patient sneezes or coughs then an aerosol formed which can be dispersed in the indoor space. At the same time the infected saliva droplets might be unstable in size as a result of the effect of room temperature, humidity or their droplet breaking up tendency due to hydrodynamics (behaviour of droplets in air). It has been found that a sneeze releases approximately 40,000 droplets, while a cough produces a considerably lower number of droplets at around 3,000 (Dhand and Li 2020). Similarly, when a person walking or touching areas full of dust infected solid particles can spread in air. However, the size of the solid particles is not changing as a result of the indoor environment conditions and thus understanding of this mode of transmission is less complicated compared to the airborne droplets transmission mechanism. Regardless their behaviour though, both saliva droplets and/or any infected solid particulates are inevitably and unconsciously inhaled by the occupants of confined indoor spaces. Both those agents, infectious or not, and depending on their size, they are diffused at different concentrations in the many different compartments of the human respiratory system. Additionally, it is widely known from engineering studies that the airflow inside a specific geometry is strongly influenced by the geometric shape of the air flow pathways. Similar rules are applied in the human respiratory system and its compartments. Therefore, understanding the human’s inhalation/exhalation geometry route is a useful step towards simulation studies of the inhaled/and exhaled pollutants and biological agents (Mutuku et al. 2020a; Mutuku et al. 2020b).
On the other hand, and for the purpose of computational modelling studies, it is useful to know that the lung of an adult man offers the incredible air exchange surface area of approximately 100 m2. The mean lug capacity of an adult man is of 1.5 L (Scharfman et al. 2016) and he is able to inhale and exhale over 10,000 L of air per day while resting (Ni et al. 2020). This huge permeable membrane surface, the lungs, is the means by which the indoor air pollutants are absorbed and diffused by mainly the air mass transfer mechanism into the human body. Specifically, air mass transfer by diffusion via membranes is the key engineering mechanism for not only transmitting viruses trapped in saliva droplets, but also, a variety of other aerosol particles and droplets into the human body (Jayaweera et al. 2020). It is also known that the air mass transfer is enhanced by the increased surface areas available to diffusion and the physiology of a human respiratory system is not only quite complicated in anatomical characteristics, but also offers an excessive total surface to enhance any such transmission of biological agents hosted in indoor air. This creates more serious respiratory problems as penetration of pollutants and biological agents can affect every other organ of the human body via their diffusion in veins and the human blood circular system.
The human respiratory system consists of and connects also the mouth, throat and pharynx with the trachea, all of them often known as Generation 0, according to the human tracheobronchial tree. After inhalation, the larger pollutants or biological agents are filtered by the nose or deposit in the oropharynx, whereas smaller particles, droplets and nuclei are possible to penetrate the deeper than Generation 0 parts of the human respiratory system. The Generation 0 system is further leading to two bronchi, commonly known as Generation 2, with then the different branches of the lungs’ system to be continued down to smaller and smaller compartments of, in total, 23 different generations. The lowest and deeper of them, Generation 23, counting at some millions of the smallest lung compartments, being the alveolar sacks and alveoli (Mutuku et al. 2020a). For example, an adult man’s lung is made from approximately 300,000,000 alveoli (~ 200 μm in diameter) where the supply of oxygen takes place through a rich network of blood vessels (Rhodes 2008). Concerning their characteristic lengths, each of the respiratory system compartments, starting from the nose and mouth and ending in the tiniest lung compartments the alveoli, has substantial different sizes. Those sizes range from 30 to 150 μm, with total lengths between 120 mm and 150 μm. Typical air velocities in the respiratory system are ranging from 9 to 4 × 10−5 m/s, with corresponding residence times of contaminated air being between 0.021 s in mouth and the incredible high residence time of 4 s in alveoli (Mutuku et al. 2020a; Rhodes 2008).
The face anatomy though of each individual person varies and at the same time plays a major role to the biological agents’ transmission. For example, the nasal airways of an adults’ narrowest section is ranging from 5 to 9 mm with a resulting cross-sectional area ranging between 20 and 60 mm2, without taking into account the unique face anatomy of each individual. The nose anatomy, for instance, accounts for the 50% of the indoor airflow resistance and creates a natural resistance to biological agents’ and other pollutants inhalation (Rhodes 2008). The typical airflow through the nasal canals ranges from 0.18 to 1 l/s, from normal breathing to strongly sniffing, respectively (Rhodes 2008). The typical airflow from mouth during normal breathing is 3 m/s and depends, as previously stated, on the physiology of the face and lungs of a person (Rhodes 2008).
Table 1 depicts the main characteristics of the human respiratory tract (size (mm), velocity of air (m/s) and residence time (s)). The specific information might be proven useful for studies on lung damage during inhalation of pollutants and biological agents. In Table 1, it can be seen that by decreasing the characteristic length size of the geometry (higher Generation) of the respiratory system part, there is an incredible increase of the residence time of the biological agents which remain in the different generation parts of the human respiratory system.Table 1 Main characteristics of the human respiratory tract of an adult (basis 60 l/min) along with the generated number of saliva droplets (adapted from Rhodes (2008))
Characteristics—body part Diameter range (mm) Length range (mm) Typical air velocity range (m/s) Typical residence time range (s)
Nasal airways, mouth and pharynx, trachea 5–30 70–120 1.4–4.4 0.021–0.027
Bronchi (main, lobar and segmental) 5–13 28–60 2.9–4.0 0.010–0.007
Bronchioles (main, secondary and terminal) 0.7–2.0 5–20 0.2–0.6 0.023–0.036
Alveolar ducts and sucks 0.3–0.8 0.5–1.0 2.3 × 10−3–7.0 × 10−4 0.44–0.75
Alveoli 0.15 0.15 4 × 10−5 4.0
It is also generally accepted that the respiratory droplets are formed from the fluid lining of the human respiratory track (Mittal et al. 2020), while the biological agents which are dispersed into indoor environments pose a new challenge. This challenge is mainly focused on the understanding of deposition/ diffusion patterns and efficiencies of the infectious aerosols generated from symptomatic and especially asymptomatic patients of infectious diseases (Mutuku et al. 2020a). Shao et al. (2021) stressed out the importance of indoor ventilation system design. More specifically, a properly designed and selected ventilation system is critical for decreasing the transmission risk of infectious diseases, while an inappropriate design can significantly limit the efficiency of droplets removal from indoor air. The local hot spots of biological agents with several orders of magnitude posing higher risks, and at the same time enhancing the droplets deposition causing surface contamination.
The site of droplet nuclei deposition in the lower Generation parts of lungs depends strongly on the droplet shape, size and mass. This transmission route is also dependent on the droplets which are carried in stable and small enough size via indoor air as respiratory droplets of some considerable size or as fine droplet nuclei (Dhand and Li 2020). The very fine droplets and particulates, entering and remaining in the lungs are often an approximate size of up to 7 μm (Jones 1999). In addition, Cheng et al. (2016) found that there was a probability of 50% for the influenza infected nuclei of sizes from 0.3 to 0.4 μm to promote influenza reproduction number (R-0) at values higher than 1, known to increase the risk of transmission of the disease. This only indicates the importance of indoor air biological agents’ size, and how influences their ability to be highly infectious. On the other side, Han et al. (2020) reported that the total dust and the respirable dust should be below 4.0 mg/m3 and 2.5 mg/m3, respectively, to ensure the health and safety of people staying in indoor environments within their usual working timeframe of 8 h. Bourouiba et al. (2014) also reported that tiny droplets and particulates can easily penetrate the respiratory tract, reaching the deeper targeted tissues of the lungs during inhalation of hazardous agents, as shown in Fig. 1.Fig. 1 Schematic representation of a variety of biological (infectious) and chemical agents’ transmission between humans via the airborne route
According to Scheuch (2020), the very fine particles are extremely difficult to separate from the indoor air environment. Those cannot even be effectively deposited in the human respiratory tract compartments, reporting that only 30% of the inhaled particles (0.1–0.5 μm) are deposited in lungs. This means that the rest 70% of the inhaled droplets/particles are exhaled back to the indoor air again. He also claims that while the deposition occurs to a small extent throughout the entire respiratory tract, ranging from nose, mouth to throat, bronchi, bronchiole and alveoli, the preferred site of biological particles deposition is the peripheral area of the lungs.
Aliabadi et al. (2011) indicated that the humidity and temperature of the human respiratory tract varies with the anatomical location of the targeted compartment of the human respiratory system. A temperature, for example, of 37 °C and a relative humidity of 99.5% may be assumed for nasal respiration. For oral respiration the same temperature of 37 °C but lower relative humidity (90%) can be assumed, as well as an increase of the relative humidity by 1% per each Generation of the human airway (branching) until a maximum of 99.5% can be assumed for modelling studies. Varying temperature and relative humidity which prevail in the human respiratory tract are both very important factors due to the impacts on the characteristics of the hygroscopic aerosols, carrying biological or any other chemical agents. As those aerosols inhaled and move along the respiratory tract, their diameter and density might be changing. This is affecting their fate: either those aerosols will be exhaled or end up in deeper Generation part of the human’s respiratory system.
To better understand the importance of the temperature and humidity especially in the survival of biological agents, Zhang et al. (2020a, b, c, d) reported that MERS-CoV exhibited a very strong ability of surviving in air. They indicated that those agents surviving even 1 h after of their atomisation, via a violent for example sneezing, at relative humidity of 79% and ambient temperature of 25 °C. However, when the temperature increased by roughly 10 °C at 38 °C, only 5% survival rate occurred in 1 h when the relative humidity was 27%.
Chemical composition of particles and biological agents
It is widely known that different contaminants and mixtures of droplets present varying physicochemical properties, and those properties affect both the droplets’ and solid particles’ behaviour. The physicochemical characteristics of droplets such as viscosity (μ), density (ρ) and/ or surface tension (σ) affect their shape and characteristic size, among others parameters of the aerosol system (Mandato et al. 2012). Aerosols of human saliva which are infected with viruses, for instance, are primarily composed by water (more than 99% wt), and secondary by traces of enzymes, mucus, white blood cells, enzymes amylase, lipase and antimicrobial agents lysozymes (Al Assaad et al. 2020; Sarkar et al. 2019). Gralton et al. (2011) reported that an increase in the droplets’ size made from saliva and release in indoor air environment is directly related to an increased mucus viscosity.
In the literature as already mentioned it is common to simulate the aerosol droplets of saliva including water (Bourouiba et al. 2014; Liu et al. 2019a, b, c, d, e). However, water has a density of 1,000 kg/m3, viscosity of 10−3 Pa·s and an interfacial surface tension of 0.0728 N/m (Viswanathan 2019) at ambient indoor air conditions, while the saliva has a viscosity 86 to 150 × 10−3 Pa·s and interfacial surface tension of 0.05898 N/m (Sarkar et al. 2019). In the case of droplets’ formation during a coughing incident, the quality of saliva, which is different between a healthy person and a patient, will impact the droplets behaviour. This is done by strengthening the elasticity of the droplets and their resistance into their breaking up to smaller nuclei droplets and residuals, while releasing in the indoor air. As a result the saliva droplets will be more resistant to break, forming a lowest number of fine droplets and fewer droplets of a large size (Zayas et al. 2012). The droplets formed by a respiratory event of a patient can unfortunately be at the same time carriers of a biological agent due to their illness. In addition, contaminated droplets travelling in air might attract other (i.e. chemical contaminants being present in the confined indoor environment). As a result, another healthy person (recipient) can be infected via the unconscious inhalation process (Fig. 1) (Vadivukkarasan et al. 2020).
Similarly, for other types of indoor air contaminants, chemical analyses and characterisation play an important role on understanding their physicochemical characteristics. For example, droplets of tobacco smoke are made only from 20% wt water among the rest several thousands of different traces of their tar constituents (Ni et al. 2020). It is obvious that such properties will be different in nature biological agents and those should be taken into account when modelling the routes of transmission for indoor air agents. Balachandar et al. (2020) claim that although the surface tension of saliva droplets measured similar to that of water, their viscosity can be 1 to 2 orders of magnitude larger than that of water, resulting in making those droplets less coalescence prone.
Shape of particles and biological agents
Another important characteristic of pollutants and biological agents is their shape which has a strong influence on droplets’ and particles’ size (Rhodes 2008). The shape of a particle affects its properties such as the surface area per unit volume (m2/m3) and/or the rate at which particles in general settle in indoor air environments (Rhodes 2008). Defining the droplet, and especially the solid particles’ shape, is dependent on their real shape, the availability and suitability of the analytical methods for their shape determination. More specifically, in the case of droplets their chemical composition has a great impact on their characteristics such as density, viscosity and the forces imparted on particles, while they are expelled and move in the indoor air.
Particles in general, and for the sake of modelling and simulation studies, are usually assumed to be represented by spheres in a 3-D system or circles in a 2-D system, respectively. However, very rarely particles maintain a spherical shape and a uniform size. In practice particles’ shape, either those being plain chemical pollutants such as ash or biological agents, their shape is usually far away from that of a perfect sphere. Simulating solid particles as spheres might not be realistic and thus the dimensionless number of sphericity (φ) is used to determine how far away the shape of a real particle is from the perfect spherical one. Sphericity is defined as the ratio of the surface area of an equal in volume sphere with the real particle to the surface area of the real particle. Sphericity values of particles are always ranging between 0 and 1, with the value of 1 to represent the sphericity of the perfect shape that of the sphere (Rhodes 2008).
When also a droplet or a particle falls freely in air, under the action of gravity, and an indoor air stream blows at an angle, several forces acting on the droplet/particle. Those are the gravitational force due to the mass of the particle, the buoyancy force due to the movement of the particle in a fluid, as well as the inertia and drag forces which oppose the travel direction of the particle. The balance of all these forces imparted on a particle will dictate the terminal velocity by which the particle or droplet of a final stable size, for the latter, will settle in indoor air (Soni et al. 2020).
In fluid dynamics studies, the dimensionless numbers are very useful in analysing the fluid flows, especially the multi-phased ones, where there is an interface between different fluids (gas–gas, gas-liquids). A widely used dimensionless number for this type of flows is the Weber number (We). We number indicates how the shape of a droplet will be in a certain fluid system or when deposited on a surface. Thus, it measures the relative importance of the inertia over the surface tension force and is mainly used to demonstrate the different break-up modes of the droplets and, as the result, the shape of the droplets. We number can also be used in describing the influence on the surface wettability under the effect of droplets. According to Liu et al. (2019e), when the value of We number is less than 0.5, droplets impact differently processes on hydrophobic, hydrophilic and super-hydrophilic surfaces which are dominated by the spreading stage and retraction is not evident.
At low We number, a droplet undergoes shape oscillations at a certain frequency (Fig. 2). As the We number increases slowly, by increasing the aerodynamic force applied on a droplet and keeping the surface tension force constant, the droplet exhibits a transition from the vibrational mode to the ‘bag’ break-up mode of droplets. When the We number is low the droplet tends to maintain its shape. On the other hand, high values of We number along with increasing the aerodynamic forces imparted on the droplet lead to the loss of the almost spherical shape of the droplet and create a ‘bag’ deformation and breakage, which also forms several smaller satellite droplets of smaller dimensions (Soni et al. 2020).Fig. 2 Schematic representation of the vibrational, transitional and bag deformation shape changes of water droplets travelling in air adapted from (Soni et al. 2020)
During the release and travelling of the formed droplets in the air, they interact with their host medium and alter their shape as move along with air, especially at high speed airflows. It is also known that high speeds prevailing when a person coughs or sneezes. Hence, the droplet shape changes depend on different mechanisms such as the vibrational changes of droplets (We = 5.13), transitional towards a bag shape, bag-stamen, dual-bag, multi-mode, shear and catastrophic break-up (We = 6.35) modes, according to the work of Soni et al. (2020). Those transitional areas of the droplet shape-change depend on the conditions under which the experiment is taking place. This spherical shape is changing rapidly in a ‘bag’ shape and breaking via ligaments with the production of finer satellite droplets based on the surface tension and the aerodynamic forces applied on droplets. Relatively little attention has been given though to the instabilities associated with the dynamics of respiratory droplets creation and expelling during especially the coughing or sneezing incidents (Vadivukkarasan et al. 2020).
Size of particles and biological agents
The size of particles, either being solid particulates or liquid droplets, is determined by their characteristic length (size). The size of solid particles very rarely depends on the ambient indoor conditions (temperature, humidity). It also depends on their natural shape and morphology, and their chemical composition (Rhodes 2008). In addition, the particulates found in nature or produced buy processes very rarely possessing the perfect shape of a sphere. Real particles quite often have irregular shapes such as acicular, flaky, spongy or any other shape.
As a result, the size characterisation of solid particles is easier compared to droplets even though their shape is not spherical. The most appropriate characteristic length then for solid particles, instead of the diameter of a perfect sphere, it might be a different size such as the equivalent circle diameter or the surface to volume diameter and Sauter mean diameter and others (Rhodes 2008). All these characteristic lengths are used to describe the real size of a particle in conjunction with their non-spherical shape and real surface area, while they are moving in a fluid under aerodynamic forces. The measurement of the characteristic diameters is achieved by analytical methods such as the scanning electron microscopy (SEM), electro zone sensing, permeatry and other less known analytical and optical methods (Morawska et al. 2009; Rhodes 2008).
On the contrary of the stable size of solid particulates, droplets’ size is not unfortunately remaining stable upon released in indoor air and the droplet size is highly dependent on indoor air conditions. When a liquid is atomised an aerosol of droplets is produced, with those droplets to usually keeping their initial spherical shape for only a short period of time after their formation. Their shape depends on several factors which have to do with the droplet’s physicochemical characteristics and environmental conditions of the indoor space where they are dispersed and move.
The size of droplets highly depends on their formation process with fine ones of less than 1 μm to be produced from engineering manufacturing processes and larger up to 100 μm from mechanical processes (Morawska 2006). It also depends on especially the humidity and temperature (Dhand and Li 2020; Gralton et al. 2011) of the indoor air. The diameter of the droplet is a dynamic property due to the liquid evaporation under certain indoor air conditions. Those conditions are resulting in droplet shrinking by time which finally leads to the formation of the stable droplet nuclei (Ji et al. 2018; Li et al. 2018; Liu et al. 2019a, b, c, d, e; Liu et al. 2017; Morawska et al. 2009; Wang et al. 2019a, b; Wei and Li 2015; Xie et al. 2007; Yang et al. 2018).
In the case of droplets, it should be also considered the effect of droplet’s evaporation (Ji et al. 2018; Li et al. 2018; Liu et al. 2019a, b, c, d, e; Liu et al. 2017; Morawska et al. 2009; Wang et al. 2019a, b; Wei and Li 2015; Xie et al. 2007; Yang et al. 2018). A characteristic example of the effect of the relative humidity (RH) of air in water droplets of 50-μm diameter is that they will evaporate at RH = 50% in less than 3 s (Vuorinen et al. 2020). Droplets also under favourable humidity conditions may even increase in size due to attachment of the surrounding humidity of air on them. As a result, the droplet size varying not only with time, but also depends highly on the environmental conditions of temperature and humidity.
On another aspect the initial formation mechanism of an aerosol of droplets occurs due to mainly water vapour condensing onto the cloud of initial nuclei. This condensation occurs only when air contains slightly more water vapour than it normally holds for a given temperature. Vuorinen et al. (2020) indicated to the importance of understanding what are the humidity supersaturation conditions of atmospheric air and their nuclei, which promote cloud droplet nucleation and growth. Carducci et al. (2020) reported that the different expiratory events such as coughing, sneezing, speaking, singing and simple breathing release droplets of sizes ranging between 1 and 2000 μm noticing, however, that the majority of them has a size between 2 and 100 μm.
Recently, Dhand and Li (2020) indicated that the size of the droplets expelled by a patient mainly depends on their site of origin from their respiratory systems. For example, droplets which are produced by the mouth (oral cavity) have a large size (~ 100 μm), while smaller droplets (~ 1 μm) are formed during talking and coughing. The difference in size of droplets is due to the fact that the smaller droplets originate from the bronchioles, while the larger droplets are generated during normal breathing and from the larynx during talking and coughing. It was also reported that the particle size distribution could be altered by the presence of viruses (Dhand and Li 2020).
The droplet size determination is usually taking place via optical methods and laser analysis (Stadnytskyi et al. 2020; Tang et al. 2009). Ni et al. (2020) reported that recent studies have demonstrated that particulate matter (PM2.5) is closely associated with the chronic lung diseases and special attention should be given to biological pollutants of this specific size range. However, special attention should also be given to the fact that only few studies have conducted with modern techniques, capable of detecting sub-micrometric size particles. Thus, it is necessary to undertake further studies in order to develop a better understanding of the formation mechanisms of fine droplets (Morawska 2006).
Size distribution of a large population of particles and biological agents
The accurate characterisation of a large population of droplets/particulates can be done by investigating their size distribution within the multi-phase cloud of particles. This size distributions changes with time and distance from the source of generation depending on environmental factors, too (Dhand and Li 2020). This can be achieved by three ways and depends on the nature of droplets/ particles. For example, a droplet of an agent, infectious or not, in equilibrium with the environment has a stable size as cannot shrinks or increases in size. The later can be determined as per their particle size distribution based on mass, or surface area or number of particles (Rhodes 2008).
Concerning the aerosols and the size distribution of particles there is a threshold distance of approximately 1.5 m, which distinguishes the two basic droplet and droplet nuclei transmission processes, namely (a) the short-range mode and (b) the long-range airborne route. The short-range mode of transmission includes the conventional, large droplet routes of parabolic travel under the effect of gravity, as well as the newly defined short-range airborne transmission (Liu et al. 2016a, b). However, Pendar and Páscoa (2020) reported lately that the infectious saliva droplets can travel up to 6 m at a wind speed of 15 km h−1and a safe distance of 2 m is not appropriate for outdoor activities.
A large number of studies highlights the importance of the size distribution regarding the particles and biological agents, as well as the occupants in indoor environments (Choi et al. 2015; Cole and Cook 1998; Dhand and Li 2020; Faridi et al. 2020; Faulkner et al. 2015; Feng et al. 2020a, b; Fernstrom and Goldblatt 2013; Ghosh et al. 2015; Gralton et al. 2011; Lv et al. 2018; Milton et al. 2013; Monn 2001; Morawska et al. 2009; Nicas 1996; Nielsen 2015; Phu et al. 2020; Sajjadi et al. 2016); Scheuch 2020; Schroeter et al. 2012; Vianello et al. 2019; Wang and Yoneda 2020; Yang et al. 2016).
Lv et al. (2018) indicated that the supply flowrate of fresh air per unit of closed space volume, defined as air changes per hour (ACH) is also an important factor which influences the indoor particle distribution. They found that the free settling of particles into indoor space for particles ranging from 0.5 μ to 1.0 μm, 1.0 to 3.0 μm and 3.0 to 5.0 μm, presenting a sedimentation rate of 0.086 h−1, 0.122 h−1 and 0.173 h−1, respectively. The same researchers reported that an increase of ACH from 0 to 2.5 yields significantly different values on the sedimentation. Recently though, special attention is given to studies with reference to the size distribution of droplets and the improvement of measurement accuracy for small scales below micrometre range. For instance, a droplet size distribution for coughing indicates a peak drop size of almost 15 μm while the associated settling speed obtained at 6.5 mm/s in an ambient winter indoor air (Bourouiba et al. 2014).
Han et al. (2020) stated that there are several empirical equations to characterise the droplet size distribution such as Nukiyama-Tanasawa, Rosin–Rammler, log-normal, root-normal and log-hyperbolic. Poon et al. (2020) found that the droplets produced by coughing present a wide size distribution of droplets ranging from 0.6 to 16 μm, with a mode of around 6 μm. Lately several studies have been devoted to the size distribution of small droplets expelled during talking, coughing and sneezing; however, uncertainties on the droplet size distribution are still present (Asadi et al. 2019; Scharfman et al. 2016).
The airborne route of transmission of particles and biological agents
The droplet or aerosol airborne transmission route seems to be the most complicated mode of dispersion of particles, droplets and biological agents into indoor environment (Dhand and Li 2020; Ai et al. 2019a, b; Ai and Melikov 2018; Aliabadi et al. 2011; Beggs 2003; Booth et al. 2005; Drossinos and Stilianakis 2020; Monn 2001) and as a result remains one of the most difficult aspects to study. Aerosols of particulates and droplets pose a major challenge: being invisible in human eye, they are transported as a cloud of submicron-sized particles generated especially by coughing and sneezing via a process which is called atomisation in engineering practise, or trapped in liquid micro-sized water droplets (Vuorinen et al. 2020) or even drifted away by being attached on solid particulates (e.g. dust and pollen) (Griffin 2007).
The airborne transmission is further classified as short and long range, with most of the scientific community to be still unclear on the determination of the safety distances need to be kept to avoid infections. This becomes even more unclear considering especially infectious diseases which have the ability to spread in short and long diseases (Bourouiba et al. 2014) and under the two most widely known modes of transmission the short and long one. It seems that the most common indirect transmission route is occurring via spreading of an infected cloud of small saliva droplets (aerosols) during talking, coughing, sneezing or breathing (Gralton et al. 2011; Tang et al. 2009; Zhao et al. 2005). Lately, Godri Pollitt et al. (2020) demonstrated that the short-range airborne route of infection may be the most common transmission way of infectious diseases. Carducci et al. (2020) also refer that droplets up to 5 μm, fall next to the donor source, within a distance of approximately 1–2 m, due to the effect of the gravitational force prevailing on the large droplets. The smaller aerosols though can remain suspended and travel at greater distances in the indoor air environment. More information on the aerosol’s nature, generation and behaviour can be found in next sections.
Aerosols of particles and biological agents
An aerosol is defined as a population of submicron particles or a suspension of droplets and droplet nuclei in the air. An aerosol of droplets is usually created by a violent respiratory event such as a cough or sneeze (Sakharov and Zhukov 2020). Jayaweera et al. (2020) claimed that up to 90% of the aerosol droplets generated by a human expiratory activities. Since aerosols are particles or biological agents of less than 50 μm, they remain suspended into indoor air due to their small size for extended periods of time. The larger airborne particles (> 50 mm) are too heavy to become suspended in the air for longer periods of time (Marui et al. 2019). In addition, the droplet nuclei residuals remain into indoor air at a fine and stable size, in the range of 5–10 μm (Bourouiba et al. 2014). This final stable size of the residual droplets/nuclei is determined by the equilibrium with the moisture of ambient indoor air (Vuorinen et al. 2020). The dynamic reduction in the size of the infectious droplets leads to a change in the pattern of transporting in air, depending also in the indoor air currents, humans moving and talking, coughing or sneezing all known to be able to create a laminar or event transient and turbulent flow of the aerosols in confined spaces.
Many researchers study how the diameter of the liquid droplets changes dynamically and strongly affected by the temperature and relative humidity (RH) of indoor air (Aliabadi et al. 2011; Dedesko and Siegel 2015; Faridi et al. 2020; Shajahan et al. 2019; Verijkazemi et al. 2018; Zhang et al. 2019). Aerosols of less than 1 μm, with the lowest density are generated by nasal breathing, while the highest density by coughing in very short time (up to 500 ms) (Bourouiba et al. 2014). Exhaled breath is also more responsible for transmitting viruses of size of approximately 0.1 μm, compared to the bacteria transmission with particle size over 1 μm (Zhang et al. 2019). From the above-mentioned, it is evident that all the above factors, chemical composition, shape and size of droplets are interconnected.
The main characteristics of an aerosol depend on the characteristics of the single droplet and the forces imparted on them as the move along with the air currents (Rhodes 2008). The shape, and as a result size, of droplets depends on the spray/aerosol angle, covering of surface, droplet velocity distribution, volume distribution and pattern is different for different aerosol systems (Broniarz-Press et al. 2009). Some physicochemical properties of the droplets, such as viscosity, might vary, and depend on the fluid environment where the droplets are hosted (other liquid or air environments). For an aerosol of droplets in air, for instance, the relative viscosity of the liquid compared to the surrounding gas viscosity is high (50%), while in a liquid host is relatively low (Ben-Tzvi and Rone 2010).
In general, the larger the droplets and particles are, the quickest they settle and in a shortest distance they travel, as this will determine how far the particles will be dispersed. This is based on the force by which they are expelled from the source, either the source being a person or a ventilation equipment. It is widely acceptable that the respiratory droplets evaporate to form smaller droplet nuclei, remain then suspended in air due to Brownian motion, and susceptible individuals from the source could inhale them even when stand far away. Scheuch (2020) indicated that for small particles, the main mechanism of their transport in air is the Brownian motion and this mechanism works relatively effectively with droplets size in the range of 5–100 nm. Scheuch (2020) stated that the second important physical mechanism of eliminating particles from the indoor air is sedimentation. This mechanism is effective for aerosol particles above 0.5 μm – 1 μm. Stilianakis and Drossinos (2010) indicated that all droplets generated by an expiratory event, either this being coughing, sneezing, laughing, talking or breathing cover a large size range from approximately 0.6 to more than 1000 µm.
Atomisation of liquids
Atomisation is the process of formation of fine droplets, or an aerosol of droplets or biological agents in the case of indoor environments (Morawska 2006). The atomisation as a process creates small fractions of the liquid droplets affecting considerably other pollutants emission and spreading (Urbán et al. 2017), especially in indoor spaces. Ai and Melikov (2018) reported that the techniques of producing aerosols are increasingly been used to investigate airborne transmission of biological and chemical agents.
For example, a sneezing or violent coughing incident in terms of engineering is a large-scale atomisation process and formation of an aerosol of saliva droplets and nuclei. The atomisation as a mechanical process is affected by the geometry of the source, the aerodynamic forces imparted on particles, the surface tension and viscosity of the droplet. The aerodynamic forces are of considerable effect on the droplet or particle, while travelling in the air with the dominant being the gravitational forces or mass body forces which are imparted on relatively large particles. Thus, larger droplets settle quickly and the smaller airborne droplet nuclei are travelling over longer distances by the indoor air streams (Dhand and Li 2020). The drag force being also opposite to the gravitational force leading to the resistance in motion of droplets/particles in air. The surface tension, too, is the natural tendency of a liquid droplet to stabilise the shape of a droplet of a certain volume, offering the minimum surface area possible. The surface tension has a consolidating influence, which contradicts with the opposite tendency of the surface of the droplet to extent and wet a surface. The viscosity is a property which describes the rheological properties-behaviour of a fluid, and is opposing any change of the shape of the liquid droplets as they flow (Morawska 2006).
Atomisation is further classified as primary, upon injection of droplets and particles i.e. by a person sneezing, and secondary atomisation (Kuznetsov et al. 2019). The secondary atomisation takes place by the droplet size disruption due to interference of a solid surface such as a collision with a wall or a substrate (e.g. hand in front of the mouth while sneezing). This creates a second wave of atomisation due to the fact that the single cloud of droplets colliding with each other, a micro-explosive break-up of droplets is taking place, especially under the effect of the increased temperature and heat, as well as the interference of an existing indoor air stream flow. Han et al. (2020) indicated that increasing the mean air velocity results in larger aerodynamic forces which reduce the droplet sizes, while an increase in air pressure reduces the droplet size. The same researchers (Han et al. 2020) reported that the droplet size distribution is a crucial parameter of the atomisation process besides the mean diameter of droplets.
Suspension and resuspension of particles and biological agents
Suspension time of indoor pollutants is defined as the time that small droplets or particles remain suspended on air, carried away at short or long distances due to airflow motion and without necessary settling on horizontal surfaces such as the floor. Their velocity also plays an important role on the analysis and simulation of the aerosol systems and their suspension time. The effects of gravity or inertia forces on droplets of less than 30 μm are negligible as they are too small in size; their transmission then is mainly influenced by the indoor airflow as those particles remain suspended for long time and as a result the risk to be inhaled is high (Zhu et al. 2006). Results of studying a coughing incident showed that more than 6.7 mg of saliva are expelled as droplets exhibiting a velocity up to 22 m/s, while at the same time a travel distance of more than 2 m has been reported (Zhu et al. 2006). On the other hand, droplets with their size range varying from 50 to 200 μm are of significant size in terms of importance. Those are affected by gravity and fall on the ground as the indoor air flow streams are weakening. Droplets of diameter of 300 μm or larger, which are mostly affected by inertia forces rather than gravitational, rarely fall (Zhu et al. 2006).
In general, the evaporation rate of droplets depends mainly on the ambient temperature and humidity. It was found that droplets of size less than 100 μm will typically become droplet nuclei before settling on the floor. Small droplets of sizes between 5 and 10 μm will rapidly evolve into droplet nuclei with extremely low settling speeds (> 0.003 m/s). As a result those droplet nuclei are able to remain suspended for longer periods of time, however, the fate of droplets are determined by the competing effect of inertia, gravity and evaporation (Mittal et al. 2020). At the same time the nuclei are expected to be crucial in the long-range airborne transmission route. Bourouiba et al. (2014) also highlighted the synergistic effect of Brownian motion in the phenomenon of suspension and resuspension of particles, where air currents are absent. The same mechanism may keep the stable in size droplet nuclei suspended for very long periods of time in such environments.
The resuspension of particles into indoor spaces is the phenomenon of the detachment of deposited particles and droplets of other pollutants or biological agents from the surfaces into the bulk air (Al Assaad et al. 2020). The reason of resuspension is usually the human activities such as walking and natural or mechanical ventilation. All these actions cause the aerodynamic and mechanical vibration disturbances of the particles. It seems that particle resuspension takes places within a very narrow time frame of less than 25 s, since the initial disturbance, prior further decreasing to negligible values (Al Assaad et al. 2020).
For different indoor open surfaces, it was found that the resuspension was the lowest for smooth surfaces such as glass, followed by marble and linoleum. When though the aerodynamic disturbances applied on those surfaces were accompanied with vibrations the resuspension of particles increased by more than 45% for all cases (Al Assaad et al. 2020). It also seems that a decrease in the roughness of the indoor space surfaces can increase the particles and droplets adhesive forces reducing considerably the vibration effects which are responsible for enhancing the resuspension in air (Al Assaad et al. 2020). For example, dust is re-suspended when people walking on carpets and has been found that the mass load of dust is generally greater in carpets than the hardwood floors (Haines et al. 2020). They reported other pollutants such as stain-protectors which were found not only in the carpet, attached to dust, but were also detected in the blood serum of the occupants (Haines et al. 2020). The same researchers found that the man-driven resuspension of particles previously settled on carpets and hard flooring is a source of coarse-mode biological agents’ pollution. When an adult, for instance, is walking across the floor, this can create a resuspension of 10 to 100 million particles per minute, many of which are likely to be of biological origin. For particles thought of less than 10 μm mass resuspension rates can exceed 10 mg/min (Haines et al. 2020).
In addition, indoor environmental conditions of temperature, humidity and air streams should not be underestimated, as it was found that 50% of the airborne biological agents could originate from the resuspension of fungi grown at equilibrium relative humidity of more than 85% on dust floor (Dannemiller et al. 2017). You and Wan (2014) based their findings both on experimental and modelling results. They showed that Bacillus anthracis particles’ concentration becomes 1.5 to 3 times and 4 to 8 times higher after the initiation of airflow for particle of sizes between 2 and 4.75 μm. Their study indicated clearly the importance of the airflow to the resuspension of particles.
Evaporation, coalescence and growth of droplets
The evaporation of droplets plays an important role in the later fate of the droplet and competing effects of inertia and gravity. The evaporation rate depends on the difference between the droplet surface saturation vapour pressure and the vapour pressure of the surrounding air, which also depends on the humidity (Mittal et al. 2020). The diffusion mechanism strongly affects the droplets surface-to-temperature difference, and the relative velocity between the droplet and surrounding gas. Thus, dimensionless numbers such as the Sherwood (Sh), Nusselt (Nu) and Reynolds (Rep) for the droplets are important to determine the evaporation phenomenon. It seems that higher temperature and lower relative humidity lead to larger evaporation rates that increase the critical droplet size (Mittal et al. 2020). The temperature effect initiates the evaporation of atomised liquid droplets affecting the overall motion and distribution of droplets. Sakharov and Zhukov (2020) indicated that smaller droplets, 5 μm, would evaporate in less than 3 s, at typical indoor relative humidity of 50%.
Evaporation is a very fast molecular process, for instance, a 20-µm droplet evaporates to 1-µm diameter droplet within only a rate of 0.24 s−1 at 50% ambient relative humidity (Yang et al. 2018; Ai and Melikov 2018). Due to the evaporation phenomenon, the size of the droplets is affected by time, as they are shrinking and this is prominent for droplets with an initial diameter of 100 μm (Yang et al. 2018). Wells (1934) although has already found that by the beginning of the twentieth century, droplets with characteristic diameter larger than 100 μm settle to the ground in less than 1 s, without being significantly affected by evaporation. Similarly Morawska et al. (2009) did not detect droplet evaporation for particle sizes varied between 0.5 and 20 μm, and, if any evaporation occurs, take place at less than 1 s. Studies of water droplets with diameters of 10 to 240 μm indicated that the medium-sized droplets vary from 50 to 170 μm, as the thermal stratification weaken the evaporation of droplets due to less heat and mass transfer between the droplets and air. When the ambient relative humidity increased to 60%, a possible condensation phenomenon occurred on droplets, increasing the suspending time of droplets in the air (Liu et al. 2019a, b, c, d, e). In addition, vapours generated due to evaporation and super-saturated wet air exhaled from the respiratory tracks form a vapour plume in front of the nose and mouth of a person, which, despite the short life time enhances significantly the evaporation of the droplets captured in it (Li et al. 2018). Due to the evaporation and density of airborne droplets and mass concentration of inhalable pathogens, the process can result in a higher risk of infection (Li et al. 2018). The study of Li et al. (2018) demonstrated the importance of considering inhomogeneous humidity field when modelling the evaporation and dispersion of cough droplets.
Droplets might collide with each other and can undergo coalescence. Droplet coalescence is the process of merging of two or more droplets during contact to form a single larger droplet. If droplets are hydroscopic they grow in size or while transported in air might trap particulates such as dust (Han et al. 2020; Morawska 2006). As a result, the coalescence mechanism leads to a change of the particle size distribution with the mode value of droplets to increase as the total number of particles decrease (Morawska 2006). Shao et al. (2021) reported that the viscosity and surface tension of droplets might be of significant importance. They influence the droplet size distribution as both controlling the coalescence and breakage of larger droplets to smaller. However, these mechanisms are important only during the ejection stage of the infected saliva droplets. Once the infected saliva droplets are below 50 μm, the coalescence and break up mechanisms are hindered. Occasionally, the particles may shatter apart into numerous smaller particles; however, this process usually occurs primarily in large particle size droplets, which cannot be considered as aerosols (Shao et al. 2021).
Aerosols and bioaerosols
An overview of airborne particle types that affect respiratory health
As previously discussed, the vast variety of abiotic (chemical agents) and biotic (biological agents) particles being present in air at considerable concentrations can have a negative effect on human respiratory system or human health in general. Such particles are usually present in the form of aerosols which either travelling or being suspended in air. As defined in the ‘Aerosols of particles and biological agents’ section, an aerosol is a suspension of fine solid or liquid particles of varying sizes in air (Fig. 3).Fig. 3 The size ranges of air particles and microorganisms
Bioaerosols can be defined as the particulate matter usually associated with compounds of pure biological origin. This definition includes all pathogenic or non-pathogenic media ranging from live or dead fungi and bacteria, viruses, high molecular weight allergens, pollens and many others (Ghosh et al. 2015). The main type of aerosols being of a significant concern for human health is the plume of droplets of micron size that are scattered in the air during breathing, talking, coughing or sneezing (see the ‘The human respiratory system’ section). As these droplets can stay suspended in the air for many minutes and contain pathogenic microorganism that can lead to respiratory diseases (Bourouiba et al. 2014; Cole and Cook 1998) (Fig. 4). Aerosols of biological agents can be also created mechanically by other ways such as emerging from water fountains, shower heads, surgical or dental procedures, as well as faulty air-conditioning or ventilation systems (Tran et al. 2012).Fig. 4 A donor-recipient model of transmission of respiratory pathogens within droplets
As discussed previously, the size of these droplets is a very important factor affecting the transmission of respiratory diseases. Usually droplets’ size range from 0.01 to 500 μm, although larger droplets have also been reported (Gralton et al. 2011). According to Guzman (2020), only droplets smaller than 5 μm are able to reach the trachea of the recipient, while droplets below 2.5 μm can penetrate to the lower respiratory system and reach the bronchioles and alveoli inside the lungs (see ‘The human respiratory system’ section). Aerosols smaller than 5 μm are considered to be airborne means of disease transmission, since they stay in the air for long periods of time, while larger aerosols are linked with droplet-associated transmission of diseases (Gralton et al. 2011).
The spread of pulmonary aerosols is a major public health concern, especially for indoor environments of hospitals and other healthcare units, where patients often have a weak immune system and at the same time multi-drug microbial pathogens might be present (Stockwell et al. 2019);Tang et al. 2006).
A second type of particles that could be potentially harmful, even though not of biological origin, is related with dust. Dust particles in domestic surfaces, such as floors, furniture or carpets (Haines et al. 2020), may also be contaminated by microbial pathogens (Dannemiller et al. 2017), inducing allergic reactions or worsen the symptoms of an already pre-existing asthma condition. Inhalation of household dust, which contains a variety of aeroallergens, can worsen the symptoms of allergies and asthma. House dust particle sizes range from 2 mm to 63 μm, with approximately 33% of the dust being smaller than 500 μm (Lanzerstorfer 2017). Examples of such allergens include the house dust mite (HDM) protein Der p 1, Can f 1 (associated with dogs) and Fel d 1 (associated with cats). Dust particles < 5 mm tend to remain suspended in the air for a number of days, whereas larger particles (> 5-mm diameter), which remain airborne for a shorter period after disturbance (Hussain et al. 2019). The dust mite itself has a diameter of 200 μm and it is considered too large for penetrating the lungs, however a small proportion of its faeces that are rich in Der p 1 can enter the lungs and cause allergy symptoms (Wilson and Platts-Mills 2018).
House dust particles can also absorb harmful microbial volatile organic compounds (MVOCs). Exposure to low levels of MVOCs in indoor air is related to a range of non-specific symptoms, including redness of the eyes and irritation of the nose and skin, that are known as the sick building syndrome (Wady and Larsson 2005). Other types of dust that could enter inside buildings via open doors or windows include sand particles, farm and coal mine dust and they can all lead to serious lung damage (Khan and Strand 2018; Penconek et al. 2019; Schuijs et al. 2015).
Fungal and bacterial spores can also lead to development of serious lung disease (Cutting and Ricca 2014; Foster et al. 2017; Han and Weiss 2017). Several microorganisms such as fungi (e.g. Aspergillus fumigatus) and bacteria (e.g. Bacillus anthracis) form spores. These are resistant structures with thick cell walls of several layers that provide resistance against extreme environmental conditions, such as adverse temperatures, drought and chemical biocides (Leggett et al. 2012; Madsen et al. 2016). These spores can be easily dispersed in the air, outside aerosols and become inhaled by humans. After inhalation, they end up in the lungs where they germinate and colonise the tissues of the human respiratory system, if they are not controlled by the immune system (Husman 1996). Bacterial spore sizes vary from 0.8 to 1.2 μm (Carrera et al. 2007), while fungal spores range from 2 to 4 μm (Madsen et al. 2016). Fungal spores and vegetative fragments can also be allergenic, bearing a variety of allergens such as Asp f 1, Alt a 1 and Cop c 1 (Crameri et al. 2006; Green et al. 2006). Anthrax spores formed by Bacillus anthracis are considered to be a highly persistent and lethal type of bioterrorism agent, therefore they are a major biosecurity concern, especially for indoor environments, such as offices or schools (Taylor et al. 2012).
Finally, plants produce pollen, which is a powdery substance consisting of pollen grains that contain the male gametes (sperm cells) of the plant. Such particles have a rigid thick exterior layer which protects the genetic material of the gamete. Pollen size ranges generally from 20 to 60 μm (Mander 2016; Rantio-Lehtimäki et al. 1994; Soares et al. 2018). There are, however, exceptions such as Pinaceae pollen which can be of size over 80 μm (Smith et al. 2014). Pollen grains can also travel long distances in air and are known to contain allergenic proteins inducing hay fever and asthma exacerbations. More than 150 different pollen allergens have been identified so far (Mothes and Valenta 2004); Rodríguez et al. 2007); White and Bernstein 2003). The most common ones are the Phl p 1 and Lop p 1. Unfortunately, allergic reactions to pollen represent the most frequent type I allergies affecting up to 30% of the industrialised population (Biedermann et al. 2019; D’Amato et al. 2007; D’Amato et al. 1998). Climatic changes are expected to influence the duration as well as the intensity of pollen seasons which might in hand with air pollution contribute to increased numbers of respiratory allergy and asthma (Pablos et al. 2016).
Major respiratory microbial pathogens and health effects
Numerous infectious agents lead to serious respiratory illness or even death. These belong to three major classes of microorganisms, namely viruses, bacteria and fungi (King and Auger 2002; Prat and Lacoma 2016; Rath et al. 2017) (Fig. 5). Viruses are not considered to be living organisms, as they do not have a metabolism and are unable to replicate outside a host cell. Their viral genetic material is usually protected by a protein capsule. Several viruses are also surrounded by a lipid envelope (Weber and Stilianakis 2008). Bacteria and fungi are living organisms. The morphology of these microbes is extremely diverse in nature, but again the genetic material is enclosed by a lipid membrane and a polysaccharide cell wall. On their surface, these agents have receptors enabling them to attach to human cells and potentially invade into the human cells. In terms of pathogens sizes viruses typically range between 20–300 nm, bacteria 1.0–5.0 μm and fungal cells 2–30 μm (Choudoir et al. 2018; Shi and Tarabara 2018; Weiser 2013) (Table 2). Some bacteria and fungi are able to build long filaments up to several centimetres (cm), while some fungi can form much larger structures in nature (e.g. mushrooms). As discussed in the previous section, the respiratory pathogens usually spread through the air via coughing or sneezing (Barmby and Larguem 2009; Srivastav et al. 2018; Xie et al. 2009), as well as being transmitted by touching contaminated surfaces and then touching the eyes, nose or mouth (Deacon 2006; Madigan 2009).Fig. 5 Images of key respiratory pathogens: a SARS-CoV-2, b Mycobacterium tuberculosis and c Aspergillus fumigatus.
Source: Public Health Image Library, CDC-USA
Table 2 Details of key respiratory pathogens in relation to their pathogenicity
Species name Size (μm) (Collier et al. 2000; Murray et al. 2013) Disease(s) (Collier et al. 2000; Murray et al. 2013) Duration of survival on surfaces (h) (Kramer et al. 2006) Minimal infectious dose (# of particles/cells) (Yezli & Otter 2011)
Rhinovirus 0.03 Common cold Up to 7 days 10
Influenza virus 0.08–0.12 Flu 24–48 h (Bean et al. 1982) 1,000
SARS virus 0.05–0.20 Respiratory syndrome 24 h (M. Y. Y. Lai et al. 2005) 280 (Watanabe et al. 2010)
MERS virus 0.10 (Hajjar et al. 2013) Respiratory syndrome 8–48 h (Kampf et al. 2020) 1,000 (Douglas et al. 2018)
SARS-CoV2 (COVID19) 0.60–0.14 (Dhama et al. 2020) Respiratory syndrome 84 h (Hirose et al. 2020) 100 (Ryan et al. 2020)
Respiratory syncytial virus 0.15–0.25 Common cold 6 h Unknown
Parainfluenza virus 0.15–0.25 Respiratory illness in children 4–10 h (Henrickson 2003) Unknown
Streptococcus pneumoniae 0.5–1.25 Pneumonia 20 days 5 × 106 (Dietert et al. 2017)
Haemophilus influenzae 1.00 Pneumonia 12 days Unknown
Legionella pneumophila 3.00–5.00 Legionnaire’s Disease 2 h (Katz & Hammel 1987) 100,000 (Gama et al. 2012)
Mycobacterium tuberculosis 2.00–4.00 Tuberculosis Up to 4 months 10 (Gama et al. 2012)
Acinetobacter baumanii 0.90–1.60 Lung infection; wound infection Up to 5 months 106 (Breslow et al. 2011)
Bordetella pertussis 0.40–0.80 Whooping cough 3–5 days 10,000 (Vidlak & Kielian 2016)
Klebsiella pneumoniae 0.50–2.00 Pneumonia Up to 30 months Unknown
Pseudomonas aeruginosa 1.50–3.00 Lung infection; wound infection Up to 5 weeks 1010 (Gama et al. 2012)
Staphylococcus aureus 1.00–1.50 Lung infection; wound infection; toxic shock syndrome Up to 7 months 100,000 (Vidlak & Kielian 2016)
Bacillus anthracis 3.00–10.00 Highly fatal lung infection; skin infection 56 days 8,000 (Gama et al. 2012)
Aspergillus fumigatus 10.00–20.00 (Loures et al. 2015) Allergic bronchopulmonary aspergillosis (ABPA); allergic Aspergillus sinusitis; Aspergilloma; chronic pulmonary aspergillosis; invasive aspergillosis 30 days (Neely & Orloff 2001) Unknown
Candida albicans 10.00–12.00 Lung infection; oral and vaginal infections Up to 3 months Unknown
Cryptococcus spp. 4.00–6.00 Lung infection; meningitis Unknown Unknown
Pneumocystis spp. 2.00–6.00 Pneumonia Unknown Unknown
Viruses
One of the most frequently encountered viral pathogens is the rhinovirus, which is the primary cause of common cold in humans, closely related to respiratory diseases. There are three species of rhinovirus (A, B and C) that include around 160 serotypes (Glanville and Johnston 2015; Pomeranz et al. 2019; Taylor-Robinson and Tyrrell 1962). The symptoms that they cause upon human infection include sore throat, runny nose, nasal congestion, sneezing and cough, muscle aches, fatigue, malaise, headache, muscle weakness and loss of appetite. However, this virus can also cause exacerbation of underlying lung disease, for instance, in critically ill patients with pneumonia, with or without co-pathogens. In terms of particle size, they are among the smallest viruses, with diameters of about 30 nm (Collier et al. 2000; To et al. 2017).
Another very common respiratory viral infectious agent is the influenza virus, which causes the common flu. There are four types of this virus (A, B, C and D) (Iwasaki and Pillai 2014; Kim et al. 2018; Lyons and Lauring 2018). Types A, B and C are known to infect humans (Kumar 2017; Peteranderl et al. 2016; Webster and Govorkova 2014), while D affects cattle. Normally, flu is characterised by systemic symptoms such as fever, myalgia, headaches and severe malaise, and respiratory symptoms such as coughing, sore throat and rhinitis. Those occur after approximately 2 days of an incubation period and can last for up to 7 to 10 days. Coughing and tiredness symptoms though can persist for even up to two weeks. If the virus reaches the alveoli of the lungs, it can result to serious viral pneumonia and interstitial pneumonitis. The influenza virus especially consist a major health risk and hazard for the elderly or immunocompromised individuals (Pleschka 2013).
Coronaviruses is another group of viruses causing diseases in humans, mammals and birds. When humans are infected by coronaviruses, this leads to respiratory infections that can range from mild effect to detrimental for the human health and even lead to death. Mild symptoms are similar to these of common cold, while more lethal strains can result in severe respiratory illnesses such as SARS, MERS and SARS-CoV-2 syndrome (de Wit et al. 2016; Hageman 2020; Yin and Wunderink 2018). The mortality rates range from 5 to 15% (Chan et al. 2003; Singh 2016; Weiss and Murdoch 2020). The SARS-CoV virus pandemic (2002–2004) resulted in 926 deaths worldwide, while the newly identified SARS-CoV-2 virus led to 279,000 deaths worldwide by 21/05/2021, only 6 months after the first outbreak (Lauxmann et al. 2020; Rothan and Byrareddy 2020). As of July 2017, 2040 MERS-CoV laboratory confirmed cases, resulting in 712 deaths, were reported globally, with a majority of these cases from the Arabian Peninsula (Chafekar and Fielding 2018). There are as yet no vaccines or antiviral drugs to prevent or treat human coronavirus infections. Finally, other airborne viral pathogens include respiratory syncytial virus (RSV) and parainfluenza virus (Collier et al. 2000).
Bacteria
Streptococcus pneumoniae is asymptomatically carried in healthy individuals, typically colonising various tissues of the upper respiratory system, as well as the sinuses. However, in susceptible individuals with weaker immune systems, such as the elderly and young children, S. pneumoniae can lead to serious pneumonia. Moreover, several strains of this species have developed resistance to many of the traditional antibiotics, which makes such infections difficult to treat (Feldman and Anderson 2016). This bacterium also causes bronchitis, rhinitis, acute sinusitis, otitis media, conjunctivitis, meningitis, sepsis, osteomyelitis, septic arthritis, endocarditis, peritonitis, pericarditis, cellulitis and brain abscess (Murray et al. 2013).
Haemophilus influenzae is a bacterium that is responsible for a wide range of topical and systemic infections. Most strains of H. influenzae are opportunistic pathogens, as they usually grow on the mucosal layers of the respiratory tract without causing any disease. However, when other factor such as a viral infection, impaired immune function or chronic inflammation create the appropriate conditions, then a disease can occur. In infants and children, H. influenzae type b (Hib) causes bacteraemia, pneumonia and acute meningitis. More rarely, it can also lead to cellulitis, osteomyelitis and infectious arthritis (Butler and Myers 2018).
Legionella pneumophila is a bacterial pathogen which invades and replicates inside macrophages via phagocytosis. Inside the macrophages, the bacteria are enclosed into a membrane-bound vacuole that protects them from degradation by cellular enzymes and allows them to multiply in large numbers. Legionella is most commonly transmitted by inhalation of contaminated aerosols produced by water sprays, jets or mists. This bacterium can cause Legionnaires’ disease and the less severe form, Pontiac fever. The common clinical symptoms of Legionella infection include high fever, cough, chills, difficulty in breathing, neurological problems, muscle weakness, diarrheal, chest pain, headache, nausea and vomiting. Legionnaires’ disease, which is a form of atypical pneumonia, has a mortality rate in the range of ∼10–50% (Murray et al. 2013; Prussin et al. 2017).
Mycobacterium tuberculosis is the causative agent of tuberculosis. Although this type of lung disease was widely controlled after the discovery of antibiotics, new emerging multidrug-resistant (MDR) strains are still a great concern in many areas of the world. Symptoms include chest pain and a prolonged productive cough. Approximately 25% of tuberculosis patients remain asymptomatic, but they can still spread the pathogen (Hunter 2016, 2018; Wang 1999). From time to time, patients may cough up blood in small amounts, while in rare cases, the infection may damage the pulmonary artery, resulting in massive bleeding (Bansal et al. 2018; Beggs et al. 2003). Other bacteria that can lead to serious lung disease are Acinetobacter baumanii, Bordetella pertussis and Klebsiella pneumoniae Pseudomonas aeruginosa, Staphylococcus aureus and Bacillus anthracis (Murray et al. 2013).
Fungi
Aspergillus fumigatus is a fungal pathogen that it is ambiguously found both indoors and outdoors. It forms thousands of tiny spores (2–3 μm) which readily become airborne and after inhalation they can easily penetrate the tissues of the lower respiratory system. The fungus is capable of growth at temperatures up to 50 °C, with spores surviving at 70 °C (Dijksterhuis 2019; Grishkan 2018; Pitt and Christian 1970). Typically, inhaled spores are quickly eliminated by the immune system in healthy individuals. However, in immunocompromised people, such as transplant recipients, AIDS or cancer patients, the fungus is more likely to become pathogenic and lead to more serious lung illnesses such as allergic bronchopulmonary aspergillosis (ABPA), aspergilloma, chronic pulmonary aspergillosis and invasive aspergillosis. Due to the extended use of immune suppressants for treating human diseases, it is estimated that A. fumigatus is the cause of over 600,000 deaths annually, with mortality rates ranging from 25 to 90% (Latgé and Chamilos 2019; Murray et al. 2013). Other important fungi that can cause respiratory disease in immunocompromised patients are Candida albicans, Cryptococcus spp. and Pneumocystis spp. (Murray et al. 2013).
Microbiological and molecular methods for microbial enumeration and identification
Nowadays science innovation arises from the multidisciplinary approach of a variety of scientific fields. Analytical methods often applied in biomedical sciences find applications in engineering. Below, the main microbial and molecular methods for identification or biological agents are reported, and might be proved very useful in engineering applications such as determination of the biological load of indoor air.
Air samplers
The microbiological quality of the air is usually determined by sampling small volumes of air, which contain various bioaerosols. Then the process of enumerating and identifying the microbes within the sample is taking place. Such microbial monitoring is done routinely at healthcare-related areas for assessing environmental quality and deciding if corrective intervention is necessary or not (Napoli et al. 2012; Razzini et al. 2020). Air samplers are the most frequently used devices for such purposes, mainly because of their low costs and easiness of handling. Air samplers draw in air and force the various particles in it to get impacted over collecting surfaces or impinged into a liquid. These samples can also utilise filters for selecting a specific range of particles, while different impaction rates can be used by adjusting the vacuum settings (Ghosh et al. 2015).
Air filtration
Another method for collecting airborne bioaerosols is filtration. During this procedure, air is drawn through a filter with a 0.2-µm pore size, trapping all particles apart from small viruses. This can be facilitated by a vacuum system. The filter can be then used for enumerating the microbes or culturing them before identification with traditional or molecular techniques. One important advantage of filtration is that the captured microorganisms remain viable. Also, the filter can be directly used for nucleic acid extraction (Ferguson et al. 2019). However, such filters are prone to overloading or damage and also desiccation can result in low recovery efficiency of the trapped microbes (Ghosh et al. 2015). Such method of air sampling by filtration was used for sampling of bioaerosols by (Predicala et al. 2002) at a swine farm environment.
Other bio-aerosol precipitation approaches
More laboratory-based approaches are available for precipitating bioaerosols or other particles from the indoor air. Those however are not used as frequently as the ones mentioned above. These include sedimentation and centrifugation, as well as electrostatic or thermal precipitation (Ghosh et al. 2015).
Cotton swabs
Medical-type swabs are often used for taking biological samples from surfaces, for subsequent microbiological or molecular analysis. The procedure is very simple, as the swab is rubbed onto or into the contaminated area and then wiped across a culture medium, such as an agar plate, where the bacteria and fungi from the swab may grow. This has to be done quickly and aseptically, in order to avoid contamination of the sample with other environmental microbes. It has been suggested that if the swab is mildly sonicated after sampling, the microbial recovery rate on the culture media is increased (Ahnrud et al. 2018).
A combination of air sampling and cotton swabs was used this year at a Milanese hospital for detecting SARS-CoV-2 genetic material (RNA) in the air and on key surfaces of the building (Razzini et al. 2020). The most contaminated surfaces were hand sanitizer dispensers (100%), medical equipment (50%), medical equipment touch screens (50%), shelves for medical equipment (40%), bedrails (33.3%) and door handles (25%) (Haun et al. 2016; Kurgat et al. 2019; D. J. Weber et al. 2019). Other recent studies that used cotton swab sampling approaches for microbiological monitoring are these by Lee et al. (2018) and Luksamijarulkul and Pipitsangjan (2015). According to these studies, it was shown that such swabs remain the easiest and most widely used method for surface sampling. A variety of more effective swabbing products, such as nylon, rayon and polyester swabs, has been lately developed (Bruijns et al. 2018).
Microscopy
One of the most traditional methods for microbial identification is observation of the microbe’s physical characteristics, such as shape, size and the types of dyes that absorbs, under a light microscope. For example, the Gram stain can distinguish different bacterial species to Gram-positive or Gram-negative, according to their cell wall structure. Other types of staining can provide information about production of spores (Schaeffer-Fulton staining), capsules (India ink or nigrosine) and mycolic acids (acid-fast staining). Light microscopy can also be used for enumeration of cells, by using of a haemocytometer. For virus identification, the use of electron microscopy is required (Ahmed et al. 2016).
Use of selective and differential culture media for microbial enumeration and identification
The method that is most frequently used for isolating bacteria and fungi involves culturing them on the surface of solid nutrient media (Brugger et al. 2012; Burmølle et al. 2009; Wiegand et al. 2008). Such media contain all the necessary nutrients for the growth of a wide range of microbes, including carbon (C), nitrogen (N) and phosphate (P) sources, amino acids, inorganic salts and trace elements. It also contains 1.5% agar, which is a polysaccharide that gives a gel-like structure to the solid medium. After incubation for 24 h at 37 °C, colonies appear on the surface of the agar, which can be counted and identified based on their morphological characteristics (Collins et al. 1989).
There is a wide range of selective and differential media that are used in clinical microbiology laboratories for microbial enumeration and identification (Bonnet et al. 2020; Reddy et al. 1972; Yoo et al. 2014). Selective media contain compounds that inhibit the growth of some microorganisms, permit however others to grow. Differential media contain ingredients making the colonies of a certain group of microorganisms appear in a different colour than this of other groups. Some differential media are also selective, for instance, MacConkey agar, which is selective for Gram-negative coliforms and can also differentiate between lactose-fermenting and non-lactose-fermenting bacteria (Ahmed et al. 2016; Nigro and Steward 2015). Such traditional microbiological culture-based approaches have been followed recently for indoor bioaerosol characterisation (Yasmeen et al. 2020; and Nasir et al. 2018).
Use of biochemical tests for microbial identification
A plethora of biochemical tests is also available for identifying microbial species. These tests usually determine the ability of a species to grow in media containing certain carbon or nitrogen sources, such as glucose, lactose and urea. As the microbes metabolise these substrates, they produce products leading to a medium colour change, which is regarded as a positive test result. Based on the results from many such tests, a microbiologist can use specific charts for identifying a bacterial pathogen. Automated identifying systems are today available for running these tests in a high-throughput mode, e.g. VITEK2® and FAME, and those have been very useful for bioaerosol profiling (Duquenne 2017). Similar approaches can be used for identifying fungi, but not viruses as they do not have a metabolism activity (Spiegelman et al. 2005).
Polymerase chain reaction
The polymerase chain reaction (PCR) is a widely used method for rapidly amplifying a specific area of the DNA of a sample (Gadsby et al. 2019; Liu et al. 2019a, b, c, d, e; Siqueira and Rôças 2003). The PCR product is then analysed by gel electrophoresis and a final result can be obtained about the identity of microbe, based on whether it contained the targeted area in its DNA or not (Järvinen et al. 2009). Real-time quantitative (qPCR) is a more advanced method and can be used for both identification and quantification of a microbial pathogen in a clinical sample. Real-time qPCR utilises fluorescent chemicals that can be detected by a detection system when amplification of the desired DNA area begins. As a result, there is not a need for gel electrophoresis. This method is more sensitive and precise than the standard PCR method (Kralik and Ricchi 2017). Real-time quantitative and standard PCR methods have been recently applied in several indoor bioaerosol surveillance studies (Coleman and Sigler 2020; Razzini et al. 2020).
Matrix-assisted laser desorption/ionisation
Matrix-assisted laser desorption/ionisation (MALDI) is an ionisation technique for mass spectrometry that uses a laser energy absorbing matrix to create ions from large molecules (Dingle and Butler-Wu 2013; Jang and Kim 2018; Singhal et al. 2015). Biological macromolecules such as DNA, proteins and peptides tend to be fragile and fragment when ionised by more conventional ionisation methods. The advantage of matrix-assisted laser desorption/ionisation (MALDI-TOF) is that it does not lead to such fragmentation, something which makes it suitable for clinical use. Colony material of the microbe in question is placed onto the sample target and overlaid with matrix. The resulting spectra are used for the identification of micro-organisms, after analysis by dedicated software and compared with stored profiles. MALDI-TOF is much faster, more accurate and cheaper than traditional methods (Madsen et al. 2015; Murray 2012; White et al. 2019).
Nucleic acid sequencing
Next-generation sequencing (NGS) is a highly advanced technology that via which millions of DNA fragments can be simultaneously and independently sequenced (Huang et al. 2020; Lin et al. 2019; Sung et al. 2018). In clinical microbiology laboratories, metagenomic NGS (mNGS) is most frequently used for detection of certain pathogens. The cost of such analyses is still very high, and most hospitals cannot afford them even when the results are obtained faster and are much more reliable.
Another advantage of NGS is that analyse DNA or RNA in a clinical sample are surveyed masse, in contrast to PCR that can only analyse few specific targets per run (Gu et al. 2019; Madsen et al. 2015; White et al. 2019). MALDI-TOF and NGS are definitely the most promising advanced technologies for microbial identification at the moment. This year’s “Viruses in the Built Environment (VIBE) meeting in Arlington, Virginia, USA, highlighted the importance of constructing bioinformatic tools and databases that will ensure a quick and accurate microbiological monitoring within buildings (Prussin et al. 2020). Other methods that can also help with microbial identification include DGGE, serological approaches, epifluorescent microscopy and flow cytometry (Ghosh et al. 2015).
New novel approaches for real-time monitoring of bioaerosols
The last few years, several novel approaches have been tested and applied for real-time monitoring and characterisation of bioaerosols. These include fluorescence spectroscopy, elastic scattering, microscopy and holography, Raman spectroscopy, mass spectrometry, breakdown spectroscopy, remote sensing, microfluidic techniques and paired aqueous techniques (Huffman et al. 2020; Nasir et al. 2019). Examples of such modern applications are provided below. In 2013, Usachev et al. (2013) applied a surface plasmon resonance-based immunosensor for real-time bioaerosol detection. The collected viral particles were mixed with a target-specific antibody and the positive aggregates were efficiently detected in less than 2 min.
Choi et al. (2015) developed and tested a micro-optofluidic platform that proved able to accurately detect, quantify and characterise bacterial aerosols, by use of fluorescent dye detection, fluidics and optical microscopy. Furthermore, an adenosine triphosphate (ATP) bioluminescence assay was developed by detecting and measuring the concentration of bacterial aerosols. This assay was coupled with a continuous aerosol sampling device. The collected bacteria were charged, added to a liquid buffer and their numbers were estimated by measurement of the ATP levels generated via microbial metabolism (Park et al. 2016).
Finally, laser-based bio-detectors were applied for characterising a great number of individual particles in seconds, by analysing optical scattering and fluorescence characteristics. Data analysis by use of Artificial Neural Networks led to construction of decision trees for aerosol classification (Leskiewicz et al. 2018). All these approaches seem extremely promising and are expected to be more widely applied for characterisation of medically important aerosols in the near future.
Survival of respiratory microbial pathogens
The duration of survival of different microbial pathogens in the environment is a major public health parameter that has significantly attracted the interest of most epidemiologists worldwide. The main factor that affects this is the structural composition of the pathogen. For instance, fungal and bacterial spores can survive for years due to their thick cell walls and dormant metabolism. Non-enveloped viruses are also very tolerant due to their resistant protein capsule. Enveloped viruses are less resistant, because their lipid bilayer is susceptible to heat, dryness and chemical agents. Finally, fungi are usually better at survival than bacteria due to their stronger cell walls (Table 2). Both bacteria and fungi often require high water activity and nutrient availability in order to survive and grow (Dedesko and Siegel 2015; Mendell et al. 2018). Furthermore, the type of surface is also important for determining the survival of microbial pathogens. For example, moist, porous and soft surfaces such as carpets and curtains are more likely to accommodate microbial growth than dry non-porous hard surfaces such as wood, plastic or metal (Thompson and Bennett 2017).
Some types of surface material such as copper, silver or antibacterial polymers can lead to microbial death and prevention of colonisation (Muller et al. 2016). Finally, environmental factors such as heat, pH, humidity, UV radiation and chemicals can affect microbial viability. Some bacteria are tolerant to adverse environmental condition (Walsh and Camilli 2011), while many bacteria and fungi can form biofilms, slimy layers made of polysaccharides and proteins that protect them from hazardous conditions (Hall-Stoodley et al. 2004). Environmental factors such as humidity and ambient temperature can also affect the survival of microbes in the air, either within or outside bioaerosol droplets, with a subsequent importance for respiratory disease (Prussin et al. 2020; Pyankov et al. 2018; Tang et al. 2006).
Transmission of respiratory microbial pathogens
Microbial pathogens can be transmitted via a variety of routes, including person-to-person (touch, saliva), airborne, foodborne/waterborne, via blood, sex, insects or fomites (non-living objects, such as door handles or towels, etc.). When it comes to airborne transmission, this can be classified as long and short range, depending on the viability of a pathogen in the air or the stability and size of the droplet that might carry it. Large-droplet diameter is considered to be > 50 to 60 μm, small droplet diameter is < 50 to 60 μm and droplet nuclei diameter < 5 to 10 μm (Tang et al. 2006) (see the ‘Size of particles and biological agents’ section). An example of short-range airborne transmission is the inhalation of droplets from a coughing or sneezing infected donor (from a < 1-m distance), while long-range airborne transmission can include inhalation of fungal or bacterial spores that have travelled a long distance in the air via the wind (see ‘The challenging nature of biological agents’ transmission in indoor environment’ section). However, several non-spore bioaerosols can also travel long distances, if certain environmental conditions permit it (e.g. indoor air circulation) or if they are inside small droplets or droplet nuclei.
Many respiratory pathogens can be also transmitted via personal contact, via dust or from fomites, if the recipient touches a contaminated area and then touch facial, oral or nasal areas, allowing the entry of the pathogen into the respiratory tract (Wei and Li 2016a, b). Even if the pathogen enters the upper respiratory system, it might not be able to cause disease unless it penetrates the lower respiratory tract (trachea, bronchi, bronchioles and the alveoli). As it was mentioned in the ‘The human respiratory system’ section, this depends on the size of the infectious agent or the droplet that carries it (< 5 μm are able to penetrate lungs).
Factors that affect the development of respiratory infectious disease
Pathogen-related factors
Several microbe-related factors can affect its ability to cause respiratory disease. Some infectious agents are more pathogenic than others and even within the same species there are often sub-species, serovars or strains that are more virulent than others. This depends on the weaponry of virulent factors that a strain carries, such as toxins, super-antigens and degradative enzymes that destroy the tissues and cause localised damage and inflammation. Moreover, some strains have the ability to form filaments, spores and biofilms that make them more invasive and tolerant to the attacks of the immune system. Finally, the ability of a strain to mutate is an additional factor that affects it virulence (Davidson 2018; Murray et al. 2013).
In addition, the number of the initial infectious agents that enter the site of infection (e.g. lungs) is very important. Usually, low numbers, e.g. 50–150 cells or virus particles, can be easily dealt by the immune system which represses the infection before it leads to disease. Higher infectious doses can be difficult to control. However, this also depends on the type of pathogen that reaches the site of infection. The infectious doses of certain infectious agents that can lead to death have been experimentally measured by use of mice or other laboratory animals (Prussin et al. 2020; Tang et al. 2006) (Table 1).
Host-related factors
There are also many different host-related factors that can determine if a respiratory disease such as pneumonia will develop or not and how severe it will be. Firstly, the age of the patient is important. Young children do not have a fully developed immune system and the elderly have a weakened one that is often unable to eradicate the infectious agent. Vaccination against agents such as the influenza virus, Mycobacterium tuberculosis or Streptococcus pneumoniae can also prevent development of respiratory disease.
Immunocompromised individuals, such as cancer patients, transplant recipients or HIV patients, are also more vulnerable to infectious agents that cannot cause respiratory disease in healthy individuals (e.g. Cryptococcus neofmans, Candida albicans). Moreover, smoking and air pollution destroy the ciliated cells of the respiratory system that are a physical defence mechanism against microbes and push mucous-trapped microorganisms out of the body. This makes smokers more susceptible to lung and airway disease. Finally, underlying disease such as diabetes, obesity or cystic fibrosis can affect the potency of the immune system (Engin et al. 2020; Lacoma et al. 2019; Murray et al. 2013).
The role of heating, ventilation and air-conditioning systems
Heating, ventilation and air-conditioning (HVAC) systems are widely recognised as the most influential engineering approach to control the airborne transmission of the pollutant agents in the internal spaces (Bhagat et al. 2020; Li et al. 2007; Luongo et al. 2016; Qian and Zheng 2018; Shajahan et al. 2019; Wei and Li 2016a, b) as their operation is associated with the movement/ flow of the indoor air due to the introduced buoyancy forces and pressure differences. The operation patterns of these systems have been analysed in several engineering and epidemiological studies in last decades, resulting in the suggestion of three individual characteristics, (a) ventilation rate, (b) airflow direction and (c) thermal plume, to be the main parameters that significantly determine the transportation and the infectious mechanisms.
Adequate ventilation rate is pointed out as an important factor for removing the pollutants in general and especially the less studied biological agents from the indoor spaces. Airflow direction leads the air from the clean zone into the pollutant source area and consequently from the polluted space to outdoors. Thermal plume influences the space stratification conditions and the kinetics of the pollutant agent. The following sections summarise and criticise the results of previous studies related to the aforementioned parameters regarding the control of the airborne transmission of the contaminant agents, and on minimising the risk of cross-infection between the occupants.
The role of ventilation
Ventilation is the supply of the outdoor air into internal building spaces, and can be categorised as natural and mechanical or forced ventilation. Both ventilation options induce different advantages and disadvantages while their combination could provide mixed characteristics (Cao et al. 2014; Gilkeson et al. 2013). Natural ventilation is of low cost and maintenance and allows the ambient air to be entered into the building by various and mixing routes. In contrast, the use of natural ventilation is directly linked with fluctuating ventilation rates that under specific outdoor and indoor conditions the air movement could be inadequate or overabundant. In addition, the intake air is unfiltered and depending on the ambient environmental conditions it may transport a variety of undesirable contaminants (e.g. dust, fumes and microbes, among others).
Mechanical ventilation could supply filtered fresh air and especially in combination with high efficiency minimum efficiency reporting value (MERV) 13–16 filters, the risk of the airborne disease transmission can be significantly reduced (Rui et al. 2008). Although the mechanical ventilation systems offer better control capability of the indoor environment characteristics, however, they introduce significant financial expenditures (Azimi and Stephens 2013; Escombe et al. 2019; Hobday and Dancer 2013). Weather using natural or mechanical ventilation the quantity of pathogens and the quality of the indoor air are not necessarily higher in case of the first or second alternative, e.g. (Qian et al. 2010; Short and Al-Maiyah 2009; Stockwell et al. 2019). This is due to the fact that in both cases the airflow rate and the airflow movement pattern are the most prominent characteristics that determine the efficacy of each option to provide the desirable indoor atmosphere. In general, the use of ventilation in buildings is associated with a dual positive and negative effect against the airborne transmission (Noakes and Andrew Sleigh 2009). The positive role is the dilution of the concentration or the dispersion of the biological agents and particulates leading to minimising the occupants’ risk. In parallel, the transportation of the bio-aerosols and particulates among adjacent spaces is a non-negligible undesirable effect.
Mechanical systems of ventilation
In mechanical ventilation systems, two different airflow patterns are commonly used such as the displacement ventilation (DV) and the entrainment or mixed ventilation (MV) flow (ASHRAE 2017b). There are also advanced mechanical systems such as personalised ventilation (PV) and personalised exhaust which can be installed stand alone or in combination with other ones in spaces with or without specific requirements (Melikov 2004). The application of PV systems in common indoor spaces becomes more attractive, as many recent studies indicate the benefits on the indoor air quality (IAQ) improvement and minimising the airborne transmission risk (Al Assaad et al. 2018; Habchi et al. 2016; Lipczynska et al. 2015; Melikov et al. 2013; Yang et al. 2015).
Except for the fact that any stand-alone or conjugated ventilation system under controlled conditions is able to supply fresh air, however, the differentiation of the airflow direction and pattern-based on the design characteristics of each system in association with ventilation rate- are the most important parameters that influence the (a) contaminant concentration; (b) contaminant removal effectiveness; (c) infection risk; and (d) human’s exposure to pollutants in general and biological agents.
Displacement and entrainment/mixed ventilation systems
Displacement ventilation (DV) system or displacement airflow describes the air movement in one direction by a piston type motion. Ideally, the air is not mixed and the pollutants are totally removed out from the indoor space. The airflow in DV systems could be either downward (ceiling-to-floor) (see Fig. 6) or upward (floor-to-ceiling) (see Fig. 7), based on the design requirements of each space. In both cases, the idea is to supply fresh and clean air with low velocity leading to a laminar airflow which intent to sweep air across the space with the minimum possible mixing (ASHRAE 2017b).Fig. 6 Displacement downward ventilation pattern
Fig. 7 Displacement upward ventilation pattern
Due to that characteristics, the downward DV system is considering as the ideal system for removing the contaminated indoor air, and is expected to minimise the cross-infection risk (Qian and Zheng 2018; Tang et al. 2006). However, either the design of DV systems with airflow pattern about 4.0 ACH or the synergies with the humans’ thermal plume are impossible to produce laminar flow, thus mixed ventilation airflows occur (Qian et al. 2008).
Entrainment/ mixed ventilation (MV) system or airflow, Fig. 8, refers to a circular pattern of air flow in which the intake fresh air is conventional mixed with the internal air and finally the mixture leaves the space.Fig. 8 Mixed ventilation (MV) pattern
In this case, the pollutants are removed by dilution. Entrainment flows, according to mixing conditions are characterised as short-circuit flow or complete/ uniform mixing (well-mixed). In the first case, the supply air leaves the space without mixing with the indoor air as a result of very poor mixing conditions, while in the second one the supply air is instantly and evenly distributed in the space leading to a perfect mixing with the room air (ASHRAE 2017b).
Underfloor air distribution (UFAD or UAD), in Fig. 9, is a hybrid ventilation method that combines the characteristics of both displacement and mixing ventilation schemes. Outdoor air is introduced into a plenum floor and supplied to the indoor space throughout floor-mounted diffusers. The diffusers produce a turbulent flow near to the floor level and the supplied air is mixing with the indoor one. Then the mixed air moves to the ceiling in a laminar flow without mixing phenomena and exhausted from the space through outtake diffusers.Fig. 9 Underfloor air distribution pattern
The ventilating performance of the underfloor distribution system is thus between upward DV and MV systems (ASHRAE 2017b). The effectiveness of the DV, MV and UAD systems on minimising the airborne transmission of the infectious agents has been evaluated in several studies using experimental and numerical approaches. A detailed analysis on these studies indicated that the majority of them deal with the assessment of cross-infection risk, while some of them focused on the assessment of the droplet dispersion mechanisms and behaviour.
Qian et al. (2006) performed a series of experiments to understand the interaction of the exhaled bio-aerosols in downward and upward DV and MV airflows in a hospital ward. They reported that downward DV with an airflow rate of 4 ACH has similar behaviour as the MV, due to the turbulent characteristics of the flow. In addition, they do not suggest the installation of upward DV system in hospital wards due to the possibility of increase the exposure level, if an occupant is located in the exhalation jet.
Olmedo et al. (2012) studied the human exposure to the contaminants of the exhaled bio-aerosol among two persons taking into consideration between other parameters the use of upward displacement and mixing ventilation. They found that in the case of upward DV, the exhaled air flows transport for longer distance with higher concentration. Lin et al. (2012) accessed the risk of pathogen inhalation under stratum and upward DV and concluded that the risk was higher when upward DV system was used.
Chen et al. (2014) analysed the person-to-person bio-aerosol transport under upward displacement and mixing ventilation and UAD systems. They indicated that the upward DV and underfloor air distribution have the same performance in reducing the contaminant exposure and were about 20% better than the MV. Although this study presents contradictory behaviour compared to the similar ones, the authors, however, reported that in cases of upward DV and UAD, significant variations in the relative effect on exposure have been noticed. This phenomenon indicates that under certain circumstances, the pointed out relationship among the alternative ventilation systems may be altered (Chen et al. 2014). Similar results and recommendations have also been reported in many studies (Ai et al. 2019a, b; Friberg et al. 1996; Jurelionis et al. 2015; Keshavarz et al. 2017; Li et al. 2012; Lin et al. 2013; Mazumdar et al. 2010; Salmanzadeh et al. 2012; Villafruela et al. 2019; Wu and Lin 2015; Yang 2013; Yin et al. 2009).
Lai and Cheng (2007) studied the droplet’s dispersion in a space under upward displacement and well-mixed ventilation flows. They concluded that for the well-mixed ventilation system, the dispersion pattern is driven by the velocity airflow. High-velocity airflow produces within 1 min a homogeneous bio-aerosol. In contrast, when upward DV with low-velocity airflow is used, the dispersion partner is dominated by the droplets’ size. In this case, 10-μm droplets begin to settle at the lower areas of the located space.
Gao et al. (2008a, b) simulate the dispersion characteristics of an exhaled bio-aerosol consisted with droplets in the range of 1 to 10 μm in an office space using upward DV, MV and UAD systems. The obtained results showed that in MV system the exhaled droplets were uniformly distributed. However, in all ventilation systems, the exhaled flow was trapped in the breathing zone of the occupant.
Mui et al. (2009) stated that the droplet dispersion and mixing in case of DV is poorer, compared to the MV. Sun and Ji (2007) proved again that the efficiency of the upward DV is higher in removing small droplets, while MX has equal efficiency for removing droplets in the range of 80 to 100 μm and higher efficiency in removing large size droplets. They concluded that this behaviour is subjected to the equilibrium between gravitational and buoyancy forces. High gravitational forces occur in the case of large droplets, while the buoyancy forces become significant in the case of small-sized droplets and high-velocity airflow. DV introduce low-velocity airflow, and for the case of upward airflow pattern, the large size particles tend to settle in the lower part of the space. However, for the case study of downward airflow systems, these particles can be efficiently removed by the outlet vents. MV systems are characterised by high airflow patterns leading to a well-mixed bio-aerosol which can be efficiently removed from the space, regardless of the droplet sizes. It is worth noticing that similar results have also been reported in the following studies (Berlanga et al. 2018; Chao and Wan 2006; Gao et al. 2012a, b; Lai and Wong 2011; Seepana and Lai 2012; Li et al. 2011; Yin et al. 2011).
Personalised ventilation systems
Personalised ventilation (PV) system or personalised airflow intents to provide fresh air into breathing zone of an occupant. The system uses air terminal devises that consist of nozzle/s allowing the control of airflow rate by the occupant to the desirable level and direction. The PV system has two main advantages: it improves the quality of the inhaled air and allows the user to control the temperature, velocity and direction of the supplied airflow (Melikov 2004). The contribution of PV systems on the mitigation of the airborne cross-infection risk has been analysed in several studies. Cermak and Melikov (2007) conducted a series of measurements to examine the capability of two PV systems in association with an UAD system to protect occupants from exhaled infectious aerosols and emissions from the floor materials. They found that the conjugated systems protect the occupants from inhaling the aerosols, while the concertation of the pollutants into the indoor air was increased. Pantelic et al. (2009) studied the protective role of a PV system against the infectious cough droplets released near the PV occupant. They addressed that the PV system significantly reduced the bio-aerosol concentration in the breathing area of the occupant. It had also reduced the risk of cross-infection particularly in cases that the source point of the bio-aerosol infection and the occupant were at a distance less than 1.75 m. He et al. (2011) assessed the airborne transmission of an exhaled bio-aerosol between two occupants under three ventilation systems, namely MV, upward DV and UAD working autonomous and in conjugation with PV. They concluded that for PV scenarios the quality of the inhaled air has been improved. A study of Mazej and Butala (2012) proved that by using a PV system, the amount of the re-inhaled bio-aerosol is extremely low; however, the dispersion of bio-aerosol to the indoor air increases the risk of cross-infection onto the occupants who are not using personalised ventilation. Li et al. (2013) analysed the exposure of occupants to the exhaled pollutants under two different conjugated ventilation systems. They concluded that the upward DV combined with PV provides better inhaled air quality compared to the alternative option of MV with personalised one. Pantelic et al. (2015) evaluated the effectiveness of a PV system to reduce the inhalation intake fraction of an infectious bio-aerosol against to MV system. The obtained results indicated that the PV system substantially reduces the intake fraction for the all analysed cases. In addition to the above studies, it is worth mentioning that similar results have also been addressed in many other cases (Bolashikov et al. 2015; Bolashikov and Melikov 2009; Cermak et al. 2006; Nielsen 2009; Nielsen et al. 2007a, b; Nielsen et al. 2007b; Pantelic and Tham 2011; Pantelic and Wai 2009; Tham and Pantelic 2011; Wai and Pantelic 2009; Yang et al. 2015; Zheng et al. 2011). Moreover, detailed reviews on the personalised ventilation systems have been published on (Liu et al. 2019d; Melikov 2004; Zhai and Metzger 2019).
Natural ventilation
Natural ventilation is the physical flow of the external air through the building vents into indoor spaces caused by a thermal and/or wind pressure difference. Under certain circumstances, it can be provided an adequate level of pollutants’ removal, which is not always controlled and acceptable. There are two types of natural ventilation airflow patterns: the cross and the single-sited ventilation. Cross ventilation is achieved using openings in both sides of the space and it is driven by the pressure difference. Single-sited ventilation occurs when one or more openings in the same façade of the building are open. Thus, the airflow could be driven by temperature and/or pressure difference. Although the role of natural ventilation on indoor air quality and comfort levels has been well studied and documented (e.g. (Allocca et al. 2003; Brager and De Dear 1998; De Dear and Brager 2002), however, the effect on the airborne transmission of pollutants and bio-aerosols between the adjusted building units and their dispersion in lower or higher building floors has attracted the research interest mainly after the SARS pandemic in 2003. Li et al. (2005) studied the SARS virus transmission between adjusted flats in a high-rise residential building in Hong Kong. They concluded that in the natural ventilated high-rise apartment buildings it is difficult to control the air leakage between flats as the flow is driven by the air-tightness and the pressure difference. This phenomenon leads to carry bio-aerosols between the apartments of the building. A study by Gao et al. (2008a, b) proved again the airborne transmission across apartments in a high-rise natural ventilated building through open windows between flats caused by buoyancy effect. They reported that the gaseous pollutant’s concentration in the immediate upper flat is 2 orders lower compared to the lower flat in which the gaseous pollutant is generated, while the risk of infection is 1 order lower, respectively. They also noticed the importance of wind speed and concluded that high-speed winds act like air-curtain reducing the pollutants’ spread. However, they clearly reported that in natural ventilated multi-family buildings the inflection control of bio-aerosols should consider the role of this airflow.
In-line with the previous study, the same research team simulated the airborne transmission of particle pollutants (Gao et al. 2009a, b). They found that the concentration of the particle pollutant in the upper floor is between two to three orders lower than in the lower source floor. They also concluded that particles up to 1-μm disperse like gaseous pollutants, while particles larger than 20 μm show a strong deposition on the source space and limit their transport to the up-floor area. Ai and Mak (2014) studied the dispersion characteristics of infectious aerosols exhausted from a building unit in association with the hypothesis that the exhausted aerosol can re-enter into another unit of the building through opened windows. They reported that the re-entry ratios can be reach up to 10% based on the wind direction and façade characteristics, non-flush walls or balconies. The high re-enter ratio is observed in the windward site following by the leeward site both in case of 45° wind direction. In addition, the balconies enhance the re-entering phenomenon of the exhausted bioaerosols, except the case of the normal incident wind direction. Wu et al. (2018) studied the role of infiltration on the airborne transmission route and evaluate the associated infection risk in a high-rise building, under different wind directions and leakage characteristics of doors and windows. They found that infiltration rates below 0.7 ACH increase the cross-infection risk up to 20% compared to the risk of 9% in case of air change rates over 3 ACH. The increase of infiltration rate along the building height leads to the increase of the cross-infection risk in the lower building floors. They also reported that the wind direction is a significant parameter that influence the cross-infection risk. The higher cross-infection risk observed in case of the contaminant source is placed on the windward site and on the adjacent units. Finally, they concluded that improving the air-tightness of the internal openings and increasing the airflow on the external ones is an effective solution for the control of inter-unit airborne transmission. The effect of natural ventilation in the airborne transmission of bio-aerosols in multi-family buildings (in both vertical and horizontal directions) together with the role of wind characteristics has also been investigated by many scientist (Ai and Mak 2016; Ai et al. 2013; Cui et al. 2018; Liu and Niu 2011; Liu et al. 2007; Liu et al. 2008; Liu et al. 2011a, b; Liu et al. 2010, 2011a, b; Mu et al. 2016; Mu et al. 2017; Niu et al. 2005; Niu and Tung 2008; Wang et al. 2010a, b; Wu et al. 2019a, b; Zhou et al. 2014), who finally conclude to similar results and suggestions.
The role of ventilation rate
A minimum level of ventilation rate is recommended by relevant Standards (ASHRAE 2017a, 2019a, b; CEN 2019), in order to maintain the quality of the indoor air to a pre-defined acceptable level and minimise the risk of human exposure to pollutants in general and biological threats. In general, there are three methods for the calculation of the ventilation rate, that based on the: (a) perceived air quality, (b) criteria for individual substances and (c) pre-defined ventilation air flow rates. According to the perceived air quality method, the ventilation rate is found by adding the required air volume for people and emissions. This method is mainly used in residential and non-residential buildings in which critical contaminant sources are not identified. In spaces with essential pollutant sources the ventilation rate is calculated based on the generation rate of the pollutant, the concentration of the pollutant on the supply air, the guideline concertation of the pollutant in the indoor air and on the effectiveness of the ventilation system. The third method introduces pre-defined ventilation air flow rates based on the local climate and building characteristics, and is also used in residential and non-residential buildings. It is worth noticing that the first and third method in case of the non-occupied hours of the building, lower the ventilation rate to a minimum air flow needed to maintain the concentration of the non-human related pollutants to the guided level (CEN 2019). In line with the above strategy, Gao et al. (2012a, b), estimate that increasing the ventilation rate up to 10 ACH in schools led to a reduction of the peak inflection to influenza up to 9% and postponed the peak of outbreak by 3 days. However, they noticed that ventilation rates over 5 ACH may be difficult to reach and suggest to be used in conjunction with alternative prevention policies. A similar study (Gao et al. 2016), regarding the potential outbreak of influenza in Hong Kong, concluded that even in cases that the airborne transmission is 20% of the total inflection the increase of ventilation rate has strong influence on transmission pathways similar to other control measures. Nardell et al. (1991) studied the air borne infection caused by the operation of building’s ventilation and concluded that of increasing the ventilation rate by 67% and 133%, reducing the infection risk by 33% and 52% respectively. The relationship between ventilation rate and infection risk has also been studied in the work of Fennelly and Nardell (1998). They found that the infection risk decreases exponentially with the increase of ventilation rate, for instance, in a moderate-exposure space operated with 6 ACH the probability of infection is 0.42 and decreasing to 0.21 by increasing the ventilation rate to 12 ACH. Similar conclusions regarding the influence of ventilation rate to the inflection risk and on the associated concentration of airborne pathogen bioaerosols into indoor air have also been reported (Beggs et al. 2003; Bergeron et al. 2011; Cao et al. 2015; Chen et al. 2014; Escombe et al. 2007; Escombe et al. 2019; Gao et al. 2009a, b; Knibbs et al. 2014; Knibbs et al. 2011; Lim et al. 2010; Lindsley et al. 2012; Menzies et al. 2000; Milton et al. 2000; Myatt et al. 2004; Nielsen et al. 2010; Qian and Li 2010; Qian et al. 2010; Stockwell et al. 2019; Sun et al. 2011; Tung and Hu 2008).
Although these conclusions led into significant revisions and changes of the recommended ventilation rates on relevant standards and guidelines, over the last years, new findings indicate that the increase of ventilation rate might lead to the increase of the cross-infection risk. This is due to the fact that higher ventilation rates under specific conditions increasing the buoyancy forces of the airborne infectious droplets resulting in the increase of aerosol transmission and associated cross-inflection risk. Bolashikov et al. (2012) examined the exposure of a health professional and a patient to the airborne pathogen caused by an infected patient in a hospital isolation room under different ventilation rates. They performed a series of experiments and concluded that at the distance of 1.1 m for the inflected patient the peak concentration of the pathogen is much higher at the ventilation rate of 12 ACH compared to the ventilation rates of 6 and 3 ACH. Pantelic and Tham (2013) evaluated the capability of the ventilation rate to act as a sole indicator of the effectiveness of an air distribution system on the mitigation of airborne infectious disease transmission. They concluded that the increase of ventilation rate can lead to the increase of exposure risk under certain circumstances (e.g. upward airflow, characteristics of local airflow patterns and airflow quality). This evidence indicates that the use of ventilation rate as a sole indicator for the evaluation of the air distribution system’s effectiveness on the control of the infectious airborne transmission is not possible. Mousavi and Grosskopf (2015) noticed again that increasing the ventilation rate is not proportionately effective for reducing the aerosol concentrations in patient rooms. Ai et al. (2019a, b) studied the airborne transmission between an infected and a healthy person under exposed to a horizontal air flow. They also confirmed that the influence of ventilation rate is not straightforward to the expose index. The obtained experimental results indicated a decrease of the exposure index when the ventilation rate was increased from 2 to 3 ACH and from 6 to 9 ACH, while the increase from 3 to 6 ACH resulted a decrease of exposure index. Similar findings have also been reported (Marshall et al. 1996; Memarzadeh and Xu 2012). It is worth noticing that the above-mentioned studies did not neglect neither the role and the importance of ventilation rate nor the contribution on minimising the airborne transmission. In general, the ventilation rate based on the quantity dilutes the concentration of the infectious airborne bio-aerosols and decreases the risk of transmission. However, based on the velocity and on the air flow pattern, the ventilation rate may lead to the increase of transmission risk. These contradictory effects need to be further studied and evaluated in parallel during the design stage of the ventilation system considering the specific requirements and/or operations of the serviced space.
In addition to the above-mentioned studies, reviews on the role of ventilation rate to the transmission of the airborne infection may be found (Li et al. 2007; Memarzadeh and Xu 2012; Sundell et al. 2011).
The role of space heating and cooling emission system
Space heating and cooling emission units are used to provide energy to end-use space in order to maintain the desirable thermal environment. In general, and considering the main heat transfer mechanism, these units are categorised as free-convention or convector unit, forced-convention or fan-coil unit and radiator or radiant panel unit, or radiant floor/ceiling/wall system. The operation of a convector/ radiator unit or system is associated with thermal plumes that affect the air movement, while a forced-convention unit increases the air velocity in the occupied zone. Both phenomena in conjunction with the ventilation type introduce different temperature and pressure stratification conditions on horizontal and vertical directions; which finally affect the contaminants distribution and the dispersion of the airborne agents into the internal spaces. Several studies analyse the effect of space heating and cooling terminal units in association with different mechanical ventilation systems on the dispersion of pollutants and biological agents. Causone et al. (2010a, b) studied the effect of floor heating system in conjunction with upward displacement ventilation in an experimental chamber. They found that due to the influence of the thermal plume, the actual airflow pattern was between mixing and displacement ventilation, and in case of contaminants production from a heat source, high ventilation rates are required to achieve high ventilation effectiveness. Wu et al. (2014) analysed the ventilation effectiveness of mixing and upward displacement ventilation patterns with floor and ceiling heating systems. They reported that both systems have slightly similar ventilation effectiveness that ranges between 0.97 for the ceiling heating with mixing ventilation system up to 1.14 for the floor heating with displacement ventilation one. Lipczynska et al. (2015) compared the effectiveness of a personalised ventilation with chilled ceiling system against to mixing ventilation, chilled ceiling combined with mixing ventilation and chilled ceiling combined with mixing ventilation and personalised ventilation. They concluded that evaluated personalised ventilation systems was up to 10 time more efficient compared to mixing ventilation ones, and resulted a strongest protection of the occupants from the cross-infection. Jurelionis et al. (2018) accessed the capability of a conjugated floor heating and mixing ventilation system on the dispersion of the air pollutants. They reported that the use of floor heating increased the effectiveness of pollutant dispersion by 5% and reduced the exposure of the occupants by 22% on average. Choi et al. (2019) measured the contaminants concentration profiles in a hospital ward equipped with radiant panel and displacement ventilation. They stated that the heat plume generated by the vertical radiant panel strongly affects the diffusion of the contaminated air. In case of heating operation, the use of radiant panel increases the exposure of a lying patient as the contaminant air is trapped above the lying level. In contrast, during the cooling operation the downward plume drives the exhaled contaminant to the lower high than that of the lying patient, and thus increasing the contaminants concentration in the near to floor levels of the ward. These results proved that the location of the radiant panel and its thermal operation are import parameters which strongly influence the contaminants concentration on the specific levels of the hospital ward. Similar results have also been reported (Causone et al. 2010a, b; Cetin et al. 2020b; Jurelionis et al. 2016; Liu et al. 2019a, b, c, d, e; Olesen et al. 2011; Ouazia et al. 2012; Ouazia et al. 2011; Schiavon et al. 2015; Shi et al. 2019; Wu et al. 2015; Wu et al. 2019a, b; Wu et al. 2020; Zhou et al. 2017). In addition to that, a comprehensive review on the integrated radiant heating and cooling systems in conjunction with the ventilation ones has been reported by Zhang et al. (2020a, b, c, d).
Computer modelling of particles and biological agents’ airborne transmission into indoor built environment
Undoubtedly, mathematical models have proven their value for predicting the high risk and impact of the chemical-biological agents’ exposure on building environment and public health (Argyropoulos et al. 2016, 2018; Bongers et al. 2008). According to Milner et al. (2011) in order to investigate numerically the indoor exposure, a selection of the three following types of IAQ models, namely statistical regression (Valero et al. 2009), micro-environmental (Duan 1982) and CFD models (Béghein et al. 2005; Choi and Edwards 2008, 2012), should be made. The first type involves models employing empirical and semi-empirical approaches to associate indoor environment exposure with significant parameters such as building characteristics, contaminant concentration levels and source strength. The second type of model, adopts the ‘well-mixed’ zone simplification assumption (Axley 2007, 1989; Emmerich 2001) at the building interior for the temperature and contaminant concentration levels and can be further classified into the mass balance (Gerharz et al. 2009; Shrubsole et al. 2012), measurement-based (Kornartit et al. 2010; Ozkaynak et al. 2007), sub-zonal (Megri and Haghighat 2007; Stewart and Ren 2006) and multi-zone models (Argyropoulos et al. 2017a, b; Ashraf et al. 2016; Zhu et al. 2020).
CFD models could be a superior alternative approach to surpass the limitation of the ‘well-mixed’ assumption which does not always hold true especially for non-uniform concentrations in large spaces and transient state (Wang and Chen 2007, 2008a; Wang et al. 2010a, b). CFD models are capable of predicting the airborne transmission of aerosols in indoor spaces, by providing valuable insights into a number of driving factors of the phenomenon such as ventilation system, droplet formation mechanisms, concentrations, turbulence effects, ambient temperature, relative humidity for the survival capability of the agent and on airflow and agent deposition in human airways. However, these models are more computational demanding but more accurate.
A coupling approach of multi-zone and CFD models is preferable for a compromise between computational demands and accuracy (Argyropoulos et al. 2017b; Argyropoulos et al. 2020; Srebric et al. 2008; Wang and Chen 2008a). For detailed evaluation of the all above models, the interested reader is directed to the review papers by Milner et al. (2011) and Wang and Zhai (2016). In the following two subsections, it is presented a review of numerical studies focused mainly on the use of multi-zone and CFD models, as well as its coupling, for investigating the dispersion of airborne pathogens within indoor spaces. Numerical studies related to aircraft and vehicle cabins fall out of the scope of the present study. Finally, few numerical studies based on CFD-PBTΚ (Physiologically-Based ToxicoKinetic) models are also mentioned. This class of models is capable of approximating the kinetic behaviour of contaminants and as a result can assess the internal dose at targeted tissues/organs (Argyropoulos et al. 2018; Feng et al. 2021; Mumtaz et al. 2012).
Single-zone and microenvironment models
This class of models is based on semi-empirical and empirical approaches which include empirical correction factors for a great variety of ingress and egress configurations, as well as different room characteristics. Mass balance models also known as dilution models are deterministic and can be also used for the prediction of indoor contaminant concentrations in different rooms or buildings both spatially and over time.
Chao and Tung (2001), developed an empirical model for the investigation of I/O ratio based on the ventilation influence using a non-steady-state mass balance approach. They found that the air exchange rate has a crucial role to the penetration of outdoor pollutants into residential buildings.
Özkaynak et al. (2008) performed numerical simulations using HAPEM model for estimating the inhalation exposures for over 30 gaseous and particulate pollutants, by examining 37 microenvironments (MEs). The numerical results obtained showed that the predictions are appear to be influenced by the exposure concentration levels due to their dependence on the pollutant type, activity and site. Similarly, Borrego et al. (2006) studied numerically the exposure of concentration levels using an indoor/outdoor function (additional source term) to their model.
A large number of numerical studies has also been devoted to investigate PM (Dimitroulopoulou et al. 2001; Dimitroulopoulou et al. 2006; Nazaroff 2004), element PM (Lunden et al. 2003a, b), airborne bacteria and fungi (Nazaroff 2016), among other contaminants such as CO, Rn, NO2, VOCs and O3 (Briggs et al. 2003; Hicklin et al. 2018; Lee et al. 2004; Li and Niu 2007; Lunden et al. 2003a, b; Mölter et al. 2012).
Multi-zone models
Multi-zone airflow modelling is characterised by great capabilities for simulating the building infiltration, exfiltration and ventilation effects into indoor spaces. Multi-zone models are constituted by a network of elements which represents flow paths (e.g. fans, doorways, opening, cracks and HVAC ducts) among different zones of a building (Fig. 10). Consequently, the air flow rate is calculated from one zone to another as a function of pressure drop along a flow path.Fig. 10 Building layout produced by CONTAM according to available HVAC data Reproduced from Reference (Argyropoulos et al. 2017a, b) with permission from Elsevier
There are many multi-zone simulation programmes such as AIRNET (Walton 1989), CONTAM (Dols and Polidoro 2015; Walton and Dols 2005), COMIS (Feustel 1999; Feustel et al. 1989), BREEZE (Evers and Waterhouse 1978) and CBSAIR (Haghighat and Rao 1991) to name only a few; however, the most popular are CONTAM by the US National Institute of Standards and Technology (NIST) and COMIS by Lawrence Berkeley National Laboratory (LBNL). The first is the most widely used, while a validation study of the last two multi-zone models can be found in the work of Haghighat and Megri (1996). More details for the multi-zone models, the interested reader is directed to the comprehensive reviews by Axley (2007; 1989), Feustel and Dieris (1992) and Emmerich (2001).
According to Dols and Polidoro (2015), the transient partial differential equations for the description of airflow in CONTAM are specified as follows in Eq. (1):1 ∂mi∂t=ρi∂Vi∂t+Vi∂ρi∂t=∑jFji+Fi
where t is the time, mi the mass of air for zone i, Vi the volume for zone i, ρi the density for zone i, Fij the air flow rate from zone j to i and Fi non-flow processes for zone i (remove or add significant amounts of air from i zone). The above terms Fij and mi can be calculated by using the following formulas (Dols and Polidoro (2015)):2 Fij=fPj-Pi
3 mi=ρiVi=PiViRTi
where Pi is the pressure for zone i, Pj the pressure for zone j, f (Pj – Pi) the function of pressure drop n along the flow path, Ti the temperature for zone i and R the ideal gas constant.
Kowalski et al. (2003) performed a multi-zone analysis using CONTAMW (Dols et al. 2000) software for predicting concentration levels and inhaled doses against intentional releases of biological agents into a 40-story commercial office building. They investigated the performance of different cleaning systems such as air-cleaning and air-disinfection systems. They concluded that the combination of ultraviolet germicidal irradiation (UVGI) and filtration as air-cleaning strategy can provide encouraging protection for the occupants and there is no significant improvement beyond the following selected characteristics, i.e. 15% outside air ventilation, filtration of MERV 13–15 and UVGI dose of 1000 μW-s/cm2, for the considered 40-story commercial office building.
An early attempt to conduct multi-zone airflow simulations using CONTAM for studying the severe acute respiratory syndrome (SARS) virus airborne transmission among flats in Block E of the Amoy Gardens was undertaken by Li et al. (2004a, b) and Yu et al. (2004). The numerical results, which describe the evolution of virus spread, presented encouraging agreement with the observed spatial infection data. They concluded that the airborne transmission route was the main reason of SARS spread and building infiltration along with natural ventilation have a positive influence on the infection control. Few years later, Chen et al. (2011) using multi-zone modelling in conjunction with experimental measurements in an environmental chamber found that the air exchange which caused by small temperature differences between cubicles has also significant effect on the transmission of the SARS virus.
Ren and Stewart (2005) modified COMIS with sub-zones (COwZ) for investigating the occupational personal exposure to pollutant sources in a ventilated room. The numerical results were validated by available experimental measurements and CFD data, exhibiting good agreement. They found that the impact of occupant’s location and orientation has significant influence and should be considered for the personal exposure assessment.
Some years later, Lim et al. (2011) performed field measurements of pressure and numerical simulations using CONTAMW to predict both concentration levels and airflow evaluation of virus (H1N1) spread in tall Hospital buildings. Their numerical results showed the possibility of airborne transmission of pathogens through the stack effect within high-rise hospital buildings, presenting encouraging agreement with measurements excluding a few floor cases. Preventive and protection measures were also suggested for minimising the virus spread.
Emmerich et al. (2013) conducted numerical simulations using CONTAM for assessing different control strategies to reduce the dispersion of airborne pathogens (e.g. tuberculosis and squame cells) into a hypothetical hospital. The obtained numerical results indicated that the use of HEPA or MERV-15 filtration have a positive effect on the protection of occupants over pollutant dispersion, as well as UVGI systems. Finally, they also observed that increasing the interior wall leakage by a factor of 5 leads to decrease of pressure difference by a factor of 2.
Recently, numerical simulations were undertaken by Karakitsios et al. (2020) using CONTAM for a hypothetically release of a contaminant within a high-rise building. The simulations examined different scenarios for meteorological conditions, building operational characteristics and source types and location. The obtained results showed that all rooms with ventilation appeared pollutants and there was also transfer of pollutants through leakages towards the stairwell and elevators. Finally, they suggested potential locations for the installation of sensor technologies in order to detect indoor chemical-biological agents.
The same year, Zhu et al. (2020) investigated experimentally and numerically the ventilation effect in two actual building geometries (residence halls) during an entire flu season. By collecting CO2 measurements, they calibrated multi-zone models (CONTAM) in order to simulate airborne transmission of influenza A within adjacent rooms and predict the concentration levels (exposure) for room occupants. The opening doors and windows of dormitory rooms within low ventilated building can improve the ventilation rates, however, this operation sacrifices the thermal comfort (e.g. low outdoor temperatures) of the room occupants. Their numerical results indicated that there is a strong trend between the low outdoor air supply and respiratory infection rates into dormitory rooms; however, more studies are needed to confirm their findings. They also concluded that the cross-infection risk for airborne transmission of influenza A should be considered based on the airflow map rather than the spatial distribution among the occupants’ rooms.
Computational fluid dynamics models
In many cases, computational fluid dynamics (CFD) modelling is the best alternative for investigating the airborne transmission in indoor ventilated spaces, as well as the transmission from human body motion, talking, coughing and breathing.
The mathematical representation of air flow in indoor spaces, based on the set of elliptic, partial differential equations, expressing the mass conservation, momentum, continuity, energy, chemical species concentration and turbulence parameters can be all written in the following general form (Eq. (4)) (Patankar 1980):4 ∂ρϕ∂t+divρuϕ=divΓϕgradϕ+Sϕ
where ρ is the air density, t the time, ϕ the dependent variable (u, υ or w for momentum, 1 for continuity, h for enthalpy, c for chemical species concentration, k the kinetic energy of turbulence and ε the eddy dissipation rate), u is the velocity vector of air, Γϕ the effective exchange coefficient of variable ϕ (1 for continuity) and Sϕ the source/sink term of variable ϕ. The four terms in Eq. (4) represent the unsteady, convection, diffusion and source terms, respectively.
An important factor of CFD modelling for examining the airborne transmission in indoor built environments is the effect of turbulence motion on the pathogen spread and mean flow field. In the literature, the most of the relevant CFD studies are based on Reynolds-Averaged Navier Stokes (RANS) models (Satheesan et al. 2020; Tao et al. 2020; Wang et al. 2020; Ye et al. 2020; Zhang et al. 2020a) for treating turbulence, and only several Large Eddy Simulation (LES) studies (Berrouk et al. 2010; Liu and You 2011; Tian et al. 2007; Vuorinen et al. 2020; Zhang et al. 2019) have been conducted the previous decades, however, an increasing number of new LES articles due to the SARS-CoV-2 pandemic period is published, as well as an integrated DNS approach for the prediction of cough/sneeze flows by Diwan et al. (2020).
In general, the selection of the appropriate turbulence model for predicting airflow and cross-infection risk in ventilated spaces including the dispersion of airborne pathogens among occupants (e.g. through talking, sneezing, breathing, coughing) is of great challenge due to the complexity of the physical phenomenon (e.g. human body micro-environment, buoyancy, contaminant concentrations, convection, circulation, reattachment and vortices (Zhai et al. 2007). Regarding RANS models for human body micro-environment, the most used are the RNG k-ε and low Reynolds number k-ε (Gao and Niu 2005; Nielsen 2015). An interesting evaluation and comparison of eight different turbulence modelling approaches (i.e. RNG k-ε, SST k-ω, low Reynolds number Launder & Sharma (LRN-LS) k-ε, v2-f), detached Eddy simulation (DES) and LES) and available experimental data from the literature for the prediction of airflow in enclosed environments can be found in the work of Zhang et al. (2007). Subsequently, Phuong and Ito (2015) compared four different turbulence models (LRN-LS k-ε, LRN-ΑΚN (Abe-Kondoh-Nagano) k-ε, RNG k-ε and SST k-ω) against PIV measurements for investigating the flow distribution in upper human airway including oral and nasal inhalation. More details about the equations, advantages, limitations and implementation of different turbulence modelling approaches, the interested reader is directed to the review paper by Argyropoulos and Markatos (2015). Finally, a recent paper by Foster and Kinzel (2021) also presents a useful comparison between CFD models and Wells–Riley mathematical models for SARS-CoV-2 spread into classrooms.
Another important parameter to investigate pathogens transport and trajectory using advanced CFD techniques is the selection of the suitable multiphase model in order to study phenomena such as droplet evaporation, turbulence dispersion forces, droplet breakup and coalescence, among others (Dbouk and Drikakis 2020a; Löhner et al. 2020). It is common practice to choose an Eulerian approach for the gas phase, while the particle (bioaerosol) transport can be simulated using both a Lagrangian or an Eulerian method (Crowe et al. 1996). According to Eulerian–Lagrangian approach, the liquid phase is treated by a Discrete Droplet model; while for the Eulerian-Eulerian method, a Continuum Formulation model is adopted (Novozhilov 2007). Both methods have advantages and drawbacks, while many researchers have investigated extensively their limitations and applications (Zhang and Chen 2007). The mathematical formulation of the aforementioned models along with useful information for their implementation is not repeated herein, due to space limitations, but it may be found in the classical textbooks by Yeoh and Tu (2010), Brennen (2005), and Azzopardi (2006) and review papers by Crowe et al. (1996), Peng et al. (2020). Finally, the equations for the motion of particles/droplets and virus loads may be found (Löhner and Antil 2020; Löhner et al. 2020).
Numerical studies focused on the infection spread into chambers and offices
SHAO et al. (2021) performed CFD simulations using OpenFOAM in conjunction with in-situ measurements to investigate the airborne transmission risk of SARS-CoV-2 by asymptomatic individuals into small classroom, elevator and supermarket. They found that the design of ventilation system in confined spaces plays a major role in the particle removal and deposition. Bad design of ventilation system results in decreasing of particle removal efficiency and increasing of particle deposition in which both increase the risk of contamination. Similarly, Vuorinen et al. (2020) investigated numerically the dispersion and inhalation of droplets in relation to SARS-CoV-2 for open office and supermarket, using an LES approach (Fig. 11). They examined four different open sources CFD codes, namely, PALM, FDS, OpenFOAM and NS3dLab, while a number of Monte Carlo simulations was also conducted to investigate susceptible and infected individuals.Fig. 11 Visualisations demonstrating the effect of particle size (and mass) on the modelled spreading of the cough-released aerosol cloud. For better sense of scale, bystanders are placed 8 m from the coughing person. Instantaneous views on the state of the cloud are shown for realisations where the particles have a no mass, b 1000 kg m−3 density and 10-μm diameter and c 1000 kg m−3 density and 20-μm diameter. Images on the left column are at t = 20 s and on the right column at t = 120 s. Below, d presents the time evolution of the mean elevation of the 99th percentile concentration highlighting the different descent rates. Droplets in these size scales have τevap < 1 s and they would become aerosol-like droplet nuclei very rapidly.
Reproduced from Reference (Vuorinen et al. 2020) with permission from Elsevier
Several relevant LES studies, including ventilation effects, different sub-grid scale models (e.g. WALE, Deardorff model, Smagorinsky) and CFD codes (e.g. ANSY FLUENT and CFX, PHOENICS, OpenFOAM, Star-CCM +), have also published in the literature (Béghein et al. 2005; Berrouk et al. 2010; Choi and Edwards 2008, 2012; Dudalski et al. 2020; Feng et al. 2020a, b; Fontes et al. 2020; Karakitsios et al. 2020; Pendar and Páscoa 2020; Tian et al. 2007; Zhang et al. 2019). It is worth mention that Diwan et al. (2020) also developed a DNS approach for the prediction of cough/sneeze flows. According to their temperature profile results, the dry cough (without liquid droplets) flow was dispersed very fast (cough duration of 0.58 s) at a distance of more than 1 m.
Pendar and Páscoa (2020) proposed a fully coupled Eulerian–Lagrangian method based on the OpenFOAM code for investigating the dispersion of saliva microdroplets generated by sneeze and cough in indoor environment. Their numerical results showed that the use of mask and a full bending of our head during sneeze can reduce significantly the risk of infection. More specifically, the latter action can cause decreasing of the microdroplets travelling distance by > 22%, while the first action can restrict the risk infection in a transmission sphere area of 0.6 m diameter. They also claimed that the social safety distance of 2 m should be increased to 4 m for providing more effective protection.
Feng et al. (2020a, b) conducted LES using ANSYS 17.0 in order to investigate the influence of human microenvironment on the transmission of infection diseases via microbial particles during human respiratory. They showed that an increase of heat flux leads to increase of the air flow flux of the thermal plume, resulting in a further increase of thermal plume ability to transfer particles upward. One year later, Zhang et al. (2019) employed an LES model combined with Lagrangian approach for studying the spread and transmission of bacteria and virus in a ventilated room. The numerical results obtained compared with experimental data from a climate chamber, presenting good agreement. They concluded that the droplet cloud velocity, which is characteristic for respiratory activities such as coughing and breathing, has great influence on the accuracy of the simulation. Choi and Edwards (2012) investigated via LES combined with an Immersed Boundary Method, the contaminant spread in room compartments. The immersed boundary method used for considering heat transfer effects and passive scalar advection. The numerical results obtained were validated by available experimental and CFD data, exhibiting good agreement. Fontes et al. (2020) performed DES for the investigation of human physiology factors (e.g. nasal and buccal passages, with or without teeth) during the human respiratory event of sneezing on the airborne virus transmission. They found that saliva properties have significant effect on the spray formation (i.e. droplet distribution, primary and secondary break-up mechanisms). They also claimed that women seem to be less effective on the transmission of airborne pathogens.
Special attention has also been given to pathogen transmission using Reynolds-averaged Navier Stokes (RANS), unsteady Reynolds-averaged Navier Stokes (URANS) and Reynolds Stress (RS) models associated with ventilation strategies for particle removal and dispersion (Cetin et al. 2020a; He et al. 2011; Katramiz et al. 2020; Murga et al. 2020; Park and Chang 2019; Shao et al. 2020; Wang et al. 2020), human movement (Li et al. 2020; Tao et al. 2020; Tao et al. 2019), comparison of Eulerian-Eulerian and Eulerian–Lagrangian approaches for the pathogens transport and trajectory (Yan et al. 2020) and human expiratory events (e.g. coughing, sneezing, speaking) (Chen and Zhao 2010; de Oliveira et al. 2021; Kang et al. 2015; Li et al. 2018; Licina et al. 2015; Liu et al. 2016a, b; Yan et al. 2019; Zhang et al. 2020a, b, c, d; Zhu et al. 2006), among others.
Ji et al. (2018) investigated numerically the effects of evaporation process of pure water droplet under different RHs (0%, 30%, 90%) and ventilation strategies (displacement and mixing). They concluded that the evaporation process for small droplets occurs rapidly and it is difficult to observe differences between mixing and displacement ventilation. However, RH has small effect on large droplets’ deposition, while displacement ventilation can delay evaporation similar to high RH.
Al Assaad et al. (2018) and Katramiz et al. (2020) performed numerical simulations using the RNG k-ε turbulence model for the investigation of intermittent personalised ventilation with respect to the protection of occupants from indoor contaminants. Their results showed that a selected average flowrate of 7.5 L/s along with an operating frequency of 0.86 Hz are acceptable for providing good ventilation and thermal comfort conditions in order to protect occupants. They also extended their study for the effect of walking occupant on the personalised ventilation in an office (Al Assaad et al. 2019a) and particle resuspension in a prayer room related to human prostration cycle (Al Assaad et al. 2019b). They concluded that the human prostration cycle due to prayers plays an important role in the particle spread from the floor to the upper levels of the confined space, while higher risk of contamination in the breath zone was found in the case of 1 μm particle concentration compared to 10 μm, according to the examined scenarios.
Dbouk and Drikakis (2020a) conducted RANS simulations combined with the k-ω turbulence model, by using the open-source code OpenFOAM. They investigated the spread of saliva droplets generated from a human cough in order to predict the influence of wind on social distancing. Their results showed that in the absence of wind effect the majority of exhaling saliva droplets during a cough can travel up to 1-m distance, while a small number can be travelled further. However, these droplets present low risk due to the low trajectory (< 1 m height). On the other hand, with the presence of wind speed in the range of 4–15 km/h, the travelled distance of the saliva droplets can reach up to 6 m, which is much farther than the recommended social safety distance of 2 m. Dbouk and Drikakis (2020c) presented a continuation of their previous study (Dbouk and Drikakis 2020a) with the aim at extending their work to consider the unsteady evaporation process of the saliva droplets, relative humidity, temperature and wind speed. They concluded that the low relative humidity combined with high temperature foster the droplet evaporation rate, resulting in significant reduction of virus viability. Similarly, Feng et al. (2020a, b) examined the transmission of SARS-CoV-2 droplets between two human bodies by means of RANS approach including evaporation and condensation effects (Fig. 12). They showed that the recommended social distance of 1.83 m (6 ft) is not sufficient to provide protection to people, under different wind conditions and static air environment (exposure at 3.05 m (10 ft)), from SARS-CoV-2 during coughing. Moreover, deposition and transport of droplets are dependent on the wake flow patterns and secondary flow between the two human bodies. Their results also indicated that high RH (99.5%) increases the deposition of droplets in the space, however, without increasing necessarily the risk of exposure. On the other hand, medium RH (40%) fosters the water evaporation phenomenon, resulting in decreasing of droplet diameter and remaining airborne for longer times. Similar studies including evaporation and condensation effects results obtained for coughing from one person have also been reported (Chen and Zhao 2010; Li et al. 2018; Yan et al. 2019).Fig. 12 Schematic of the computational domain with two virtual humans and the hybrid mesh details.
Reproduced from Ref (Feng et al. 2020a, b) with permission from Elsevier
Numerical studies focused on the infection spread in hospitals and patient wards
A large number of numerical studies have been undertaken by many scientists and engineers for preventing the nosocomial airborne infection in hospitals and patient wards including ventilation and turbulence effects (Qian and Li 2010; Saarinen et al. 2015; Seymour et al. 2000; Shajahan et al. 2019; Wan et al. 2007; Yang 2013), while the current number of relative published papers (Anghel et al. 2020; Borro et al. 2020; Gu et al. 2020; Satheesan et al. 2020; Villafruela et al. 2019; Wang et al. 2021) is continuously increasing due to SARS-CoV-2 pandemic. This is mainly attribute to the strong interest in the SARS-CoV-2 Coronavirus modes transmission among patients, visitors and healthcare personnel in order to protect them.
An early attempt to perform RANS computations using a k-ε turbulence model for the prediction of airborne pathogens transmission in a hospital isolation room, including the effects of ventilation systems, was undertaken by Seymour et al. (2000). Furthermore, Li et al. (2004a, b) conducted CFD simulations to investigate the spread of virus-laden bio-aerosols in a hospital ward during SARS outbreak in Hong Kong, by using the commercial CFD code Fluent 6.1. The numerical results showed that the predicted spread of the viral respiratory disease is in good agreement with the reported SARS cases. Chau et al. (2006) also examined the effects of the local exhaust ventilation system in a hospital patient ward for the protection of healthcare workers from virus diseases such as SARS.
Huang and Tsao (2005) presented numerical and experimental results for the removal of airborne pathogens in negative pressure isolation rooms. Their results showed that the buoyancy effects play an important role to flow and the removal of bacteria, while the redesign of the isolation room can improve the pathogen’s removal. Qian and Li (2010) performed numerical simulations and experiments for studying the ventilation and deposition effects in a six-bed room. They presented CFD simulations using the RNG k-ε turbulence model along with a Lagrangian method for the prediction of particles trajectory. The numerical results, which describe the characteristics of the flow and the distribution of exhaled particles, indicated that the removal of particles is achieved more efficiently by ceiling-level exhausts compared to floor-level exhausts. In a similar way, Yang (2013) investigated the different types of ventilation in a four-bed sickroom using the commercial CFD code Star CD, while Chao et al. (2008) presented numerical and experimental results for the characteristics of the expiratory droplets in a three-bed hospital ward. Recently, Satheesan et al. (2020) presented numerical results for the Middle East respiratory syndrome coronavirus (MERS-CoV) in a six-bed inpatient ward.
King et al. (2015) also exhibited CFD simulations using an RSM closure model in conjunction with particle deposition data for predicting the cross-contamination risk among healthcare workers in single- and four-bed isolation rooms. Their results showed that the cross-infection risk in a single-patient room can be decreased significantly, while the ventilation, infection patients’ location, type of patient’s care and room layout may also affect the infection spread inside four-bed rooms (Sadrizadeh et al. 2014a, b; Sadrizadeh et al. 2018; Sadrizadeh et al. 2014a, b) and Wang et al. (2019a, b) studied numerically the effects of door opening on airborne particle movement, as well as the ventilation and stuff number in operating rooms during simulated surgery. They concluded that the use of a positive-pressure system can be more effective to reduce the airborne particle spread, while the door opening combined with the ventilation system and increased number of staff may expand the contamination risk for the patient into the surgical site.
Borro et al. (2020) performed URANS simulations combined with a Lagrangian approach for the investigation of ventilation system at the Vatican State Children’s hospital (Fig. 13). The numerical results indicated that the proposed methodology is capable of predicting the contamination risk and optimising the ventilation flow in hospitals. They also showed that the installed HVAC system can diffuse the formed droplets from a coughing event, while the turbulence effects of the flow also enhance the pathogens spread and particle suspension for longer time in the room. Finally, they concluded that the use of a LEV unit placed above the face of patients can remove the particles and infected air in just a few seconds after the cough event.Fig. 13 Prospective view of the Scenarios A, B and C at t = 1 s (left) and 5 s (right). The spheres represent the droplets coloured by the diameter size (top right legend). The contaminated air is represented by different iso-surfaces coloured by mass fraction.
Reproduced from Reference (Feng et al. 2020a, b) with permission from Elsevier
Gu et al. (2020) developed and demonstrated a numerical simulation framework based on LES approach and FDS software, for assisting the design of ventilation systems in temporary hospitals, such as the first SARS-CoV-2 Wuhan Huoshenshan hospital in China. The numerical results showed that the proposed methodology is capable of assisting HVAC engineers to select and design the appropriate ventilation system in temporary hospitals. Finally, they claimed that there is no case for contamination risk to the surrounding buildings or the fresh-air intakes due to the release of the infected air from the air outlets of the temporary hospital.
Numerical studies focused on the preventive role of mask against airborne droplet transmission
A significant number of concerns has been raised due to the SARS-CoV-2 pandemic for the efficacy of face masks and coverings in controlling and limiting the transport of infective droplets which are formed during cough and sneeze events. Special attention, therefore, is given to investigate the effectiveness of face masks regarding the transmission of respiratory droplets and the recommended social distancing guidelines, respectively.
An early attempt to investigate the aerodynamics of a gas mask canister numerically and experimentally was undertaken by Li (2009). The numerical and experimental results showed that the proposed methodology can be a useful tool for the design of gas mask canister, even though with a low respiratory drop.
Lei et al. (2012a, b) proposed a CFD approach for the investigation of studying the leakages between a headform and an N95-filtering facepiece respirator (FFR). The numerical results were compared with infrared images of respiratory leakage. Their results also indicated that the use of N95 FFR may cause thermal discomfort due to the temperature increase near the lip. They concluded that the most leak presented at the region of nose (40%), left (26%) and right (26%) cheek. The same group (Lei et al. 2012a, b) also investigated numerically and experimentally the effect of pressure contact on digital headforms.
Dbouk and Drikakis (2020b) performed multiphase CFD simulations for the prediction of the droplet transmission from a headform with and without a surgical mask. Their results showed that during a mild cough event the droplets can reach up to 70-cm distance without the use of surgical mask, and wearing mask the droplets may travel about the half above mentioned distance. They also observed that after 10 cough cycles the efficiency of the surgical mask can be reduced by ~ 8%, while for severe cough events the efficiency drops significantly. Finally, the diameter of the transmitted droplets without the presence of mask on the headform was larger across the cough cycles.
Khosronejad et al. (2020) performed LES using very fine grids for the investigation of saliva droplets transmission during a cough event with and without facial mask (Fig. 14). They also examined the effects of indoor and outdoor conditions during the cough event, namely stagnant background air and unidirectional mild breeze. Their numerical results showed that during a cough event without mask and stagnant background air condition the travelling distance of fine droplets can reach up to 2.62 m, while the larger in diameter droplets fall down in the area between the human and the previous mention distance in less than 2 min.Fig. 14 Simulated evolution of the 10-µm saliva particulate concentration (volume fraction) after the cough under outdoor conditions (mild breeze) without (top) and with (bottom) the facial mask. [(a) and (f)], [(b) and (g)], [(c) and (h)], [(d) and (i)] and [(e) and (j)] show the simulated saliva particulate concentration fields after 0.24 s, 0.3 s, 0.4 s, 0.5 s and 0.6 s, respectively, on the sagittal plane. The outdoor simulations were stopped after 0.6 s, when the saliva particulates travel ∼2.0 m and 2.2 m without (top) and with (bottom) the facial mask, respectively.
Reproduced from Reference (Khosronejad et al. 2020) with permission from AIP
Furthermore, a number of fine droplets can also be remained suspended for several minutes in the air. They also observed that the wearing of a medical and non-medical mask can reduce the travelling distance of saliva droplets at 0.48 m and 0.73 m, respectively. Finally, the droplet evaporation phenomenon can increase the travelling distance to 2.84 m without wearing mask and to 0.91 m for using non-medical mask.
Numerical studies focused on the airflow and aerosols deposition in human airways
CFD modelling can also be helpful to investigate the influence of airflow and aerosols deposition in human airflow. More details regarding the transport of particles and characteristics of the transitional flow mechanisms in the human lungs may be found in the recent review papers by Islam et al. (2020) and Mutuku et al. (2020a).
Ito (2014) proposed an integrated method for investigating the airborne infection transmission of pathogens in a hospital using a combination of CFD and SIR epidemiological models. This approach can allow the consideration of the hospital space in conjunction with the human nasal airway. As a result, the proposed methodology is capable of evaluating the exposure risk of occupants and estimate the contaminant dose. Phuong and Ito (2015) performed RANS simulations using four different turbulence models (i.e. two LRN k-ε, RNG k-ε and SST k-ω) to investigate the airflow in human realistic respiratory tract for three constant breathing conditions (7.5, 1.5 and 30 L/min). The numerical results obtained with LRN-AKN model were compared with PIV measurements, presenting better agreement. Recently, an extension of the previous work was proposed by Phuong et al. (2020).
Haghnegahdar (2019a) et al. developed a Computational Fluid Particle Dynamics (CFPD) model combined with Host Cell Dynamic (HCD) model (Fig. 15) for the prediction of influenza A virus droplets trajectory and deposition in the pulmonary tracts.Fig. 15 The framework of the multiscale CFPD-HCD model for the human-to-human IAV infection with a subject-specific airway geometry. The description of the HCD model is given in the ‘The human respiratory system’ section of the original paper (Haghnegahdar et al. 2019a). The detail of the final polyhedral-core mesh is provided at the right nostril and an airway outlet (RUL: right upper lobe, RML: right middle lobe, RLL: right lower lobe, LUL: left upper lobe, LLL: left lower lobe).
Reproduced from Reference (Haghnegahdar et al. 2019a with permission from Elsevier
Their numerical results showed that the proposed model is capable of predicting the spread of virus and population variations in the upper airways tissues. They also predicted particle deposition fractions values of 26.4%, 23.7% and 24.1% for droplet mass fraction of 0, 0.068 and 0.104, respectively, in the oral cavity, while for the nasal cavity the fraction values are 48.1%, 45.2% and 47.6%, respectively. Finally, the average diameter of deposited droplets on the oral cavity is less than nasal cavity.
Mutuku et al. (2020b) performed CFD simulations for investigating the characteristics of airflow and particle deposition effects of PM2.5 on healthy and Chronic Obstructive Pulmonary Disease (COPD) patients. The numerical results showed that the deposition fractions are between 0.12% and 1.18% for healthy case and between 0.05% and 0.49% for COPD case, while carina 5 was found to be the most important place of particle deposition.
Coupling of multi-zone and CFD models
Multi-zone models suffer from the well-mixing assumption which clearly is not valid in cases for Archimedes number (Ar) smaller than 400 and dimensionless temperature gradient (τ) greater than 0.03 (Wang and Chen, (2008b). To surpass this issue, a combination of a multi-zone model and CFD model can be adopted which is superior for more realistic prediction of pollutant concentration levels and airflow characteristics. The coupling of the models provides a satisfactory compromise between accuracy and computational sources.
Wang and Chen (2008a) presented a coupling approach of CFD and multi-zone model for estimating the concentration levels in case of chemical-biological-radiological agent release within complex three-floor building. They showed that the combination of CFD and multi-zone models is superior and capable of identifying the optimal location of emergency sensors, ventilation strategies for emergency response, as well as to examine proposed routes for evacuation.
Jiang et al. (2009) performed multi-zone simulations for predicting the virus concentration and the required ventilation rate for sufficient air dilution in two Hospitals in Beijing and Guangzhou, respectively. It is worth mentioning that the pressure coefficient was predicted by the commercial CFD software PHOENICS 3.2. (Spalding 1981) and used as input parameter for CONTAM model. Their numerical results were validated against field experiments using tracer gas (SF6), with promising and encouraging results.
Recently, Karakitsios et al. (2020) used COMIS for calculating the inflow and outflow conditions from the openings (windows and doors) and then induced them into ADREA-HF CFD code (Efthimiou et al. 2018; Kovalets et al. 2018) in order to investigate the release of a hazardous agent through the HVAC system in a large office.
CFD-PKTE or CFD-PTBK models
Another interesting combination of models for the investigation of the transport and deposition of particles into human airways is CFD-physiologically based pharmacokinetic (PBPK) or CFD-physiologically based toxicokinetic (PBTK) (Mumtaz et al. 2012), respectively. Recently, Feng et al. (2021) presented a detailed tutorial paper regarding the development, implementation and validation of CFD-PBPK and CFD-PBTK models for investigating the human lung aerosol dynamics numerically.
Yoo and Ito (2018) proposed a computational framework based on CFD, Computer Simulated Person (CSP) and PBPK models for the prediction of inhaled formaldehyde internal dose at human respiratory system. The numerical results indicated that the computational framework is capable of tackling many different types of pollutants and not only the examined formaldehyde. It is also important to mention that the proposed numerical methodology can also provide useful information regarding the exposure to pollutants and health risk assessment into indoor environments. The same group of researchers extended their work (Yoo and Ito (2018) for unsteady breath conditions using the aforementioned computational approach, predicting different concentration levels of formaldehyde inside the room and around the human zone, while for the person breathing zone the concentration values were lower than inside the room.
Murga et al. (2019) conducted health risk assessment in a working environment for the toxic inhalation of breathing air and how affects the human respiratory system, by means of CFD, CSP and PBTK models. The results revealed that the nose area is primarily influenced in all examined cases according to the considered working conditions and there is high risk of acute exposures during the working period.
Haghnegahdar et al. (2019b) also developed a CFD-PTBK model for investigating the transport of xenon gas and how the inhaled dose affects the human body. The numerical results obtained were compared with experimental data, exhibiting good agreement. Finally, the multiscale model is capable of predicting the concentration levels of xenon in the human respiratory system and can also be used for future non-invasive studies regarding patient specific pulmonary diseases.
Conclusions—future directions
The nature and physicochemical characteristics of particles, either those being solids or liquids forming droplets, play an important role in the mode of transmission of a variety of pollutants, contaminants and biological agents in indoor air environments. Understanding the engineering aspects of particle technology plays a major role in designing better depollution or prevention of pollutants and biological agents’ strategies and as a result minimising the risk of transmission in indoor air environments. The current study shade light on the possible gaps and directions for future research in the field of particles transmission in indoor air, focusing especially on biological agents’ transmission:Only a handful of studies conducted with the application of modern techniques capable of detecting sub-micrometer-sized particles, it is important that more work is done in this area to develop a better understanding of the mechanism of droplet generation (Morawska 2006).
It is evident that aqueous solution of mucin alone cannot fully represent various physicochemical and biophysical properties of saliva. Systematic studies on designing saliva targeted tribological properties have to be investigated in future (Sarkar et al. 2019).
However, studies of how surface contamination is propagated by human touching are scare as there are no experimental data (Xiao et al. 2018).
Both studies on the absolute and relative humidity which are known to affect the viral survival need to be further investigated (Poon et al. 2020).
The fundamental science underlying the virus –microorganisms transfer mechanisms on soft matter domain (Poon et al. 2020).
Studies of how the solid surfaces contamination is propagated by human touching are scarce due to mainly the lack of experimental data (Xiao et al. 2018) due to complexity of such types of experiments and quite intense health and safety protocols, laboratories mainly accessible by medical scientists and difficulties in introducing other disciplines in the field etc.
Better understanding the pollutants in general and even more biological agents’ mechanisms in the deeper generation parts of the human tracheobronchial system.
The mechanisms of how the droplets are formed near the mouth has not been studied (Vadivukkarasan et al. 2020)
Sneezes especially have received much less attention in literature and is a field which needs further investigation (Scharfman et al. 2016).
Concentration of biological agents in the droplets (Zhang et al. 2020a, b, c, d).
Viral survival on the skin (Zhang et al. 2020a, b, c, d).
Dependence of evaporation on the temperature and humidity regarding seasonal and geographic variations in transmission rates (Tang et al. 2009).
The installed ventilation system plays significant role on the transmission of the pollutants and biological agents into indoor spaces. The positive and/ or negative influences of either mechanical systems or natural one have been extensive analysed and documented in a series of studies. Hereafter some of the main conclusions regarding HVAC systems are summarised:Mixing ventilation leading to well-mixed homogeneous bio-aerosols and high dispersion rates, regardless of the droplet sizes.
Upward displacement ventilation provides high efficiently removing of small size droplets.
Downward displacement ventilation is ideally removing the contaminated indoor air, and minimise the cross-infection risk.
Human’s thermal plume and walking velocity are significantly influence the dispersion mechanism and kinetics of the bio-aerosols.
Relative position and orientation among occupants are critical parameters that influence the cross-infection risk. Face-to-face position and upward exhaled bio-aerosol airflow are of high cross-infection risk.
Personalised ventilation decreases the cross-infection risk of the user and increases the risk of cross-infection to the non-personalised occupants of the space.
Natural ventilation minimises the cross-infection risk due to high airflow rates and mixed airflow distribution; however, in some cases, the concentration of pollutants in the unfiltered air is significant high.
Thermal plumes from radiant heating and convective heating and cooling panels strongly influence the contaminants concentration and the associated cross-infection risk. In general, upward thermal plumes in-line with the airflow pattern have positive effect on the dispersion of airborne agents, while the downward of crossflow ones might need higher airflow rates in order to maintain the ventilation effectiveness.
In addition to the above reporting findings, it is worth noticing that today the main scientific interest on ventilation systems has been turn to the more sophisticated ones, such as personalised systems, which still remaining in a developing stage. The opportunities for future research and the still remaining open research questions in this area have been recently presented by Zhai and Metzger (2019).
Since the 1970’s, the first documented attempts to use CFD in ventilation industry (Chow 1996; Nielsen 2015), the progress of CFD has been tremendous for indoor environments, with promising results for the prediction of pollutant dispersion and concentration levels, as well as for the design of ventilation strategies. Nowadays, the further development of CFD techniques along with the continuing progress of computer-hardware development has established the use of CFD as the main tool for the prediction of air movement and design of HVAC systems for the controlled ventilation of indoor spaces. In the current pandemic of SARS-CoV-2, CFD simulations have played a major role to investigate the airborne saliva droplets transmission among people in enclosed spaces. Below, we present some final comments regarding the aforementioned computational approaches for the pollutants in general and biological agents’ airborne transmission into indoor built environment:Human’s thermal plume and walking velocity are significantly influenced by the dispersion mechanism and kinetics.
More research should be devoted to evaluate the probability of droplet vs. viral transmission during airborne droplets transport and coughing (Dbouk and Drikakis 2020a).
Regarding face masks and protection from the dispersion of airborne infected saliva droplets further research must be directed to the composition, properties of saliva droplets and mask high-filter efficiency for the prediction of airborne droplet transmission (Dbouk and Drikakis 2020b).
LES seems to be the most appropriate method for practical computation for the investigation of droplet transmission, however, it is time consuming and computationally demanding. Furthermore, there are still challenges such as the development of advanced sub-grid scale models, high-order discretisation schemes for the elimination of the numerical errors, implementation on unstructured grids, and interaction with other physical mechanisms (Argyropoulos and Markatos 2015).
The combination of CFD and multi-zone models can be very useful for more realistic prediction of pollutant concentration levels and airflow characteristics, and can provide a compromise between accuracy and computational sources.
It is important to mention that the CFD-PKTE or CFD-PTBK models for the transport prediction of particles in human respiratory system exhibit many difficulties and should be further improved by developing the next generation of virtual lung computational framework (Feng et al. (2021).
There is also a need for further improvement and validation of the current numerical methods in order to be fully capable of predicting accurately complex phenomena of the biopathogens’ transmission mechanisms, such as evaporation, dispersion, droplet distribution, primary and secondary break-up mechanisms, coalescence, turbulence, inhalation and pulmonary transport.
Acknowledgements
The authors gratefully acknowledge Mrs. Eirini Kyritsi and Mr. Christos Italos for the visualisation of the enclosed figures, and the anonymous reviewers for their valuable recommendations and efforts.
Funding
Dr. Skoulou received partial support from the European Commission H2020 MSCA programme (DEW-COOL-4-CDC project, Grant agreement ID: 734340).
Data availability
Data sharing is not applicable to this manuscript as no datasets were generated or developed during the current study. All the included information has been retrieved from the existing literature.
Declarations
Ethics approval and consent to participate
Not applicable.
Conflict of interest
The authors declare no competing interests.
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
Ahmed N, Glencross H, Wang Q (2016) Biomedical Science practice: experimental & professional skills. Oxford University Press, Oxford
Ahnrud GP, Mendoza AJ, Hurley MJ, Marek PJ (2018) Efficacy of a sonicating swab for removal and capture of microorganisms from experimental and natural contaminated surfaces. Appl Environ Microbiol 84(9):208–218. 10.1128/aem.00208-18
Ai Z Hashimoto K Melikov AK Airborne transmission between room occupants during short-term events: measurement and evaluation Indoor Air 2019 29 4 563 576 10.1111/ina.12557 30980555
Ai ZT Huang T Melikov AK Airborne transmission of exhaled droplet nuclei between occupants in a room with horizontal air distribution Build Environ 2019 163 106328 10.1016/j.buildenv.2019.106328
Ai ZT Mak CM A study of interunit dispersion around multistory buildings with single-sided ventilation under different wind directions Atmos Environ 2014 88 1 13 10.1016/j.atmosenv.2014.01.049
Ai ZT Mak CM Large eddy simulation of wind-induced interunit dispersion around multistory buildings Indoor Air 2016 26 2 259 273 10.1111/ina.12200 25787963
Ai ZT Mak CM Niu JL Numerical investigation of wind-induced airflow and interunit dispersion characteristics in multistory residential buildings Indoor Air 2013 23 5 417 429 10.1111/ina.12041 23495766
Ai ZT Melikov AK Airborne spread of expiratory droplet nuclei between the occupants of indoor environments: a review Indoor Air 2018 28 4 500 524 10.1111/ina.12465 29683213
Al Assaad D Ghali K Ghaddar N Effect of flow disturbance induced by walking on the performance of personalized ventilation coupled with mixing ventilation Build Environ 2019 160 106217 10.1016/j.buildenv.2019.106217
Al Assaad D Ghali K Ghaddar N Particles dispersion due to human prostration cycle and ventilation system in a prayer room Build Environ 2019 150 44 59 10.1016/j.buildenv.2019.01.005
Al Assaad D Ghali K Ghaddar N Habchi C Coupled CFD and particle resuspension models under combined effect of mechanical and aerodynamic disturbances Build Environ 2020 169 106567 10.1016/j.buildenv.2019.106567
Al Assaad D Habchi C Ghali K Ghaddar N Effectiveness of intermittent personalized ventilation in protecting occupant from indoor particles Build Environ 2018 128 22 32 10.1016/j.buildenv.2017.11.027
Aliabadi AA Rogak SN Bartlett KH Green SI Preventing Airborne disease transmission: review of methods for ventilation design in health care facilities Adv Prev Med 2011 2011 124064 10.4061/2011/124064 22162813
Allocca C Chen Q Glicksman LR Design analysis of single-sided natural ventilation Energy Build 2003 35 8 785 795 10.1016/S0378-7788(02)00239-6
Anghel L, Popovici C-G, Stătescu C, Sascău R, Verdeș M, Ciocan V, Serban I-L, Mărănducă MA, Hudișteanu S-V, Țurcanu F-E (2020) Impact of HVAC-Systems on the dispersion of infectious aerosols in a cardiac intensive care unit. Int J Environ Res Public Health 17(18):6582 10.3390/ijerph17186582
Argyropoulos CD, Abraham M, Hassan H, Ashraf AM, Fthenou E, Sadoun E, Kakosimos KE (2016) Modeling of PM10 and PM2.5 building infiltration during a dust event in Doha, Qatar. Paper presented at the 2nd International Conference on Atmospheric Dust, Castellaneta Marina, Italy
Argyropoulos CD Ashraf AM Markatos NC Kakosimos KE Mathematical modelling and computer simulation of toxic gas building infiltration Process Saf Environ Prot 2017 111 687 700 10.1016/j.psep.2017.08.038
Argyropoulos CD, Ashraf AM, Vechot L, Kakosimos KE (2017b) Coupling multi-zone and CFD models for investigating indoor air quality. Paper presented at the Second International Conference on Energy and Indoor Environment for Hot Climates, Doha, Qatar
Argyropoulos CD, Elkhalifa S, Fthenou E, Efthimiou GC, Andronopoulos S, Venetsanos A, Kovalets IV, Kakosimos KE (2018) Source reconstruction of airborne toxics based on acute health effects information. Sci Rep 8(1):5596. 10.1038/s41598-018-23767-8
Argyropoulos CD Hassan H Kumar P Kakosimos KE Measurements and modelling of particulate matter building ingress during a severe dust storm event Build Environ 2020 167 106441 10.1016/j.buildenv.2019.106441
Argyropoulos CD Markatos NC Recent advances on the numerical modelling of turbulent flows Appl Math Model 2015 39 2 693 732 10.1016/j.apm.2014.07.001
Asadi S Wexler AS Cappa CD Barreda S Bouvier NM Ristenpart WD Aerosol emission and superemission during human speech increase with voice loudness Sci Rep 2019 9 1 2348 10.1038/s41598-019-38808-z 30787335
ASHRAE (American Society of Heating Refrigerating and Air-Conditioning Engineers) (2017a) ANSI/ASHRAE/ASHE 170:2017 Ventilation of health care facilities. American Society of Heating Refrigerating and Air-Conditioning Engineers (ASHRAE) Inc., Atlanta
ASHRAE (American Society of Heating Refrigerating and Air-Conditioning Engineers) (2017b) Ventilation and infiltration. In American Society of Heating Refrigerating and Air-Conditioning Engineers (ed) ASHRAE Handbook of Fundamentals 2017. American Society of Heating Refrigerating and Air-Conditioning Engineers, Atlanta, pp 16.11–16.39
ASHRAE (American Society of Heating Refrigerating and Air-Conditioning Engineers) (2019a) ANSI/ASHRAE Standard 62.1:2019 Ventilation for acceptable indoor air quality. American Society of Heating Refrigerating and Air-Conditioning Engineers (ASHRAE) Inc., Atlanta
ASHRAE (American Society of Heating Refrigerating and Air-Conditioning Engineers) (2019b) ANSI/ASHRAE Standard 62.2:2019 Ventilation and acceptable indoor air quality in residential buildings. American Society of Heating Refrigerating and Air-Conditioning Engineers (ASHRAE) Inc., Atlanta
Ashraf AM, Argyropoulos CD, Olewski T, Vechot L, Kakosimos KE (2016) Comparative study on toxic gas infiltration in a nonprocess area using CFD and multi-zone models. Institution of Chemical Engineers. https://www.icheme.org/media/11816/hazards-26-poster-11-comparative-study-on-toxic-gas-infiltration-in-a-non-process-area-using-multizone-and-cfd-models.pdf. Accessed 24 Nov 2022
Axley J Multizone Airflow modeling in buildings: history and theory HVAC&R Res 2007 13 6 907 928 10.1080/10789669.2007.10391462
Axley JW Multi-zone dispersal analysis by element assembly Build Environ 1989 24 2 113 130 10.1016/0360-1323(89)90001-2
Azimi P Stephens B HVAC filtration for controlling infectious airborne disease transmission in indoor environments: predicting risk reductions and operational costs Build Environ 2013 70 150 160 10.1016/j.buildenv.2013.08.025 32288024
Azzopardi BJ Gas-liquid flows 2006 New York Begell House
Balachandar S Zaleski S Soldati A Ahmadi G Bourouiba L Host-to-host airborne transmission as a multiphase flow problem for science-based social distance guidelines Int J Multiph Flow 2020 132 103439 10.1016/j.ijmultiphaseflow.2020.103439
Bansal R Sharma D Singh R Tuberculosis and its treatment: an overview Mini Rev Med Chem 2018 18 1 58 71 10.2174/1389557516666160823160010 27553018
Barmby T Larguem M Coughs and sneezes spread diseases: an empirical study of absenteeism and infectious illness J Health Econ 2009 28 5 1012 1017 10.1016/j.jhealeco.2009.06.006 19619909
Bean B Moore BM Sterner B Peterson LR Gerding DN Balfour HH Jr Survival of influenza viruses on environmental surfaces J Infect Dis 1982 146 1 47 51 10.1093/infdis/146.1.47 6282993
Beggs CB The airborne transmission of infection in hospital buildings: fact or fiction? Indoor and Built Environment 2003 12 1–2 9 18 10.1177/1420326X03012001002
Beggs CB, Noakes CJ, Sleigh PA, Fletcher LA, Siddiqi K (2003) The transmission of tuberculosis in confined spaces: an analytical review of alternative epidemiological models. Int J Tuberc Lung Dis 7(11):1015–1026
Béghein C Jiang Y Chen QY Using large eddy simulation to study particle motions in a room Indoor Air 2005 15 4 281 290 10.1111/j.1600-0668.2005.00373.x 15982275
Ben-Tzvi P Rone W Microdroplet generation in gaseous and liquid environments Microsyst Technol 2010 16 3 333 356 10.1007/s00542-009-0962-7
Bergeron V Chalfine A Misset B Moules V Laudinet N Carlet J Lina B Supplemental treatment of air in airborne infection isolation rooms using high-throughput in-room air decontamination units Am J Infect Control 2011 39 4 314 320 10.1016/j.ajic.2010.06.013 21095042
Berlanga FA de Adana MR Olmedo I Villafruela JM San José JF Castro F Experimental evaluation of thermal comfort, ventilation performance indices and exposure to airborne contaminant in an airborne infection isolation room equipped with a displacement air distribution system Energy Build 2018 158 209 221 10.1016/j.enbuild.2017.09.100
Berrouk AS Lai ACK Cheung ACT Wong SL Experimental measurements and large eddy simulation of expiratory droplet dispersion in a mechanically ventilated enclosure with thermal effects Build Environ 2010 45 2 371 379 10.1016/j.buildenv.2009.06.016 32288007
Bhagat RK, Davies Wykes MS, Dalziel SB, Linden PF (2020) Effects of ventilation on the indoor spread of COVID-19. J Fluid Mech 903:F1–1-F1–18. 10.1017/jfm.2020.720
Biedermann T Winther L Till SJ Panzner P Knulst A Valovirta E Birch pollen allergy in Europe Allergy 2019 74 7 1237 1248 10.1111/all.13758 30829410
Bolashikov Z, Lu P, Malinowski T, Melikov A (2015) Air quality performance of ductless personalized ventilation in conjunction with displacement ventilation: Impact of walking person. Paper presented at the Healthy Buildings Europe 2015, HB 2015 - Conference Proceedings
Bolashikov ZD Melikov AK Methods for air cleaning and protection of building occupants from airborne pathogens Build Environ 2009 44 7 1378 1385 10.1016/j.buildenv.2008.09.001 32288004
Bolashikov ZD Melikov AK Kierat W Popioek Z Brand M Exposure of health care workers and occupants to coughed airborne pathogens in a double-bed hospital patient room with overhead mixing ventilation HVAC R Res 2012 18 4 602 615 10.1080/10789669.2012.682692
Bongers S, Janssen NAH, Reiss B, Grievink L, Lebret E, Kromhout H (2008) Challenges of exposure assessment for health studies in the aftermath of chemical incidents and disasters. J Expos Sci Environ Epidemiol 18(4):341–359. Retrieved from 10.1038/jes.2008.23
Bonnet M Lagier JC Raoult D Khelaifia S Bacterial culture through selective and non-selective conditions: the evolution of culture media in clinical microbiology New Microbes New Infect 2020 34 100622 10.1016/j.nmni.2019.100622 31956419
Booth TF, Kournikakis B, Bastien N, Ho J, Kobasa D, Stadnyk L, Li Y, Spence M, Paton S, Henry B, Mederski B, White D, Low DE, McGeer A, Simor A, Vearncombe M, Downey J, Jamieson FB, Tang P, Plummer F (2005) Detection of airborne severe acute respiratory syndrome (SARS) Coronavirus and environmental contamination in SARS outbreak units. J Infect Dis 191(9):1472–1477. 10.1086/429634
Borrego C Tchepel O Costa AM Martins H Ferreira J Miranda AI Traffic-related particulate air pollution exposure in urban areas Atmos Environ 2006 40 37 7205 7214 10.1016/j.atmosenv.2006.06.020
Borro L, Mazzei L, Raponi M, Piscitelli P, Miani A, Secinaro A (2020) The role of air conditioning in the diffusion of Sars-CoV-2 in indoor environments: a first computational fluid dynamic model, based on investigations performed at the Vatican State Children’s hospital. Environ Res 193:110343. 10.1016/j.envres.2020.110343
Bourouiba L Dehandschoewercker E Bush JWM Violent expiratory events: on coughing and sneezing J Fluid Mech 2014 745 537 563 10.1017/jfm.2014.88
Brager GS, De Dear RJ (1998) Thermal adaptation in the built environment: a literature review. Energy Build 27(1):83–96
Brennen EC (2005) Fundamentals of multiphase flows. Cambridge University Press, Cambridge
Breslow JM Meissler JJ Jr Hartzell RR Spence PB Truant A Gaughan J Eisenstein TK Innate immune responses to systemic Acinetobacter baumannii infection in mice: neutrophils, but not interleukin-17, mediate host resistance Infect Immun 2011 79 8 3317 3327 10.1128/iai.00069-11 21576323
Briggs DJ, Denman AR, Gulliver J, Marley RF, Kennedy CA, Philips PS, Field K, Crockett RM (2003) Time activity modelling of domestic exposures to radon. J Environ Manag 67(2):107–120. 10.1016/S0301-4797(02)00159-7
Broniarz-Press L Ochowiak M Rozanski J Woziwodzki S The atomization of water–oil emulsions Exp Thermal Fluid Sci 2009 33 6 955 962 10.1016/j.expthermflusci.2009.04.002
Brugger SD Baumberger C Jost M Jenni W Brugger U Mühlemann K Automated counting of bacterial colony forming units on agar plates PLoS One 2012 7 3 e33695 10.1371/journal.pone.0033695 22448267
Bruijns BB Tiggelaar RM Gardeniers H The extraction and recovery efficiency of pure DNA for different types of swabs J Forensic Sci 2018 63 5 1492 1499 10.1111/1556-4029.13837 29890011
Brundage JF Scott RM Lednar WM Smith DW Miller RN Building-associated risk of febrile acute respiratory diseases in army trainees JAMA 1988 259 14 2108 2112 10.1001/jama.1988.03720140028029 3346987
Burmølle M Johnsen K Abu Al-Soud W Hansen LH Sørensen SJ The presence of embedded bacterial pure cultures in agar plates stimulate the culturability of soil bacteria J Microbiol Methods 2009 79 2 166 173 10.1016/j.mimet.2009.08.006 19699242
Butler DF Myers AL Changing epidemiology of Haemophilus influenzae in children Infect Dis Clin North Am 2018 32 1 119 128 10.1016/j.idc.2017.10.005 29233576
Cao G Awbi H Yao R Fan Y Sirén K Kosonen R Zhang J A review of the performance of different ventilation and airflow distribution systems in buildings Build Environ 2014 73 171 186 10.1016/j.buildenv.2013.12.009
Cao G Nielsen PV Jensen RL Heiselberg P Liu L Heikkinen J Protected zone ventilation and reduced personal exposure to airborne cross-infection Indoor Air 2015 25 3 307 319 10.1111/ina.12142 24995998
Carducci A, Federigi I, Verani M (2020) Covid-19 Airborne transmission and its prevention: waiting for evidence or applying the precautionary principle? Atmosphere 11(7):710. 10.3390/atmos11070710
Carrera M Zandomeni RO Fitzgibbon J Sagripanti JL Difference between the spore sizes of Bacillus anthracis and other Bacillus species J Appl Microbiol 2007 102 2 303 312 10.1111/j.1365-2672.2006.03111.x 17241334
Causone F Baldin F Olesen BW Corgnati SP Floor heating and cooling combined with displacement ventilation: possibilities and limitations Energy Build 2010 42 12 2338 2352 10.1016/j.enbuild.2010.08.001
Causone F Olesen BW Corgnati SP Floor heating with displacement ventilation: an experimental and numerical analysis HVAC R Res 2010 16 2 139 160 10.1080/10789669.2010.10390898
CEN (Εuropean Committee for Standardization) (2019) EN 16798–1:2019 Energy performance of buildings. Ventilation for buildings. Indoor environmental input parameters for design and assessment of energy performance of buildings addressing indoor air quality, thermal environment, lighting and acoustics. European Committee for Standardization, Brussels
Cermak R Melikov AK Protection of occupants from exhaled infectious agents and floor material emissions in rooms with personalized and underfloor ventilation HVAC R Res 2007 13 1 23 38 10.1080/10789669.2007.10390942
Cermak R Melikov AK Forejt L Kovar O Performance of personalized ventilation in conjunction with mixing and displacement ventilation HVAC R Res 2006 12 2 295 311 10.1080/10789669.2006.10391180
Cetin YE Avci M Aydin O Influence of ventilation strategies on dispersion and removal of fine particles: an experimental and simulation study Sci Technol Built Environ 2020 26 3 349 365 10.1080/23744731.2019.1701332
Cetin YE Avci M Aydin O Particle dispersion and deposition in displacement ventilation systems combined with floor heating Sci Technol Built Environ 2020 10.1080/23744731.2020.1760637
Chafekar A Fielding BC MERS-CoV: understanding the latest human coronavirus threat Viruses 2018 10 2 22 10.3390/v10020093 29300333
Chan KS, Zheng JP, Mok YW, Li YM, Liu YN, Chu CM, Ip MS (2003) SARS: prognosis, outcome and sequelae. Respirology 8 Suppl(Suppl 1):S36–40. 10.1046/j.1440-1843.2003.00522.x
Chao CYH Tung TC An empirical model for outdoor contaminant transmission into residential buildings and experimental verification Atmos Environ 2001 35 9 1585 1596 10.1016/S1352-2310(00)00458-1
Chao CYH Wan MP A study of the dispersion of expiratory aerosols in unidirectional downward and ceiling-return type airflows using a multiphase approach Indoor Air 2006 16 4 296 312 10.1111/j.1600-0668.2006.00426.x 16842610
Chao CYH Wan MP Sze To GN Transport and removal of expiratory droplets in hospital ward environment Aerosol Sci Technol 2008 42 5 377 394 10.1080/02786820802104973
Chau OKY Liu CH Leung MKH CFD analysis of the performance of a local exhaust ventilation system in a hospital ward Indoor and Built Environment 2006 15 3 257 271 10.1177/1420326X06066123
Chen C Zhao B Some questions on dispersion of human exhaled droplets in ventilation room: answers from numerical investigation Indoor Air 2010 20 2 95 111 10.1111/j.1600-0668.2009.00626.x 20002792
Chen C Zhao B Yang X Li Y Role of two-way airflow owing to temperature difference in severe acute respiratory syndrome transmission: revisiting the largest nosocomial severe acute respiratory syndrome outbreak in Hong Kong J R Soc Interface 2011 8 58 699 710 10.1098/rsif.2010.0486 21068029
Chen C Zhu J Qu Z Lin CH Jiang Z Chen Q Systematic study of person-to-person contaminant transport in mechanically ventilated spaces (RP-1458) HVAC R Res 2014 20 1 80 91 10.1080/10789669.2013.834778
Cheng YH Wang CH You SH Hsieh NH Chen WY Chio CP Liao CM Assessing coughing-induced influenza droplet transmission and implications for infection risk control Epidemiol Infect 2016 144 2 333 345 10.1017/S0950268815001739 26211781
Choi J Kang M Jung JH Integrated micro-optofluidic platform for real-time detection of airborne microorganisms Sci Rep 2015 5 15983 10.1038/srep15983 26522006
Choi JI Edwards JR Large eddy simulation and zonal modeling of human-induced contaminant transport Indoor Air 2008 18 3 233 249 10.1111/j.1600-0668.2008.00527.x 18422965
Choi JI Edwards JR Large-eddy simulation of human-induced contaminant transport in room compartments Indoor Air 2012 22 1 77 87 10.1111/j.1600-0668.2011.00741.x 21906177
Choi N, Yamanaka T, Sagara K, Momoi Y, Suzuki T (2019) Displacement ventilation with radiant panel for hospital wards: measurement and prediction of the temperature and contaminant concentration profiles. Build Environ 160:106197. 10.1016/j.buildenv.2019.106197
Choudoir MJ Barberán A Menninger HL Dunn RR Fierer N Variation in range size and dispersal capabilities of microbial taxa Ecology 2018 99 2 322 334 10.1002/ecy.2094 29160898
Chow WK Application of computational fluid dynamics in building services engineering Build Environ 1996 31 5 425 436 10.1016/0360-1323(96)00012-1
Cole EC Cook CE Characterization of infectious aerosols in health care facilities: an aid to effective engineering controls and preventive strategies Am J Infect Control 1998 26 4 453 464 10.1016/S0196-6553(98)70046-X 9721404
Coleman KK Sigler WV Airborne influenza A virus exposure in an elementary school Sci Rep 2020 10 1 1859 10.1038/s41598-020-58588-1 32024882
Collier LH, Oxford JS, Pipkin J (2000) Human virology: a text for students of medicine, dentistry, and microbiology. Oxford University Press, Oxford
Collins CH, Lyne PM, Grange JM (1989) Collins and lyne’s microbiological methods. Butterworths, London
Cooke TF Indoor air pollutants a literature review Rev Environ Health 1991 9 3 137 160 10.1515/REVEH.1991.9.3.137 1792387
Crameri R Weichel M Flückiger S Glaser AG Rhyner C Fungal allergies: a yet unsolved problem Chem Immunol Allergy 2006 91 121 133 10.1159/000090276 16354954
Crowe CT Troutt TR Chung JN Numerical models for two-phase turbulent flows Annu Rev Fluid Mech 1996 28 11 43 10.1146/annurev.fl.28.010196.000303
Cui D Ai Z Mak CM Kwok K Xue P The influence of envelope features on interunit dispersion around a naturally ventilated multi-story building Build Simul 2018 11 6 1245 1253 10.1007/s12273-018-0460-x 32218905
Cutting SM Ricca E Bacterial spore-formers: friends and foes FEMS Microbiol Lett 2014 358 2 107 109 10.1111/1574-6968.12572 25168928
D’Amato G Liccardi G Frenguelli G Thunderstorm-asthma and pollen allergy Allergy 2007 62 1 11 16 10.1111/j.1398-9995.2006.01271.x 17156336
D’Amato G, Spieksma FT, Liccardi G, Jäger S, Russo M, Kontou-Fili K, Nikkels H, Wüthrich B, Bonini S (1998) Pollen-related allergy in Europe. Allergy 53(6):567–578. 10.1111/j.1398-9995.1998.tb03932.x
Dannemiller KC Weschler CJ Peccia J Fungal and bacterial growth in floor dust at elevated relative humidity levels Indoor Air 2017 27 2 354 363 10.1111/ina.12313 27272645
Davidson S Treating Influenza infection, from now and into the future Front Immunol 2018 9 1946 10.3389/fimmu.2018.01946 30250466
Dbouk T Drikakis D On coughing and airborne droplet transmission to humans Phys Fluids 2020 32 5 053310 10.1063/5.0011960
Dbouk T Drikakis D On respiratory droplets and face masks Phys Fluids 2020 32 6 063303 10.1063/5.0015044
Dbouk T Drikakis D Weather impact on airborne coronavirus survival Phys Fluids 2020 32 9 093312 10.1063/5.0024272
De Dear RJ Brager GS Thermal comfort in naturally ventilated buildings: revisions to ASHRAE standard 55 Energy Build 2002 34 6 549 561 10.1016/S0378-7788(02)00005-1
de Oliveira PM Mesquita LCC Gkantonas S Giusti A Mastorakos E Evolution of spray and aerosol from respiratory releases: theoretical estimates for insight on viral transmission Proc R Soc A Math Phys Eng Sci 2021 477 2245 20200584 10.1098/rspa.2020.0584
de Wit E van Doremalen N Falzarano D Munster VJ SARS and MERS: recent insights into emerging coronaviruses Nat Rev Microbiol 2016 14 8 523 534 10.1038/nrmicro.2016.81 27344959
Deacon JW (2006) Fungal biology. Blackwell Publishing, Malden
Dedesko S Siegel JA Moisture parameters and fungal communities associated with gypsum drywall in buildings Microbiome 2015 3 71 10.1186/s40168-015-0137-y 26642923
Dhama K, Khan S, Tiwari R, Sircar S, Bhat S, Malik YS, Singh KP, Chaicumpa W, Bonilla-Aldana K, Rodriguez-Morales AJ (2020) Coronavirus disease 2019-COVID-19. Clin Microbiol Revi 33(4):e00028–00020. 10.1128/CMR.00028-20
Dhand R Li J Coughs and sneezes: their role in transmission of respiratory viral infections, including SARS-CoV-2 Am J Respir Crit Care Med 2020 202 5 651 659 10.1164/rccm.202004-1263PP 32543913
Dietert K, Gutbier B, Wienhold SM, Reppe K, Jiang X, Yao L, Chaput C, Naujoks J, Brack M, Kupke A, Peteranderl C, Becker S, Lachner C, Baal N, Slevogt H, Hocke AC, Witzenrath M, Opitz B, Herold S, Hackstein H, Sander LE, Suttorp N, Gruber AD (2017) Spectrum of pathogen- and model-specific histopathologies in mouse models of acute pneumonia. Plos One 12(11):e0188251–e0188251. 10.1371/journal.pone.0188251
Dijksterhuis J Fungal spores: highly variable and stress-resistant vehicles for distribution and spoilage Food Microbiol 2019 81 2 11 10.1016/j.fm.2018.11.006 30910084
Dimitroulopoulou C Ashmore MR Byrne MA Kinnersley RP Modelling of indoor exposure to nitrogen dioxide in the UK Atmos Environ 2001 35 2 269 279 10.1016/S1352-2310(00)00176-X
Dimitroulopoulou C Ashmore MR Hill MTR Byrne MA Kinnersley R INDAIR: a probabilistic model of indoor air pollution in UK homes Atmos Environ 2006 40 33 6362 6379 10.1016/j.atmosenv.2006.05.047
Dingle TC Butler-Wu SM Maldi-tof mass spectrometry for microorganism identification Clin Lab Med 2013 33 3 589 609 10.1016/j.cll.2013.03.001 23931840
Diwan SS Ravichandran S Govindarajan R Narasimha R Understanding transmission dynamics of COVID-19-Type infections by direct numerical simulations of cough/sneeze flows Trans Indian Natl Acad Eng 2020 5 2 255 261 10.1007/s41403-020-00106-w
Dols WS, Polidoro BJ (2015) CONTAM user guide and program documentation. National Institute of Standards and Technology. https://nvlpubs.nist.gov/nistpubs/TechnicalNotes/NIST.TN.1887r1.pdf. Accessed 24 Nov 2022
Dols WS, Walton GN, Denton KR (2000) CONTAMW 1.0 user manual. National Institute of Standards and Technology, Gaithersburg
Domingo JL Marquès M Rovira J Influence of airborne transmission of SARS-CoV-2 on COVID-19 pandemic. A Review Environ Res 2020 188 109861 10.1016/j.envres.2020.109861 32718835
Douglas MG Kocher JF Scobey T Baric RS Cockrell AS Adaptive evolution influences the infectious dose of MERS-CoV necessary to achieve severe respiratory disease Virology 2018 517 98 107 10.1016/j.virol.2017.12.006 29277291
Drossinos Y Stilianakis NI What aerosol physics tells us about airborne pathogen transmission Aerosol Sci Technol 2020 54 6 639 643 10.1080/02786826.2020.1751055
Duan N Indoor Air Pollution Models for human exposure to air pollution Environ Int 1982 8 1 305 309 10.1016/0160-4120(82)90041-1
Dudalski N Mohamed A Mubareka S Bi R Zhang C Savory E Experimental investigation of far-field human cough airflows from healthy and influenza-infected subjects Indoor Air 2020 30 5 966 977 10.1111/ina.12680 32304605
Duquenne P On the identification of culturable microorganisms for the assessment of biodiversity in bioaerosols Ann Work Expo Health 2017 62 2 139 146 10.1093/annweh/wxx096
Efthimiou GC Kovalets IV Argyropoulos CD Venetsanos A Andronopoulos S Kakosimos KE Evaluation of an inverse modelling methodology for the prediction of a stationary point pollutant source in complex urban environments Build Environ 2018 143 107 119 10.1016/j.buildenv.2018.07.003
Emmerich SJ (2001) Validation of multizone IAQ modeling of residential-scale buildings: a review. Trans Am Soc Heat Refrigerating Air Cond Eng 107(2):619–628
Emmerich SJ Heinzerling D Choi J-I Persily AK Multizone modeling of strategies to reduce the spread of airborne infectious agents in healthcare facilities Build Environ 2013 60 105 115 10.1016/j.buildenv.2012.11.013
Engin AB Engin ED Engin A Two important controversial risk factors in SARS-CoV-2 infection: obesity and smoking Environ Toxicol Pharmacol 2020 78 103411 10.1016/j.etap.2020.103411 32422280
Escombe AR, Oeser CC, Gilman RH, Navincopa M, Ticona E, Pan W, Martínez C, Chacaltana J, Rodríguez R, Moore DAJ, Friedland JS, Evans CA (2007) Natural ventilation for the prevention of airborne contagion. PLoS Med 4(2):309–317. 10.1371/journal.pmed.0040068
Escombe AR, Ticona E, Chávez-Pérez V, Espinoza M, Moore DAJ (2019) Improving natural ventilation in hospital waiting and consulting rooms to reduce nosocomial tuberculosis transmission risk in a low resource setting. BMC Infect Dis 19(1). 10.1186/s12879-019-3717-9
Evers E, Waterhouse A (1978) A computer model for analysing smoke movement in Buildings. Building Research Establishment (BRE), Borehamwood
Faridi S, Niazi S, Sadeghi K, Naddafi K, Yavarian J, Shamsipour M, Jandaghi NZS, Sadeghniiat K, Nabizadeh R, Yunesian M, Momeniha F, Mokamel A, Hassanvand, MS, MokhtariAzad T (2020) A field indoor air measurement of SARS-CoV-2 in the patient rooms of the largest hospital in Iran. Sci Total Environ 725:138401. 10.1016/j.scitotenv.2020.138401
Faulkner WB Memarzadeh F Riskowski G Kalbasi A Ching-Zu Chang A Effects of air exchange rate, particle size and injection place on particle concentrations within a reduced-scale room Build Environ 2015 92 246 255 10.1016/j.buildenv.2015.04.034
Feldman C Anderson R The role of Streptococcus pneumoniae in Community-acquired pneumonia Semin Respir Crit Care Med 2016 37 6 806 818 10.1055/s-0036-1592074 27960205
Feng G Bi Y Zhang Y Cai Y Huang K Study on the motion law of aerosols produced by human respiration under the action of thermal plume of different intensities Sustain Cities Soc 2020 54 101935 10.1016/j.scs.2019.101935 32288992
Feng Y Marchal T Sperry T Yi H Influence of wind and relative humidity on the social distancing effectiveness to prevent COVID-19 airborne transmission: a numerical study J Aerosol Sci 2020 147 105585 10.1016/j.jaerosci.2020.105585 32427227
Feng Y Zhao J Hayati H Sperry T Yi H Tutorial: understanding the transport, deposition, and translocation of particles in human respiratory systems using computational fluid-particle dynamics and physiologically based toxicokinetic models J Aerosol Sci 2021 151 105672 10.1016/j.jaerosci.2020.105672
Fennelly KP Nardell EA The relative efficacy of respirators and room ventilation in preventing occupational tuberculosis Infect Control Hosp Epidemiol 1998 19 10 754 759 10.2307/30141420 9801283
Ferguson RMW Garcia-Alcega S Coulon F Dumbrell AJ Whitby C Colbeck I Bioaerosol biomonitoring: sampling optimization for molecular microbial ecology Mol Ecol Resour 2019 19 3 672 690 10.1111/1755-0998.13002 30735594
Fernstrom A Goldblatt M Aerobiology and its role in the transmission of infectious diseases Journal of Pathogens 2013 2013 493960 10.1155/2013/493960 23365758
Feustel HE COMIS—an international multizone air-flow and contaminant transport model Energy and Buildings 1999 30 1 3 18 10.1016/S0378-7788(98)00043-7
Feustel HE, Allard F, Dorer VB, Grosso M, Herrlin M, Mingsheng L, Phaff JC, Utsumi Y, Yoshino H (1989) The COMIS infiltration model. U.S. Department of Energy. https://www.osti.gov/servlets/purl/7259831. Accessed 24 Nov 2022
Feustel HE Dieris J A survey of airflow models for multizone structures Energy Build 1992 18 2 79 100 10.1016/0378-7788(92)90040-N
Fontes D Reyes J Ahmed K Kinzel M A study of fluid dynamics and human physiology factors driving droplet dispersion from a human sneeze Phys Fluids 2020 32 11 111904 10.1063/5.0032006
Foster A Kinzel M Estimating COVID-19 exposure in a classroom setting: a comparison between mathematical and numerical models Phys Fluids 2021 33 2 021904 10.1063/5.0040755
Foster KN Chundu KR Lal S Caruso DM Invasive aspergillus infection leading to vascular thrombosis and amputation in a severely burned child J Burn Care Res 2017 38 1 e464 e468 10.1097/bcr.0000000000000366 27355652
Friberg B Friberg S Burman LG Lundholm R Östensson R Inefficiency of upward displacement operating theatre ventilation J Hosp Infect 1996 33 4 263 272 10.1016/S0195-6701(96)90012-2 8999051
Gadsby NJ McHugh MP Forbes C MacKenzie L Hamilton SKD Griffith DM Templeton KE Comparison of Unyvero P55 Pneumonia Cartridge, in-house PCR and culture for the identification of respiratory pathogens and antibiotic resistance in bronchoalveolar lavage fluids in the critical care setting Eur J Clin Microbiol Infect Dis 2019 38 6 1171 1178 10.1007/s10096-019-03526-x 30859358
Galvin CJ Li YCJ Malwade S Syed-Abdul S COVID-19 preventive measures showing an unintended decline in infectious diseases in Taiwan Int J Infect Dis 2020 98 18 20 10.1016/j.ijid.2020.06.062 32585283
Gama JA Abby SS Vieira-Silva S Dionisio F Rocha EP Immune subversion and quorum-sensing shape the variation in infectious dose among bacterial pathogens PLoS Pathog 2012 8 2 e1002503 10.1371/journal.ppat.1002503 22319444
Gao N He Q Niu J Numerical study of the lock-up phenomenon of human exhaled droplets under a displacement ventilated room Build Simul 2012 5 1 51 60 10.1007/s12273-012-0068-5
Gao N Niu J Morawska L Distribution of respiratory droplets in enclosed environments under different air distribution methods Build Simul 2008 1 4 326 335 10.1007/s12273-008-8328-0
Gao NP Niu JL CFD Study of the thermal environment around a human body: a review Indoor Built Environ 2005 14 1 5 16 10.1177/1420326X05050132
Gao NP Niu JL Perino M Heiselberg P The airborne transmission of infection between flats in high-rise residential buildings: tracer gas simulation Build Environ 2008 43 11 1805 1817 10.1016/j.buildenv.2007.10.023 32288000
Gao NP Niu JL Perino M Heiselberg P The airborne transmission of infection between flats in high-rise residential buildings: particle simulation Build Environ 2009 44 2 402 410 10.1016/j.buildenv.2008.03.016 32288003
Gao X Li Y Leung GM Ventilation control of indoor transmission of airborne diseases in an Urban community Indoor Built Environ 2009 18 3 205 218 10.1177/1420326X09104141
Gao X Li Y Xu P Cowling BJ Evaluation of intervention strategies in schools including ventilation for influenza transmission control Build Simul 2012 5 1 29 37 10.1007/s12273-011-0034-7
Gao X Wei J Cowling BJ Li Y Potential impact of a ventilation intervention for influenza in the context of a dense indoor contact network in Hong Kong Sci Total Environ 2016 569–570 373 381 10.1016/j.scitotenv.2016.06.179
Gerharz LE Krüger A Klemm O Applying indoor and outdoor modeling techniques to estimate individual exposure to PM2.5 from personal GPS profiles and diaries: a pilot study Sci Total Environ 2009 407 18 5184 5193 10.1016/j.scitotenv.2009.06.006 19577794
Ghosh B Lal H Srivastava A Review of bioaerosols in indoor environment with special reference to sampling, analysis and control mechanisms Environ Int 2015 85 254 272 10.1016/j.envint.2015.09.018 26436919
Gilkeson CA Camargo-Valero MA Pickin LE Noakes CJ Measurement of ventilation and airborne infection risk in large naturally ventilated hospital wards Build Environ 2013 65 35 48 10.1016/j.buildenv.2013.03.006
Glanville N Johnston SL Challenges in developing a cross-serotype rhinovirus vaccine Curr Opin Virol 2015 11 83 88 10.1016/j.coviro.2015.03.004 25829255
Godri Pollitt KJ, Peccia J, Ko AI, Kaminski N, Dela Cruz CS, Nebert DW, Reichardt JKV, Thompson DC, Vasiliou V (2020) COVID-19 vulnerability: the potential impact of genetic susceptibility and airborne transmission. Hum Genomics 14(1):17. 10.1186/s40246-020-00267-3
Gralton J Tovey E McLaws M-L Rawlinson WD The role of particle size in aerosolised pathogen transmission: a review J Infect 2011 62 1 1 13 10.1016/j.jinf.2010.11.010 21094184
Green BJ Tovey ER Sercombe JK Blachere FM Beezhold DH Schmechel D Airborne fungal fragments and allergenicity Med Mycol 2006 44 Suppl 1 S245 255 10.1080/13693780600776308 17050446
Griffin DW Atmospheric Movement of microorganisms in clouds of desert dust and implications for human health Clin Microbiol Rev 2007 20 3 459 10.1128/CMR.00039-06 17630335
Grishkan I Thermotolerant mycobiota of Israeli soils J Basic Microbiol 2018 58 1 30 40 10.1002/jobm.201700517 29144537
Gu D, Zheng Z, Zhao P, Xie L, Xu Z, Lu X (2020) High-Efficiency simulation framework to analyze the impact of exhaust air from COVID-19 Temporary hospitals and its typical applications. Appl Sci 10(11):3949. 10.3390/app10113949
Gu W Miller S Chiu CY Clinical metagenomic next-generation sequencing for pathogen detection Annu Rev Pathol 2019 14 319 338 10.1146/annurev-pathmechdis-012418-012751 30355154
Guzman M (2020) Bioaerosol size effect in COVID-19 transmission. Preprints org. https://www.preprints.org/manuscript/202004.0093/v2. Accessed 24 Nov 2022
Habchi C Ghali K Ghaddar N Chakroun W Alotaibi S Ceiling personalized ventilation combined with desk fans for reduced direct and indirect cross-contamination and efficient use of office space Energy Convers Manag 2016 111 158 173 10.1016/j.enconman.2015.12.067
Hageman JR The Coronavirus Disease 2019 (COVID-19) Pediatr Ann 2020 49 3 e99 e100 10.3928/19382359-20200219-01 32155273
Haghighat F Megri AC A comprehensive validation of two airflow models - COMIS and CONTAM Indoor Air 1996 6 278 288 10.1111/j.1600-0668.1996.00007.x
Haghighat F Rao J Computer-aided building ventilation system design — a system-theoretic approach Energy Build 1991 17 2 147 155 10.1016/0378-7788(91)90007-P
Haghnegahdar A Zhao J Feng Y Lung aerosol dynamics of airborne influenza A virus-laden droplets and the resultant immune system responses: an in silico study J Aerosol Sci 2019 134 34 55 10.1016/j.jaerosci.2019.04.009 31983771
Haghnegahdar A Zhao J Kozak M Williamson P Feng Y Development of a hybrid CFD-PBPK model to predict the transport of xenon gas around a human respiratory system to systemic regions Heliyon 2019 5 4 e01461 10.1016/j.heliyon.2019.e01461 31011641
Haines SR, Adams RI, Boor BE, Bruton TA, Downey J, Ferro AR, Gall E,Green BJ, Hegarty B, Horner E, Jacobs DE, Lemieux P, Misztal PK, Morrison G, Perzanowski M, Reponen T, Rush RE, Virgo T, Alkhayri C, Bope A, Cochran S, Cox J, Donohue A, May AA, Nastasi N, Nishioka M, Renninger N, Tian Y, Uebel-Niemeier C, Wilkinson D, Wu T, Zambrana J, Dannemiller KC (2020) Ten questions concerning the implications of carpet on indoor chemistry and microbiology. Build Environ 170:106589. 10.1016/j.buildenv.2019.106589
Hajjar SA Memish ZA McIntosh K Middle East Respiratory syndrome coronavirus (MERS-CoV): a perpetual challenge Ann Saudi Med 2013 33 5 427 436 10.5144/0256-4947.2013.427 24188935
Hall-Stoodley L Costerton JW Stoodley P Bacterial biofilms: from the natural environment to infectious diseases Nat Rev Microbiol 2004 2 2 95 108 10.1038/nrmicro821 15040259
Han B Weiss LM Microsporidia: obligate intracellular pathogens within the fungal kingdom Microbiol Spectr 2017 5 2 1 26 10.1128/microbiolspec.FUNK-0018-2016
Han H Wang P Li Y Liu R Tian C Effect of water supply pressure on atomization characteristics and dust-reduction efficiency of internal mixing air atomizing nozzle Adv Powder Technol 2020 31 1 252 268 10.1016/j.apt.2019.10.017
Haun N Hooper-Lane C Safdar N Healthcare personnel attire and devices as fomites: a systematic review Infect Control Hosp Epidemiol 2016 37 11 1367 1373 10.1017/ice.2016.192 27609491
He Q Niu J Gao N Zhu T Wu J CFD study of exhaled droplet transmission between occupants under different ventilation strategies in a typical office room Build Environ 2011 46 2 397 408 10.1016/j.buildenv.2010.08.003 32288015
Henrickson KJ Parainfluenza viruses Clin Microbiol Rev 2003 16 2 242 264 10.1128/cmr.16.2.242-264.2003 12692097
Hicklin W Farrugia PS Sinagra E Investigations of VOCs in and around buildings close to service stations Atmos Environ 2018 172 93 101 10.1016/j.atmosenv.2017.10.012
Hirose R, Ikegaya H, Naito Y, Watanabe N, Yoshida T, Bandou R, Daidoji T, Itoh Y, Nakaya T (2020) Survival of SARS-CoV-2 and influenza virus on the human skin: importance of hand hygiene in COVID-19. Clin Infect Dis. 10.1093/cid/ciaa1517
Hobday RA Dancer SJ Roles of sunlight and natural ventilation for controlling infection: historical and current perspectives J Hosp Infect 2013 84 4 271 282 10.1016/j.jhin.2013.04.011 23790506
Huang J Jiang E Yang D Wei J Zhao M Feng J Cao J Metagenomic next-generation sequencing versus traditional pathogen detection in the diagnosis of peripheral pulmonary infectious lesions Infect Drug Resist 2020 13 567 576 10.2147/idr.s235182 32110067
Huang J-M Tsao S-M The Influence of air motion on bacteria removal in negative pressure isolation rooms HVAC&R Res 2005 11 4 563 585 10.1080/10789669.2005.10391155
Huffman JA, Perring AE, Savage NJ, Clot B, Crouzy B, Tummon F, Shoshanim O, Damit B, Schneider J, Sivaprakasam V, Zawadowicz MA, Crawford I, Gallagher M, Topping D, Doughty DC, Hill SC, Pan Y (2020) Real-time sensing of bioaerosols: review and current perspectives. Aerosol Sci Technol 54(5):465–495. 10.1080/02786826.2019.1664724
Hunter RL Tuberculosis as a three-act play: a new paradigm for the pathogenesis of pulmonary tuberculosis Tuberculosis (Edinb) 2016 97 8 17 10.1016/j.tube.2015.11.010 26980490
Hunter RL The Pathogenesis of tuberculosis: the early infiltrate of post-primary (adult pulmonary) tuberculosis: a distinct disease entity Front Immunol 2018 9 2108 10.3389/fimmu.2018.02108 30283448
Husman T (1996) Health effects of indoor-air microorganisms. Scand J Work Environ Health 22(1):5–13. 10.5271/sjweh.103
Hussain S Parker S Edwards K Finch J Jeanjean A Leigh R Gonem S Effects of indoor particulate matter exposure on daily asthma control Ann Allergy Asthma Immunol 2019 123 4 375 380.e373 10.1016/j.anai.2019.07.020 31351980
Islam MS, Paul G, Ong HX, Young PM, Gu YT, Saha SC (2020) A review of respiratory anatomical development, air flow characterization and particle deposition. Int J Environ Res Public Health 17(2):380. 10.3390/ijerph17020380
Ito K Integrated numerical approach of computational fluid dynamics and epidemiological model for multi-scale transmission analysis in indoor spaces Indoor Built Environ 2014 23 7 1029 1049 10.1177/1420326X13516658
Iwasaki A Pillai PS Innate immunity to influenza virus infection Nat Rev Immunol 2014 14 5 315 328 10.1038/nri3665 24762827
Jang KS Kim YH Rapid and robust MALDI-TOF MS techniques for microbial identification: a brief overview of their diverse applications J Microbiol 2018 56 4 209 216 10.1007/s12275-018-7457-0 29492868
Järvinen AK, Laakso S, Piiparinen P, Aittakorpi A, Lindfors M, Huopaniemi L, Piiparinen H, Mäki M (2009) Rapid identification of bacterial pathogens using a PCR- and microarray-based assay. BMC Microbiol 9:161. 10.1186/1471-2180-9-161
Jayaweera M Perera H Gunawardana B Manatunge J Transmission of COVID-19 virus by droplets and aerosols: A CRITICAL review on the unresolved dichotomy Environ Res 2020 188 109819 10.1016/j.envres.2020.109819 32569870
Ji Y Qian H Ye J Zheng X The impact of ambient humidity on the evaporation and dispersion of exhaled breathing droplets: a numerical investigation J Aerosol Sci 2018 115 164 172 10.1016/j.jaerosci.2017.10.009
Jiang Y Zhao B Li X Yang X Zhang Z Zhang Y Investigating a safe ventilation rate for the prevention of indoor SARS transmission: an attempt based on a simulation approach Build Simul 2009 2 4 281 289 10.1007/s12273-009-9325-7 32218907
Jones AP Indoor air quality and health Atmos Environ 1999 33 28 4535 4564 10.1016/S1352-2310(99)00272-1
Jurelionis A Gagyte L Prasauskas T Čiužas D Krugly E Šeduikyte L Martuzevičius D The impact of the air distribution method in ventilated rooms on the aerosol particle dispersion and removal: the experimental approach Energy Build 2015 86 305 313 10.1016/j.enbuild.2014.10.014
Jurelionis A Gagyte L Seduikyte L Prasauskas T Ciuzas D Martuzevicius D Combined air heating and ventilation increases risk of personal exposure to airborne pollutants released at the floor level Energy Build 2016 116 263 273 10.1016/j.enbuild.2016.01.011
Jurelionis A Stasiuliene L Prasauskas T Martuzevicius D Dispersion of indoor air pollutants emitted at near-floor levels in rooms with floor heating and mixing ventilation Indoor Built Environ 2018 27 2 205 218 10.1177/1420326X16669975
Kampf G Todt D Pfaender S Steinmann E Persistence of coronaviruses on inanimate surfaces and their inactivation with biocidal agents J Hosp Infect 2020 104 3 246 251 10.1016/j.jhin.2020.01.022 32035997
Kang Z Zhang Y Fan H Feng G Numerical Simulation of coughed droplets in the air-conditioning room Procedia Eng 2015 121 114 121 10.1016/j.proeng.2015.08.1031
Karakitsios S, Busker R, Tjärnhage T, Armand P, Dybwad M, Nielsen MF, Burman J, Burke J, Brinek J, Bartzis J, Maggos T, Theocharidou M, Gattinesi P, Giannopoulos G, Sarigiannis D (2020) Challenges on detection, identification and monitoring of indoor airborne chemical-biological agents. Saf Sci 129:104789. 10.1016/j.ssci.2020.104789
Katramiz E Al Assaad D Ghaddar N Ghali K The effect of human breathing on the effectiveness of intermittent personalized ventilation coupled with mixing ventilation Build Environ 2020 174 106755 10.1016/j.buildenv.2020.106755
Katz SM Hammel JM The effect of drying, heat, and pH on the survival of Legionella pneumophila Ann Clin Lab Sci 1987 17 3 150 156 3606021
Keshavarz SA Salmanzadeh M Ahmadi G Computational modeling of time resolved exposure level analysis of a heated breathing manikin with rotation in a room J Aerosol Sci 2017 103 117 131 10.1016/j.jaerosci.2016.09.005
Khan RK Strand MA Road dust and its effect on human health: a literature review Epidemiol Health 2018 40 e2018013 10.4178/epih.e2018013 29642653
Khosronejad A Santoni C Flora K Zhang Z Kang S Payabvash S Sotiropoulos F Fluid dynamics simulations show that facial masks can suppress the spread of COVID-19 in indoor environments AIP Adv 2020 10 12 125109 10.1063/5.0035414
Kim H Webster RG Webby RJ Influenza Virus: dealing with a drifting and shifting pathogen Viral Immunol 2018 31 2 174 183 10.1089/vim.2017.0141 29373086
King MF Noakes CJ Sleigh PA Modeling environmental contamination in hospital single- and four-bed rooms Indoor Air 2015 25 6 694 707 10.1111/ina.12186 25614923
King N Auger P Indoor air quality, fungi, and health. How do we stand? Can Fam Physician 2002 48 298 302 11889889
Knibbs LD, Johnson GR, Kidd TJ, Cheney J, Grimwood K, Kattenbelt JA, O'Rourke P, Ramsay KA, Sly PD, Wainwright CE, Wood ME, Morawska L, Bell SC (2014) Viability of Pseudomonas aeruginosa in cough aerosols generated by persons with cystic fibrosis. Thorax 69(8):740–745. 10.1136/thoraxjnl-2014-205213
Knibbs LD Morawska L Bell SC Grzybowski P Room ventilation and the risk of airborne infection transmission in 3 health care settings within a large teaching hospital Am J Infect Control 2011 39 10 866 872 10.1016/j.ajic.2011.02.014 21658810
Kornartit C Sokhi RS Burton MA Ravindra K Activity pattern and personal exposure to nitrogen dioxide in indoor and outdoor microenvironments Environ Int 2010 36 1 36 45 10.1016/j.envint.2009.09.004 19878999
Kovalets IV Efthimiou GC Andronopoulos S Venetsanos AG Argyropoulos CD Kakosimos KE Inverse identification of unknown finite-duration air pollutant release from a point source in urban environment Atmos Environ 2018 181 82 96 10.1016/j.atmosenv.2018.03.028
Kowalski W Bahnfleth W Musser A Modeling immune building systems for bioterrorism defense J Archit Eng 2003 9 2 86 96 10.1061/(ASCE)1076-0431(2003)9:2(86)
Kralik P Ricchi M A Basic guide to real time PCR in microbial diagnostics: definitions, parameters, and everything Front Microbiol 2017 8 108 10.3389/fmicb.2017.00108 28210243
Kramer A Schwebke I Kampf G How long do nosocomial pathogens persist on inanimate surfaces? A systematic review BMC Infect Dis 2006 6 130 130 10.1186/1471-2334-6-130 16914034
Kumar V Influenza in children Indian J Pediatr 2017 84 2 139 143 10.1007/s12098-016-2232-x 27641976
Kurgat EK, Sexton JD, Garavito F, Reynolds A, Contreras RD, Gerba CP, Leslie RA, Edmonds-Wilson SL, Reynolds KA (2019) Impact of a hygiene intervention on virus spread in an office building. Int J Hyg Environ Health 222(3):479–485. 10.1016/j.ijheh.2019.01.001
Kuznetsov GV Shlegel NE Solomatin Y Strizhak PA Combined techniques of secondary atomization of multi-component droplets Chem Eng Sci 2019 209 115199 10.1016/j.ces.2019.115199
Lacoma A, Mateo L, Blanco I, Méndez MJ, Rodrigo C, Latorre I, Villar-Hernandez R, Domínguez J, Prat C (2019) Impact of host genetics and biological response modifiers on respiratory tract infections. Front Immunol 10:1013. 10.3389/fimmu.2019.01013
Lai ACK Cheng YC Study of expiratory droplet dispersion and transport using a new Eulerian modeling approach Atmos Environ 2007 41 35 7473 7484 10.1016/j.atmosenv.2007.05.045
Lai ACK Wong SL Expiratory aerosol transport in a scaled chamber under a variety of emission characteristics: an experimental study Aerosol Sci Technol 2011 45 8 909 917 10.1080/02786826.2011.571308
Lai MYY Cheng PKC Lim WWL Survival of severe acute respiratory syndrome coronavirus Clin Infect Dis 2005 41 7 67 71 10.1086/433186 15937765
Lanzerstorfer C Variations in the composition of house dust by particle size J Environ Sci Health A Tox Hazard Subst Environ Eng 2017 52 8 770 777 10.1080/10934529.2017.1303316 28394695
Latgé JP, Chamilos G (2019) Aspergillus fumigatus and Aspergillosis in 2019. Clin Microbiol Rev 33(1):e00140-18. 10.1128/cmr.00140-18
Lauxmann MA Santucci NE Autrán-Gómez AM The SARS-CoV-2 coronavirus and the COVID-19 outbreak Int Braz J Urol 2020 46 1 6 18 10.1590/s1677-5538.ibju.2020.s101 32549071
Lee C Park S Cho K Yoo JE Lee S Ko G Comparison of swab sampling methods for norovirus recovery on surfaces Food Environ Virol 2018 10 4 378 385 10.1007/s12560-018-9353-5 30094754
Lee K Parkhurst WJ Xue J Ozkaynak AH Neuberg D Spengler JD Outdoor/indoor/personal ozone exposures of children in Nashville, Tennessee J Air Waste Manag Assoc 2004 54 3 352 359 10.1080/10473289.2004.10470904 15061616
Leggett MJ McDonnell G Denyer SP Setlow P Maillard JY Bacterial spore structures and their protective role in biocide resistance J Appl Microbiol 2012 113 3 485 498 10.1111/j.1365-2672.2012.05336.x 22574673
Lei Z Yang J Zhuang Z Headform and N95 filtering facepiece respirator interaction: contact pressure simulation and validation J Occup Environ Hyg 2012 9 1 46 58 10.1080/15459624.2011.635130 22168255
Lei Z Yang J Zhuang Z Roberge R Simulation and evaluation of respirator faceseal leaks using computational fluid dynamics and infrared imaging Ann Occup Hyg 2012 57 4 493 506 10.1093/annhyg/mes085 23243192
Leskiewicz M Kaliszewski M Włodarski M Młynczak J Mierczyk Z Kopczynski K Improved real-time bio-aerosol classification using artificial neural networks Atmos Meas Tech 2018 11 6259 6270 10.5194/amt-11-6259-2018
Li C-C Aerodynamic behavior of a gas mask canister containing two porous media Chem Eng Sci 2009 64 8 1832 1843 10.1016/j.ces.2009.01.009
Li F Niu J Control of volatile organic compounds indoors—development of an integrated mass-transfer-based model and its application Atmos Environ 2007 41 11 2344 2354 10.1016/j.atmosenv.2006.11.022
Li X Niu J Gao N Co-occupant’s exposure to exhaled pollutants with two types of personalized ventilation strategies under mixing and displacement ventilation systems Indoor Air 2013 23 2 162 171 10.1111/ina.12005 23002790
Li X Shang Y Yan Y Yang L Tu J Modelling of evaporation of cough droplets in inhomogeneous humidity fields using the multi-component Eulerian-Lagrangian approach Build Environ 2018 128 68 76 10.1016/j.buildenv.2017.11.025 32287977
Li X-P Niu J-L Gao N-P Characteristics of physical blocking on co-occupant’s exposure to respiratory droplet residuals J Cent South Univ 2012 19 3 645 650 10.1007/s11771-012-1051-0 32288947
Li Y Duan S Yu ITS Wong TW Multi-zone modeling of probable SARS virus transmission by airflow between flats in Block E, Amoy Gardens Indoor Air 2004 15 2 96 111 10.1111/j.1600-0668.2004.00318.x
Li Y Duan S Yu ITS Wong TW Multi-zone modeling of probable SARS virus transmission by airflow between flats in Block E, Amoy Gardens Indoor Air 2005 15 2 96 111 10.1111/j.1600-0668.2004.00318.x 15737152
Li Y Huang X Yu ITS Wong TW Qian H Role of air distribution in SARS transmission during the largest nosocomial outbreak in Hong Kong Indoor Air 2004 15 2 83 95 10.1111/j.1600-0668.2004.00317.x
Li Y, Leung GM, Tang JW, Yang X, Chao CYH, Lin JZ, Lu JW, Nielsen PV, Niu J, Qian H, Sleigh AC, Su H-JJ, Sundell J, Wong TW,Yuen PL (2007) Role of ventilation in airborne transmission of infectious agents in the built environment – a multidisciplinary systematic review. Indoor Air 17(1):2–18. 10.1111/j.1600-0668.2006.00445.x
Xiaoping L Jianlei N Naiping G Spatial distribution of human respiratory droplet residuals and exposure risk for the co-occupant under different ventilation methods HVAC R Res 2011 17 4 432 445 10.1080/10789669.2011.578699
Li Z Wang H Zhang X Wu T Yang X Effects of space sizes on the dispersion of cough-generated droplets from a walking person Phys Fluids 2020 32 12 121705 10.1063/5.0034874
Licina D Melikov A Pantelic J Sekhar C Tham KW Human convection flow in spaces with and without ventilation: personal exposure to floor-released particles and cough-released droplets Indoor Air 2015 25 6 672 682 10.1111/ina.12177 25515610
Lim T Cho J Kim BS The predictions of infection risk of indoor airborne transmission of diseases in high-rise hospitals: tracer gas simulation Energy Build 2010 42 8 1172 1181 10.1016/j.enbuild.2010.02.008
Lim T Cho J Kim BS Predictions and measurements of the stack effect on indoor airborne virus transmission in a high-rise hospital building Build Environ 2011 46 12 2413 2424 10.1016/j.buildenv.2011.04.015 32288013
Lin Y Wang BX Zhang NN Zhang L Gao ZB Tian J Jiang X Metagenomic analysis identified Stenotrophomonas maltophilia Pneumonia in an infant suffering from unexplained very severe pneumonia Front Pediatr 2019 7 380 10.3389/fped.2019.00380 31608265
Lin Z Wang J Yao T Chow TT Investigation into anti-airborne infection performance of stratum ventilation Build Environ 2012 54 29 38 10.1016/j.buildenv.2012.01.017
Lin Z Wang J Yao T Chow TT Fong KF Numerical comparison of dispersion of human exhaled droplets under different ventilation methods World Rev Sci Technol Sustain Dev 2013 10 1–3 142 161 10.1504/WRSTSD.2013.050790
Lindsley WG King WP Thewlis RE Reynolds JS Panday K Cao G Szalajda JV Dispersion and exposure to a cough-generated aerosol in a simulated medical examination room J Occup Environ Hyg 2012 9 12 681 690 10.1080/15459624.2012.725986 23033849
Lipczynska A Kaczmarczyk J Melikov AK Thermal environment and air quality in office with personalized ventilation combined with chilled ceiling Build Environ 2015 92 603 614 10.1016/j.buildenv.2015.05.035
Liu F Qian H Zheng X Song J Cao G Liu Z Evaporation and dispersion of exhaled droplets in stratified environment IOP Conf Ser Mater Sci Eng 2019 609 042059 10.1088/1757-899x/609/4/042059
Liu HY Hopping GC Vaidyanathan U Ronquillo YC Hoopes PC Moshirfar M Polymerase chain reaction and its application in the diagnosis of infectious keratitis Med Hypothesis Discov Innov Ophthalmol 2019 8 3 152 155 31598517
Liu J Dalgo DA Zhu S Li H Zhang L Srebric J Performance analysis of a ductless personalized ventilation combined with radiant floor cooling system and displacement ventilation Build Simul 2019 12 5 905 919 10.1007/s12273-019-0521-9
Liu J, Zhu S, Kim MK, Srebric J (2019d) A review of CFD analysis methods for personalized ventilation (PV) in indoor built environments. Sustainability 11(15):4166. 10.3390/su11154166
Liu L Li Y Nielsen PV Wei J Jensen RL Short-range airborne transmission of expiratory droplets between two people Indoor Air 2016 27 2 452 462 10.1111/ina.12314 27287598
Liu L Wei J Li Y Ooi A Evaporation and dispersion of respiratory droplets from coughing Indoor Air 2017 27 1 179 190 10.1111/ina.12297 26945674
Liu W You X-Y Transportation and risk analysis of influenza indoor and outdoor transportation and exposure risk analysis of influenza aerosol Indoor Built Environ 2011 21 5 614 621 10.1177/1420326X11426164
Liu X Niu J Wind tunnel test of indoor air pollutant dispersion around high-rise building ASHRAE IAQ 2011 2010 1 6
Liu X, Niu J, Gao N, Perino M, Heiselberg P (2007) Cfd simulation of inter-flat air cross-contamination-A possible transmission path of infectious diseases. Air infiltration and Ventilation Centre. https://www.aivc.org/sites/default/files/p646_final.pdf. Accessed 24 Nov 2022
Liu X Niu J Perino M Heiselberg P Numerical simulation of inter-flat air cross-contamination under the condition of single-sided natural ventilation J Build Perform Simul 2008 1 2 133 147 10.1080/19401490802250462
Liu X Zhang X Min J Spreading of droplets impacting different wettable surfaces at a Weber number close to zero Chem Eng Sci 2019 207 495 503 10.1016/j.ces.2019.06.058
Liu XP Niu JL Kwok KCS Analysis of concentration fluctuations in gas dispersion around high-rise building for different incident wind directions J Hazard Mater 2011 192 3 1623 1632 10.1016/j.jhazmat.2011.06.090 21784582
Liu XP Niu JL Kwok KCS Wang JH Li BZ Investigation of indoor air pollutant dispersion and cross-contamination around a typical high-rise residential building: Wind tunnel tests Build Environ 2010 45 8 1769 1778 10.1016/j.buildenv.2010.02.003
Liu XP Niu JL Kwok KCS Wang JH Li BZ Local characteristics of cross-unit contamination around high-rise building due to wind effect: mean concentration and infection risk assessment J Hazard Mater 2011 192 1 160 167 10.1016/j.jhazmat.2011.04.106 21616594
Liu Y Zhao Y Liu Z Luo J Numerical investigation of the unsteady flow characteristics of human body thermal plume Build Simul 2016 9 6 677 687 10.1007/s12273-016-0296-1
Löhner R, Antil H (2020) High fidelity modeling of aerosol pathogen propagation in built environments with moving pedestrians. Int J Numer Meth in Biomed Eng 37:e3428. 10.1002/cnm.3428
Löhner R Antil H Idelsohn S Oñate E Detailed simulation of viral propagation in the built environment Comput Mech 2020 66 5 1093 1107 10.1007/s00466-020-01881-7 32836601
Loures FV, Röhm M, Lee CK, Santos E, Wang JP, Specht CA, Calich VLG, Urban CF, Levitz SM (2015) Recognition of Aspergillus fumigatus hyphae by human plasmacytoid dendritic cells is mediated by dectin-2 and results in formation of extracellular traps. PLoS Pathog 11(2):e1004643–e1004643. 10.1371/journal.ppat.1004643
Luengas A Barona A Hort C Gallastegui G Platel V Elias A A review of indoor air treatment technologies Rev Environ Sci Biotechnol 2015 14 3 499 522 10.1007/s11157-015-9363-9
Luksamijarulkul P Pipitsangjan S Microbial air quality and bacterial surface contamination in ambulances during patient services Oman Med J 2015 30 2 104 110 10.5001/omj.2015.23 25960835
Lunden MM Revzan KL Fischer ML Thatcher TL Littlejohn D Hering SV Brown NJ The transformation of outdoor ammonium nitrate aerosols in the indoor environment Atmos Environ 2003 37 39 5633 5644 10.1016/j.atmosenv.2003.09.035
Lunden MM Thatcher TL Hering SV Brown NJ Use of Time- and chemically resolved particulate data to characterize the infiltration of outdoor PM2.5 into a residence in the San Joaquin valley Environ Sci Technol 2003 37 4724 4732 10.1021/es026387i 14594384
Luongo JC Fennelly KP Keen JA Zhai ZJ Jones BW Miller SL Role of mechanical ventilation in the airborne transmission of infectious agents in buildings Indoor Air 2016 26 5 666 678 10.1111/ina.12267 26562748
Lv Y Wang H Wei S The transmission characteristics of indoor particles under two ventilation modes Energy Build 2018 163 1 9 10.1016/j.enbuild.2017.12.028
Lyons DM, Lauring AS (2018) Mutation and epistasis in influenza virus evolution. Viruses 10:407. 10.3390/v10080407
Madigan MT (2009) Brock Biology of microorganisms. Pearson/Benjamin Cummings, San Francisco
Madsen AM, Larsen ST, Koponen IK, Kling KI, Barooni A, Karottki DG, Tendal K, Wolkoff P (2016) Generation and characterization of indoor fungal aerosols for inhalation studies. Appl Environ Microbiol 82(8):2479–2493. 10.1128/aem.04063-15
Madsen AM Zervas A Tendal K Nielsen JL Microbial diversity in bioaerosol samples causing ODTS compared to reference bioaerosol samples as measured using Illumina sequencing and MALDI-TOF Environ Res 2015 140 255 267 10.1016/j.envres.2015.03.027 25880607
Mandato S, Rondet E, Delaplace G, Barkouti A, Galet L, Accart P, Ruiz T, Cuq B (2012) Liquids’ atomization with two different nozzles: modeling of the effects of some processing and formulation conditions by dimensional analysis. Powder Technol 224:323–330. 10.1016/j.powtec.2012.03.014
Mander L (2016) A combinatorial approach to angiosperm pollen morphology. Proc Biol Sci. 10.1098/rspb.2016.2033
Marshall JW Vincent JH Kuehn TH Brosseau LM Studies of ventilation efficiency in a protective isolation room by the use of a scale model Infect Control Hosp Epidemiol 1996 17 1 5 10 10.2307/30142358 8789680
Marui VC Souto MLS Rovai ES Romito GA Chambrone L Pannuti CM Efficacy of preprocedural mouthrinses in the reduction of microorganisms in aerosol: A systematic review J Am Dent Assoc 2019 150 12 1015 1026.e1011 10.1016/j.adaj.2019.06.024 31761015
Mazej M Butala V Investigation in the characteristics of the personal ventilation using computational fluid dynamics Indoor Built Environ 2012 21 6 749 771 10.1177/1420326X11420456
Mazumdar S Yin Y Guity A Marmion P Gulick B Chen Q Impact of moving objects oncontaminant concentration distributions in an inpatient ward with displacement ventilation HVAC R Res 2010 16 5 545 563 10.1080/10789669.2010.10390921
Megri AC Haghighat F Zonal modeling for simulating indoor environment of buildings: review, recent developments, and applications HVAC&R Res 2007 13 6 887 905 10.1080/10789669.2007.10391461
Melikov AK (2004) Personalized ventilation. Indoor Air Supplement, 14(SUPPL. 7):157–167. 10.1111/j.1600-0668.2004.00284.x
Melikov AK Skwarczynski MA Kaczmarczyk J Zabecky J Use of personalized ventilation for improving health, comfort, and performance at high room temperature and humidity Indoor Air 2013 23 3 250 263 10.1111/ina.12012 23130883
Memarzadeh F Xu W Role of air changes per hour (ACH) in possible transmission of airborne infections Build Simul 2012 5 1 15 28 10.1007/s12273-011-0053-4 32218911
Mendell MJ Macher JM Kumagai K Measured moisture in buildings and adverse health effects: a review Indoor Air 2018 28 4 488 499 10.1111/ina.12464 29683210
Menzies D, Fanning A, Yuan L, Fitzgerald JM, Blanchette G, Boldiic Ph, Brassard P, Gharbonneau M, Déry S, Libman M, Miller M, Robiliard P, Schwartzman K, Tannenbaum TN, Tremblay J, Avandano M, Gonly J, Simor A, Naus M, Taylor G, Ma M, Montaner J (2000) Hospital ventilation and risk for tuberculous infection in Canadian health care workers. Ann Intern Med 133(10):779–789. 10.7326/0003-4819-133-10-200011210-00010
Milner J Vardoulakis S Chalabi Z Wilkinson P Modelling inhalation exposure to combustion-related air pollutants in residential buildings: application to health impact assessment Environ Int 2011 37 1 268 279 10.1016/j.envint.2010.08.015 20875687
Milton DK Fabian MP Cowling BJ Grantham ML McDevitt JJ Influenza virus aerosols in human exhaled breath: particle size, culturability, and effect of surgical masks PLoS Pathog 2013 9 3 e1003205 10.1371/journal.ppat.1003205 23505369
Milton DK Glencross MP Walters MD Risk of sick leave associated with outdoor air supply rate, humidification, and occupant complaints Indoor Air 2000 10 4 212 221 10.1034/j.1600-0668.2000.010004212.x 11089326
Mittal R Ni R Seo J-H The flow physics of COVID-19 J Fluid Mech 2020 894 F2 10.1017/jfm.2020.330
Mölter A, Lindley S, de Vocht F, Agius R, Kerry G, Johnson K, Ashmore M, Terry A, Dimitroulopoulou S, Simpson A (2012) Performance of a microenviromental model for estimating personal NO2 exposure in children. Atmos Environ 51:225–233. 10.1016/j.atmosenv.2012.01.030
Monn C Exposure assessment of air pollutants: a review on spatial heterogeneity and indoor/outdoor/personal exposure to suspended particulate matter, nitrogen dioxide and ozone Atmos Environ 2001 35 1 1 32 10.1016/S1352-2310(00)00330-7
Morawska L Droplet fate in indoor environments, or can we prevent the spread of infection? Indoor Air 2006 16 5 335 347 10.1111/j.1600-0668.2006.00432.x 16948710
Morawska L, Johnson GR, Ristovski ZD, Hargreaves M, Mengersen K, Corbett S, Chao CYH, Li Y, Katoshevski D (2009) Size distribution and sites of origin of droplets expelled from the human respiratory tract during expiratory activities. J Aerosol Sci 40(3):256–269. 10.1016/j.jaerosci.2008.11.002
Mothes N Valenta R Biology of tree pollen allergens Curr Allergy Asthma Rep 2004 4 5 384 390 10.1007/s11882-004-0089-y 15283879
Mousavi ES Grosskopf KR Ventilation rates and airflow pathways in patient rooms: a case study of bioaerosol containment and removal Ann Occup Hyg 2015 59 9 1190 1199 10.1093/annhyg/mev048 26187326
Mu D Gao N Zhu T Wind tunnel tests of inter-flat pollutant transmission characteristics in a rectangular multi-storey residential building, part A: effect of wind direction Build Environ 2016 108 159 170 10.1016/j.buildenv.2016.08.032 32287967
Mu D Shu C Gao N Zhu T Wind tunnel tests of inter-flat pollutant transmission characteristics in a rectangular multi-storey residential building, part B: Effect of source location Build Environ 2017 114 281 292 10.1016/j.buildenv.2016.12.031 32287970
Mui KW Wong LT Wu CL Lai ACK Numerical modeling of exhaled droplet nuclei dispersion and mixing in indoor environments J Hazard Mater 2009 167 1–3 736 744 10.1016/j.jhazmat.2009.01.041 19232824
Muller MP MacDougall C Lim M Antimicrobial surfaces to prevent healthcare-associated infections: a systematic review J Hosp Infect 2016 92 1 7 13 10.1016/j.jhin.2015.09.008 26601608
Mumtaz M Fisher J Blount B Ruiz P Application of physiologically based pharmacokinetic models in chemical risk assessment Journal of Toxicology 2012 2012 904603 10.1155/2012/904603 22523493
Murga A Kuga K Yoo S-J Ito K Personal inhalation risk assessment based on a hybrid method using CFD-CSP-PBTK modelling: quantification of time-averaged and peak concentration differences IOP Conf Ser Mater Sci Eng 2019 609 042003 10.1088/1757-899x/609/4/042003
Murga A Long Z Yoo S-J Sumiyoshi E Ito K Decreasing inhaled contaminant dose of a factory worker through a hybrid emergency ventilation system: performance evaluation in worst-case scenario Energy Built Environ 2020 1 3 319 326 10.1016/j.enbenv.2020.04.007
Murray PR What is new in clinical microbiology-microbial identification by MALDI-TOF mass spectrometry: a paper from the 2011 William Beaumont Hospital Symposium on molecular pathology J Mol Diagn 2012 14 5 419 423 10.1016/j.jmoldx.2012.03.007 22795961
Murray PR, Rosenthal KS, Pfaller MA (2013) Medical microbiology. Elsevier Saunders, Philadelphia
Mutuku JK Hou W-C Chen W-H An overview of experiments and numerical simulations on airflow and aerosols deposition in human airways and the role of bioaerosol motion in COVID-19 transmission Aerosol Air Qual Res 2020 20 6 1172 1196 10.4209/aaqr.2020.04.0185
Mutuku JK, Hou W-C, Chen W-H (2020b) Two-phase flow dynamics and PM2.5 Deposition in healthy and obstructed human airways during inhalation. Aerosol Air Qual Res 20:1094–1110. 10.4209/aaqr.2020.03.0107
Myatt TA Johnston SL Zuo Z Wand M Kebadze T Rudnick S Milton DK Detection of airborne rhinovirus and its relation to outdoor air supply in office environments Am J Respir Crit Care Med 2004 169 11 1187 1190 10.1164/rccm.200306-760OC 14754759
Napoli C Marcotrigiano V Montagna MT Air sampling procedures to evaluate microbial contamination: a comparison between active and passive methods in operating theatres BMC Public Health 2012 12 594 10.1186/1471-2458-12-594 22853006
Nardell EA Keegan J Cheney SA Etkind SC Airborne Infection: Theoretical limits of protection achievable by building ventilation Am Rev Respir Dis 1991 144 2 302 306 10.1164/ajrccm/144.2.302 1907115
Nasir ZA, Hayes E, Williams B, Gladding T, Rolph C, Khera S, Jackson S, Bennett A, Collins S, Parks S, Attwood A, Kinnersley RP, Walsh K, Alcega SG, Pollard SJT, Drew G, Coulon F, Tyrrel S (2019) Scoping studies to establish the capability and utility of a real-time bioaerosol sensor to characterise emissions from environmental sources. Sci Total Environ 648:25–32. 10.1016/j.scitotenv.2018.08.120
Nasir ZA Rolph CA Collins S Stevenson DW Gladding TL Hayes ET Williams B Khera S Jackson S Bennett A Parks S Kinnersley RP Walsh K Pollard SJT Drew G Alcega SG Coulon F Tyrrel SF A Controlled study on the characterisation of bioaerosols emissions from compost Atmosphere 2018 9 10 379 10.3390/atmos9100379
Nazaroff WW Indoor particle dynamics Indoor Air 2004 14 Suppl 7 175 183 10.1111/j.1600-0668.2004.00286.x 15330785
Nazaroff WW Indoor bioaerosol dynamics Indoor Air 2016 26 1 61 78 10.1111/ina.12174 25483392
Neely AN Orloff MM Survival of some medically important fungi on hospital fabrics and plastics J Clin Microbiol 2001 39 9 3360 3361 10.1128/jcm.39.9.3360-3361.2001 11526178
Netz RR Mechanisms of airborne infection via evaporating and sedimenting droplets produced by speaking J Phys Chem B 2020 10.1021/acs.jpcb.0c05229
Ni Y Shi G Qu J Indoor PM2.5, tobacco smoking and chronic lung diseases: a narrative review Environ Res 2020 181 108910 10.1016/j.envres.2019.108910 31780052
Nicas M An analytical framework for relating dose, risk, and incidence: an application to occupational tuberculosis infection Risk Anal 1996 16 4 527 538 10.1111/j.1539-6924.1996.tb01098.x 8819343
Nielsen PV (2009) Control of airborne infectious diseases in ventilated spaces. J R Soc Interface 6(suppl_6):S747–S755. 10.1098/rsif.2009.0228.focus
Nielsen PV Fifty years of CFD for room air distribution Build Environ 2015 91 78 90 10.1016/j.buildenv.2015.02.035
Nielsen PV, Bartholomaeussen NM, Jakubowska E, Jiang H, Jonsson OT, Krawiecka K,Mierzejewski A, Jessica SJ, Trampczynska K, Polak M, Soennichsen M (2007a) Chair with integrated personalized ventilation for minimizing cross infection. Aalborg University. https://vbn.aau.dk/ws/portalfiles/portal/13819115/Chair_with_Integrated_Personalized_Ventilation_for_Minimizing_Cross_Infection. Accessed 24 Nov 2022
Nielsen PV Hyldgaard CE Melikov A Andersen H Soennichsen M Personal exposure between people in a room ventilated by textile terminals—with and without personalized ventilation HVAC R Res 2007 13 4 635 643 10.1080/10789669.2007.10390976
Nielsen PV Li Y Buus M Winther FV Risk of cross-infection in a hospital ward with downward ventilation Build Environ 2010 45 9 2008 2014 10.1016/j.buildenv.2010.02.017
Nigro OD Steward GF Differential specificity of selective culture media for enumeration of pathogenic vibrios: advantages and limitations of multi-plating methods J Microbiol Methods 2015 111 24 30 10.1016/j.mimet.2015.01.014 25602161
Niu J, Tung C, Wan J, Cheng J (2005) CFD simulation of interflat air flow for the study of the spread of aerosol transmitted infectious diseases. International Building Performance Simulation Association. http://www.ibpsa.org/proceedings/BS2005/BS05_0853_858.pdf. Accessed 24 Nov 2022
Niu J Tung TCW On-site quantification of re-entry ratio of ventilation exhausts in multi-family residential buildings and implications Indoor Air 2008 18 1 12 26 10.1111/j.1600-0668.2007.00500.x 18093125
Noakes CJ Andrew Sleigh P Mathematical models for assessing the role of airflow on the risk of airborne infection in hospital wards J R Soc Interface 2009 6 SUPPL. 6 S791 S800 10.1098/rsif.2009.0305.focus 19812072
Novozhilov V Fire suppression studies Therm Sci 2007 11 2 161 180 10.2298/TSCI0702161N
Olesen BW Simone A Krajčík M Causone F De Carli M Experimental study of air distribution and ventilation effectiveness in a room with a combination of different mechanical ventilation and heating/cooling systems Int J Vent 2011 9 4 371 383 10.1080/14733315.2011.11683895
Olmedo I Nielsen PV Ruiz de Adana M Jensen RL Grzelecki P Distribution of exhaled contaminants and personal exposure in a room using three different air distribution strategies Indoor Air 2012 22 1 64 76 10.1111/j.1600-0668.2011.00736.x 21815935
Ouazia B Macdonald I Tardif M Thompson A Booth D Contaminant removal effectiveness of displacement ventilationsystems during heating season; summary results from three field studies ASHRAE Trans 2012 118 2 292 299
Ouazia B, Tardif M, Macdonald I, Thompson A, Booth D (2011) In-situ performance of displacement ventilation system in Canadian schools with radiant heating systems. ASHRAE Annual Conference, Montreal, Quebec, pp 1–13
Ozkaynak H, Palma T, Touma JS, Thurman J (2007) Modeling population exposures to outdoor sources of hazardous air pollutants. J Expos Sci Environ Epidemiol 18(1):45–58. 10.1038/sj.jes.7500612
Özkaynak H Palma T Touma JS Thurman J Modeling population exposures to outdoor sources of hazardous air pollutants J Eposure Sci Environ Epidemiol 2008 18 1 45 58 10.1038/sj.jes.7500612
Pablos I Wildner S Asam C Wallner M Gadermaier G Pollen allergens for molecular diagnosis Curr Allergy Asthma Rep 2016 16 4 31 10.1007/s11882-016-0603-z 27002515
Pantelic J Sze-To GN Tham KW Chao CYH Khoo YCM Personalized ventilation as a control measure for airborne transmissible disease spread J R Soc Interface 2009 6 SUPPL. 6 S715 S726 10.1098/rsif.2009.0311.focus 19812074
Pantelic J, Tham KW (2011) Assessment of the ability of different ventilation systems to serve as a control measure against airborne infectious disease transmission using Wells-Riley approach. American Society of Heating Refrigeration and Airconditioning Engineers. https://scholarbank.nus.edu.sg/handle/10635/45912. Accessed 24 Nov 2022
Pantelic J Tham KW Adequacy of air change rate as the sole indicator of an air distribution system’s effectiveness to mitigate airborne infectious disease transmission caused by a cough release in the room with overhead mixing ventilation: a case study HVAC R Res 2013 19 8 947 961 10.1080/10789669.2013.842447
Pantelic J Tham KW Licina D Effectiveness of a personalized ventilation system in reducing personal exposure against directly released simulated cough droplets Indoor Air 2015 25 6 683 693 10.1111/ina.12187 25615014
Pantelic J, Wai TK (2009) Effect of the room air motion on the dispersion of expiratory droplets in the personalized ventilated room. International Society of Indoor Air Quality and Climate. https://scholarbank.nus.edu.sg/handle/10635/45936. Accessed 24 Nov 2022
Park DY Chang S Numerical investigation of thermal comfort and transport of expiratory contaminants in a ventilated office with an air curtain system Indoor Built Environ 2019 28 3 401 421 10.1177/1420326x18770238
Park JW Kim HR Hwang J Continuous and real-time bioaerosol monitoring by combined aerosol-to-hydrosol sampling and ATP bioluminescence assay Anal Chim Acta 2016 941 101 107 10.1016/j.aca.2016.08.039 27692374
Patankar SV (1980) Numerical heat transfer and fluid flow. Hemisphere Publishing Corporation, New York
Penconek A Michalczuk U Sienkiewicz A Moskal A The effect of desert dust particles on rheological properties of saliva and mucus Environ Sci Pollut Res Int 2019 26 12 12150 12157 10.1007/s11356-019-04628-x 30830668
Pendar M-R Páscoa JC Numerical modeling of the distribution of virus carrying saliva droplets during sneeze and cough Phys Fluids 2020 32 8 083305 10.1063/5.0018432
Peng S Chen Q Liu E The role of computational fluid dynamics tools on investigation of pathogen transmission: prevention and control Sci Total Environ 2020 746 142090 10.1016/j.scitotenv.2020.142090 33027870
Peteranderl C Herold S Schmoldt C Human Influenza virus infections Semin Respir Crit Care Med 2016 37 4 487 500 10.1055/s-0036-1584801 27486731
Phu H-T, Park Y, Andrews AJ, Marabella I, Abraham A, Mimmack R, Olson BA, Chaika J, Floersch E, Remskar M, Hume JR, Fischer GA, Belani K, Hogan CJ (2020) Design and evaluation of a portable negative pressure hood with HEPA filtration to protect health care workers treating patients with transmissible respiratory infections. Am J Infect Control. 10.1016/j.ajic.2020.06.203
Phuong NL Ito K Investigation of flow pattern in upper human airway including oral and nasal inhalation by PIV and CFD Build Environ 2015 94 504 515 10.1016/j.buildenv.2015.10.002
Phuong NL Khoa ND Ito K Comparative numerical simulation of inhaled particle dispersion in upper human airway to analyse intersubject differences Indoor Built Environ 2020 29 6 793 809 10.1177/1420326x19894128
Pitt JI Christian JH Heat resistance of xerophilic fungi based on microscopical assessment of spore survival Appl Microbiol 1970 20 5 682 686 10.1128/am.20.5.682-686.1970 5485080
Pleschka S Overview of influenza viruses Curr Top Microbiol Immunol 2013 370 1 20 10.1007/82_2012_272 23124938
Pomeranz G, Pando R, Hindiyeh M, Sherbany H, Meningher T, Sharabi S, Kolet L, Pomeranz A, Mendelson E, Mandelboim M (2019) Rhinovirus infections in infants suggest that early detection can prevent unnecessary treatment. J Clin Virol 115:11–17. 10.1016/j.jcv.2019.03.012
Poon WCK, Brown AT, Direito SOL, Hodgson DJM, Le Nagard L, Lips A, MacPhee CA, Marenduzzo D, Royer JR, Silva AF, Thijssen JH, Titmuss S (2020) Soft matter science and the COVID-19 pandemic. Soft Matter 16(36):8310–8324. 10.1039/D0SM01223H
Prat C Lacoma A Bacteria in the respiratory tract-how to treat? Or do not treat? Int J Infect Dis 2016 51 113 122 10.1016/j.ijid.2016.09.005 27776777
Predicala BZ Urban JE Maghirang RG Jerez SB Goodband RD Assessment of bioaerosols in swine barns by filtration and impaction Curr Microbiol 2002 44 2 136 140 10.1007/s00284-001-0064-y 11815859
Prussin AJ, Belser JA, Bischoff W, Kelley ST, Lin K, Lindsley WG, Nshimyimana JP, Schuit M, Wu Z, Bibby K, Marr LC (2020) Viruses in the Built Environment (VIBE) meeting report. Microbiome 8(1):1. 10.1186/s40168-019-0777-4
Prussin AJ Schwake DO Marr LC Ten questions concerning the aerosolization and transmission of Legionella in the built environment Build Environ 2017 123 684 695 10.1016/j.buildenv.2017.06.024 29104349
Pyankov OV Bodnev SA Pyankova OG Agranovski IE Survival of aerosolized coronavirus in the ambient air J Aerosol Sci 2018 115 158 163 10.1016/j.jaerosci.2017.09.009 32226116
Qian H Li Y Removal of exhaled particles by ventilation and deposition in a multibed airborne infection isolation room Indoor Air 2010 20 4 284 297 10.1111/j.1600-0668.2010.00653.x 20546037
Qian H Li Y Nielsen PV Hyldgaard CE Dispersion of exhalation pollutants in a two-bed hospital ward with a downward ventilation system Build Environ 2008 43 3 344 354 10.1016/j.buildenv.2006.03.025
Qian H Li Y Nielsen PV Hyldgaard CE Wong TW Chwang ATY Dispersion of exhaled droplet nuclei in a two-bed hospital ward with three different ventilation systems Indoor Air 2006 16 2 111 128 10.1111/j.1600-0668.2005.00407.x 16507039
Qian H Li Y Seto WH Ching P Ching WH Sun HQ Natural ventilation for reducing airborne infection in hospitals Build Environ 2010 45 3 559 565 10.1016/j.buildenv.2009.07.011 32288008
Qian H Zheng X Ventilation control for airborne transmission of human exhaled bio-aerosols in buildings J Thorac Dis 2018 10 S2295 S2304 10.21037/jtd.2018.01.24 30116608
Rantio-Lehtimäki A Viander M Koivikko A Airborne birch pollen antigens in different particle sizes Clin Exp Allergy 1994 24 1 23 28 10.1111/j.1365-2222.1994.tb00912.x 8156441
Rath B, Conrad T, Myles P, Alchikh M, Ma X, Hoppe C, Tief F, Chen X, Obermeier P, Kisler B, Schweiger B (2017) Influenza and other respiratory viruses: standardizing disease severity in surveillance and clinical trials. Expert Rev Anti Infect Ther 15(6):545–568. 10.1080/14787210.2017.1295847
Razzini K, Castrica M, Menchetti L, Maggi L, Negroni L, Orfeo NV, Pizzoccheri A, Stocco M, Muttini S, Balzaretti CM (2020) SARS-CoV-2 RNA detection in the air and on surfaces in the COVID-19 ward of a hospital in Milan, Italy. Sci Total Environ 742:140540. 10.1016/j.scitotenv.2020.140540
Reddy MS Vedamuthu ER Washam CJ Reinbold GW Agar medium for differential enumeration of lactic streptococci Appl Microbiol 1972 24 6 947 952 10.1128/am.24.6.947-952.1972 16349952
Ren Z Stewart J Prediction of personal exposure to contaminant sources in industrial buildings using a sub-zonal model Environ Model Softw 2005 20 5 623 638 10.1016/j.envsoft.2004.03.007
Rhodes M (2008) Introduction to particle technology. John Wiley & Sons, Ltd, Chichester
Rodríguez R Villalba M Batanero E Palomares O Salamanca G Emerging pollen allergens Biomed Pharmacother 2007 61 1 1 7 10.1016/j.biopha.2006.09.014 17188831
Rothan HA Byrareddy SN The epidemiology and pathogenesis of coronavirus disease (COVID-19) outbreak J Autoimmun 2020 109 102433 10.1016/j.jaut.2020.102433 32113704
Rui Z Guangbei T Jihong L Study on biological contaminant control strategies under different ventilation models in hospital operating room Build Environ 2008 43 5 793 803 10.1016/j.buildenv.2007.01.018
Ryan KA, Bewley KR, Fotheringham SA, Brown P, Hall Y, Marriott AC, Cavell BE, Tree JA, Allen L, Aram MJ, Bean TJ, Brunt E, Buttigieg KR, Carter DP, Cobb R, Coombes NS, Findlay-Wilson SJ, Godwin KJ, Gooch KE, Gouriet J, Halkerston R, Harris DJ, Hender TH, Humphries HE, Hunter L, Ho CMK, Kennard CL, Leung S, Longet S, Ngabo D, Osman KL, Paterson J, Penn EJ, Steven T. Pullan, Emma Rayner, Oliver Skinner, Kimberley Steeds, Irene Taylor, Tipton T, Thomas S, Turner C, Watson RJ, Wiblin NR, Charlton S, Hallis B, Hiscox JA, Funnell S, Dennis MJ, Whittaker CJ, Catton MG, Druce J, Salguero FJ, Carroll MW (2020) Dose-dependent response to infection with SARS-CoV-2 in the ferret model: evidence of protection to re-challenge. bioRxiv, 2020.2005.2029.123810. 10.1101/2020.05.29.123810
Saarinen PE Kalliomäki P Tang JW Koskela H Large eddy simulation of air escape through a hospital isolation room single hinged doorway—validation by using tracer gases and simulated smoke videos PLoS One 2015 10 7 e0130667 10.1371/journal.pone.0130667 26151865
Sadrizadeh S Holmberg S Tammelin A A numerical investigation of vertical and horizontal laminar airflow ventilation in an operating room Build Environ 2014 82 517 525 10.1016/j.buildenv.2014.09.013
Sadrizadeh S Pantelic J Sherman M Clark J Abouali O Airborne particle dispersion to an operating room environment during sliding and hinged door opening J Infect Public Health 2018 11 5 631 635 10.1016/j.jiph.2018.02.007 29526441
Sadrizadeh S Tammelin A Ekolind P Holmberg S Influence of staff number and internal constellation on surgical site infection in an operating room Particuology 2014 13 42 51 10.1016/j.partic.2013.10.006
Sajjadi H Salmanzadeh M Ahmadi G Jafari S Simulations of indoor airflow and particle dispersion and deposition by the lattice Boltzmann method using LES and RANS approaches Build Environ 2016 102 1 12 10.1016/j.buildenv.2016.03.006
Sakharov AS Zhukov K Study of an air curtain in the context of individual protection from exposure to coronavirus (SARS-CoV-2) contained in cough-generated fluid particles Physics 2020 2 3 340 351 10.3390/physics2030018
Salmanzadeh M Zahedi G Ahmadi G Marr DR Glauser M Computational modeling of effects of thermal plume adjacent to the body on the indoor airflow and particle transport J Aerosol Sci 2012 53 29 39 10.1016/j.jaerosci.2012.05.005
Salthammer T Emerging indoor pollutants Int J Hyg Environ Health 2020 224 113423 10.1016/j.ijheh.2019.113423 31978722
Sarkar A Xu F Lee S Human saliva and model saliva at bulk to adsorbed phases – similarities and differences Adv Coll Interface Sci 2019 273 102034 10.1016/j.cis.2019.102034
Satheesan MK Mui KW Wong LT A numerical study of ventilation strategies for infection risk mitigation in general inpatient wards Build Simul 2020 13 4 887 896 10.1007/s12273-020-0623-4 32211123
Scharfman BE Techet AH Bush JWM Bourouiba L Visualization of sneeze ejecta: steps of fluid fragmentation leading to respiratory droplets Exp Fluids 2016 57 2 24 10.1007/s00348-015-2078-4 32214638
Scheuch G Breathing is enough: for the spread of influenza virus and SARS-CoV-2 by breathing only J Aerosol Med Pulm Drug Deliv 2020 10.1089/jamp.2020.1616
Schiavon S Bauman FS Tully B Rimmer J Chilled ceiling and displacement ventilation system: laboratory study with high cooling load Sci Technol Built Environ 2015 21 7 944 956 10.1080/23744731.2015.1034061
Schroeter JD Kimbell JS Asgharian B Tewksbury EW Singal M Computational fluid dynamics simulations of submicrometer and micrometer particle deposition in the nasal passages of a Sprague-Dawley rat J Aerosol Sci 2012 43 1 31 44 10.1016/j.jaerosci.2011.08.008
Schuijs MJ, Willart MA, Vergote K, Gras D, Deswarte K, Ege MJ, Madeira FB, Beyaert R, Loo G, Bracher F, Mutius E, Chanez P, Lambrecht BN, Hammad H (2015) Farm dust and endotoxin protect against allergy through A20 induction in lung epithelial cells. Science 349(6252):1106–1110. 10.1126/science.aac6623
Seepana S Lai ACK Experimental and numerical investigation of interpersonal exposure of sneezing in a full-scale chamber Aerosol Sci Technol 2012 46 5 485 493 10.1080/02786826.2011.640365
Seymour MJ, Alani A, Manning A, Jiang J (2000) CFD based airflow modelling to investigate the effectiveness of control methods intended to prevent the transmission of airborne organisms. Paper presented at the Air Distribution in Rooms: Ventilation for Health and Sustainable Environment, Reading, UK
Shajahan A Culp CH Williamson B Effects of indoor environmental parameters related to building heating, ventilation, and air conditioning systems on patients’ medical outcomes: a review of scientific research on hospital buildings Indoor Air 2019 29 2 161 176 10.1111/ina.12531 30588679
Shao S, Zhou D, He R, Li J, Zou S, Mallery K, Kumar S, Yang S, Hong J (2021) Risk assessment of airborne transmission of COVID-19 by asymptomatic individuals under different practical settings. J Aerosol Sci 151:105661. 10.1016/j.jaerosci.2020.105661
Shao X Liang S Li X Liang C Yan S Quantitative effects of supply air and contaminant sources on steady contaminant distribution in ventilated space with air recirculation Build Environ 2020 171 106672 10.1016/j.buildenv.2020.106672
Shi H Tarabara VV Charge, size distribution and hydrophobicity of viruses: Effect of propagation and purification methods J Virol Methods 2018 256 123 132 10.1016/j.jviromet.2018.02.008 29438732
Shi Z, Lu Z, Chen Q (2019) Indoor airflow and contaminant transport in a room with coupled displacement ventilation and passive-chilled-beam systems. Build Environ 161:106244. 10.1016/j.buildenv.2019.106244
Short CA Al-Maiyah S Design strategy for low-energy ventilation and cooling of hospitals Build Res Inf 2009 37 3 264 292 10.1080/09613210902885156
Shrubsole C, Ridley I, Biddulph P, Milner J, Vardoulakis S, Ucci M, Wilkinson P, Chalabi Z, Davies M (2012) Indoor PM2.5 exposure in London’s domestic stock: modelling current and future exposures following energy efficient refurbishment. Atmos Environ 62:336–343
Singh SK Middle East respiratory syndrome virus pathogenesis Semin Respir Crit Care Med 2016 37 4 572 577 10.1055/s-0036-1584796 27486737
Singhal N Kumar M Kanaujia PK Virdi JS MALDI-TOF mass spectrometry: an emerging technology for microbial identification and diagnosis Front Microbiol 2015 6 791 10.3389/fmicb.2015.00791 26300860
Siqueira JF Jr Rôças IN PCR methodology as a valuable tool for identification of endodontic pathogens J Dent 2003 31 5 333 339 10.1016/s0300-5712(03)00051-4 12799118
Smith M Berger U Behrendt H Bergmann KC Pollen and pollinosis Chem Immunol Allergy 2014 100 228 233 10.1159/000358743 24925402
Snyder H Literature review as a research methodology: an overview and guidelines J Bus Res 2019 104 333 339 10.1016/j.jbusres.2019.07.039
Soares TL Jesus ON Souza EH Rossi ML Oliveira EJ Comparative pollen morphological analysis in the subgenera Passiflora and Decaloba An Acad Bras Cienc 2018 90 2 suppl 1 2381 2396 10.1590/0001-3765201720170248 29044321
Soni SK Kirar PK Kolhe P Sahu KC Deformation and breakup of droplets in an oblique continuous air stream Int J Multiph Flow 2020 122 103141 10.1016/j.ijmultiphaseflow.2019.103141
Spalding DB A general purpose computer program for multi-dimensional one- and two-phase flow Math Comput Simul 1981 23 3 267 276 10.1016/0378-4754(81)90083-5
Spiegelman D Whissell G Greer CW A survey of the methods for the characterization of microbial consortia and communities Can J Microbiol 2005 51 5 355 386 10.1139/w05-003 16088332
Srebric J, Yuan J, Novoselac A (2008) On-site experimental validation of a coupled multizone and CFD model for building contaminant transport simulations. ASHRAE Transactions, 14(1):273–281
Srivastav A, Santibanez TA, Lu PJ, Stringer MC, Dever JA, Bostwick M, Stanley-Kurtz M, Qualls NL, Williams WW (2018) Preventive behaviors adults report using to avoid catching or spreading influenza, United States, 2015-16 influenza season. Plos One 13(3):e0195085. 10.1371/journal.pone.0195085
Stadnytskyi V Bax CE Bax A Anfinrud P The airborne lifetime of small speech droplets and their potential importance in SARS-CoV-2 transmission Proc Natl Acad Sci 2020 117 22 11875 10.1073/pnas.2006874117 32404416
Steiner S Herve P Pak C Majeed S Sandoz A Kuczaj A Hoeng J Development and testing of a new-generation aerosol exposure system: the independent holistic air-liquid exposure system (InHALES) Toxicol in Vitro 2020 67 104909 10.1016/j.tiv.2020.104909 32512146
Stewart J Ren Z COwZ—a subzonal indoor airflow, temperature and contaminant dispersion model Build Environ 2006 41 12 1631 1648 10.1016/j.buildenv.2005.06.015
Stilianakis NI Drossinos Y Dynamics of infectious disease transmission by inhalable respiratory droplets J R Soc Interface 2010 7 50 1355 1366 10.1098/rsif.2010.0026 20164087
Stockwell RE Ballard EL O’Rourke P Knibbs LD Morawska L Bell SC Indoor hospital air and the impact of ventilation on bioaerosols: a systematic review J Hosp Infect 2019 103 2 175 184 10.1016/j.jhin.2019.06.016 31279762
Sun W Ji J Transport of droplets expelled by coughing in ventilated rooms Indoor Built Environ 2007 16 6 493 504 10.1177/1420326X07084290
Sun Y, Wang Z, Zhang Y, Sundell J (2011) In China, students in crowded dormitories with a low ventilation rate have more common colds: evidence for airborne transmission. Plos One 6(11):27140. 10.1371/journal.pone.0027140
Sundell J, Levin H, Nazaroff WW, Cain WS, Fisk WJ, Grimsrud DT, Gyntelberg F, Li Y, Persily AK, Pickering AC, Samet JM, Spengler JD, Taylor ST, Weschler CJ (2011) Ventilation rates and health: multidisciplinary review of the scientific literature. Indoor Air 21(3):191–204. 10.1111/j.1600-0668.2010.00703.x
Sung JY Hwang Y Shin MH Park MS Lee SH Yong D Lee K Utility of conventional culture and MALDI-TOF MS for identification of microbial communities in bronchoalveolar lavage fluid in comparison with the GS junior next generation sequencing system Ann Lab Med 2018 38 2 110 118 10.3343/alm.2018.38.2.110 29214754
Tang JW Li Y Eames I Chan PKS Ridgway GL Factors involved in the aerosol transmission of infection and control of ventilation in healthcare premises J Hosp Infect 2006 64 2 100 114 10.1016/j.jhin.2006.05.022 16916564
Tang JW, Liebner TJ, Craven BA, Settles GS (2009) A schlieren optical study of the human cough with and without wearing masks for aerosol infection control. J R Soc Interface 6(suppl_6):S727-S736. 10.1098/rsif.2009.0295.focus
Tao Y Inthavong K Petersen P Mohanarangam K Yang W Tu J Vortex structures and wake flow analysis from moving manikin models Indoor Built Environ 2020 30 3 347 362 10.1177/1420326x19893013
Tao Y Yang W Ito K Inthavong K Computational fluid dynamics investigation of particle intake for nasal breathing by a moving body Exp Comput Multiph Flow 2019 1 3 212 218 10.1007/s42757-019-0014-1
Taylor J Lai KM Nasir ZA Human factors and bioagent transmission following an indoor bioterror attack J Bioterror Biodef 2012 3 1 1000116 10.4172/2157-2526.1000116
Taylor-Robinson D Tyrrell DA Serotypes of viruses (rhinoviruses) isolated from common colds Lancet 1962 1 7227 452 454 10.1016/s0140-6736(62)91418-6 13920011
Tham KW, Pantelic J (2011) Cough released airborne infection disease transmission control with ventilation at various infectorsusceptible distances. Paper presented at the IAQ Conference
Thompson KA Bennett AM Persistence of influenza on surfaces J Hosp Infect 2017 95 2 194 199 10.1016/j.jhin.2016.12.003 28139390
Tian ZF Tu JY Yeoh GH Yuen RKK Numerical studies of indoor airflow and particle dispersion by large Eddy simulation Build Environ 2007 42 10 3483 3492 10.1016/j.buildenv.2006.10.047
To KKW Yip CCY Yuen KY Rhinovirus - from bench to bedside J Formos Med Assoc 2017 116 7 496 504 10.1016/j.jfma.2017.04.009 28495415
Tran K Cimon K Severn M Pessoa-Silva CL Conly J Aerosol generating procedures and risk of transmission of acute respiratory infections to healthcare workers: a systematic review PLoS One 2012 7 4 e35797 10.1371/journal.pone.0035797 22563403
Tung YC Hu SC Infection risk of indoor airborne transmission of diseases in multiple spaces Archit Sci Rev 2008 51 1 14 20 10.3763/asre.2008.5103
Urbán A Zaremba M Malý M Józsa V Jedelský J Droplet dynamics and size characterization of high-velocity airblast atomization Int J Multiph Flow 2017 95 1 11 10.1016/j.ijmultiphaseflow.2017.02.001
Usachev EV Usacheva OV Agranovski IE Surface plasmon resonance-based real-time bioaerosol detection J Appl Microbiol 2013 115 3 766 773 10.1111/jam.12267 23725222
Vadivukkarasan M Dhivyaraja K Panchagnula MV Breakup morphology of expelled respiratory liquid: from the perspective of hydrodynamic instabilities Phys Fluids 2020 32 9 094101 10.1063/5.0022858
Valero N Aguilera I Llop S Esplugues A de Nazelle A Ballester F Sunyer J Concentrations and determinants of outdoor, indoor and personal nitrogen dioxide in pregnant women from two Spanish birth cohorts Environ Int 2009 35 8 1196 1201 10.1016/j.envint.2009.08.002 19740538
Verijkazemi K Mansouri N Moattar F Khezri SM Evaluation of indoor PM distribution by CONTAM Airflow model and real time measuring: model description and validation Avicenna J Environ Health Eng 2018 5 1 42 49 10.15171/ajehe.2018.06
Vianello A Jensen RL Liu L Vollertsen J Simulating human exposure to indoor airborne microplastics using a breathing thermal manikin Sci Rep 2019 9 1 8670 10.1038/s41598-019-45054-w 31209244
Vidlak D Kielian T Infectious dose dictates the host response during Staphylococcus aureus orthopedic-implant biofilminfection Infect Immun 2016 84 7 1957 10.1128/IAI.00117-16 27091926
Villafruela JM Olmedo I Berlanga FA Ruiz de Adana M Assessment of displacement ventilation systems in airborne infection risk in hospital rooms PLoS One 2019 14 1 e0211390 10.1371/journal.pone.0211390 30699182
Viswanathan H Breakup and coalescence of drops during transition from dripping to jetting in a Newtonian fluid Int J Multiph Flow 2019 112 269 285 10.1016/j.ijmultiphaseflow.2018.09.016
Vogazianos P, Argyropoulos CD, Haralambous C, Mikellidou CV, Boustras G, Andreou M, Silvestros V, Theofanous F, Soteriou S, Gregoriou I, Jaber-Apostolidou A, Demetriou A, Athanasiadou M, Stylianou C, Michael M, Herodotou H, Alexandrou D, Kalakouta O, Pana ZD (2021) Impact assessment of COVID-19 non-pharmaceutical interventions in long term care facilities in Cyprus: Safety improvement strategy. Saf Sci 143:105415. 10.1016/j.ssci.2021.105415
Vuorinen V, Aarnio M, Alava M, Alopaeus V, Atanasova N, Auvinen M, Balasubramanian N, Bordbar H, Erästö P, Grande R, Hayward N, Hellsten A, Hostikka S, Hokkanen J, Kaario O, Karvinen A, Kivistö I, Korhonen M, Kosonen R, Kuusela J, Lestinen S, Laurila E, Nieminen HJ, Peltonen P, PokkJ i, Puisto A, Råback P, Salmenjoki H, Sironen T, Österberg M (2020) Modelling aerosol transport and virus exposure with numerical simulations in relation to SARS-CoV-2 transmission by inhalation indoors. Saf Sci 130:104866. 10.1016/j.ssci.2020.104866
Wady L Larsson L Determination of microbial volatile organic compounds adsorbed on house dust particles and gypsum board using SPME/GC-MS Indoor Air 2005 15 Suppl 9 27 32 10.1111/j.1600-0668.2005.00293.x
Wai TK, Pantelic J (2009) Influence of different personalized air terminal Devices on the motion of expiratory droplets released in the closed vicinity of the occupant. Paper presented at the 9th International Conference and Exhibition - Healthy Buildings 2009, HB 2009
Walsh RL, Camilli A (2011) Streptococcus pneumoniae is desiccation tolerant and infectious upon rehydration. mBio 2(3):e00092–00011. 10.1128/mBio.00092-11
Walton GN (1989) AIRNET - a computer program for building airflow network modeling. U.S. Department of Commerce. https://nvlpubs.nist.gov/nistpubs/Legacy/IR/nistir89-4072.pdf. Accessed 24 Nov 2022
Walton GN, Dols WS (2005) CONTAM 2.4 User Guide and Program Documentation. National Institute of Standards and Technology. https://nvlpubs.nist.gov/nistpubs/Legacy/IR/nistir7251.pdf. Accessed 24 Nov 2022
Wan MP Chao CYH Ng YD Sze To GN Yu WC Dispersion of expiratory droplets in a general hospital ward with ceiling mixing type mechanical ventilation system Aerosol Sci Technol 2007 41 3 244 258 10.1080/02786820601146985
Wang C, Holmberg S, Sadrizadeh S (2019a) Impact of door opening on the risk of surgical site infections in an operating room with mixing ventilation. Indoor Built Environ 0(0):1420326X19888276. 10.1177/1420326x19888276
Wang CT Diagnosing and treating asymptomatic tuberculosis infection Can Fam Physician 1999 45 2397 2404 10540699
Wang H Zhai Z Advances in building simulation and computational techniques: a review between 1987 and 2014 Energy and Buildings 2016 128 319 335 10.1016/j.enbuild.2016.06.080
Wang J Huo Q Zhang T Wang S Battaglia F Numerical investigation of gaseous pollutant cross-transmission for single-sided natural ventilation driven by buoyancy and wind Build Environ 2020 172 106705 10.1016/j.buildenv.2020.106705 32287994
Wang J-X Cao X Chen Y-P An air distribution optimization of hospital wards for minimizing cross-infection J Clean Prod 2021 279 123431 10.1016/j.jclepro.2020.123431 32836912
Wang JH Niu JL Liu XP Yu CWF Assessment of pollutant dispersion in the re-entrance space of a high-rise residential building, using wind tunnel simulations Indoor Built Environ 2010 19 6 638 647 10.1177/1420326X10386669
Wang L Chen Q Theoretical and numerical studies of coupling multizone and CFD models for building air distribution simulations Indoor Air 2007 17 5 348 361 10.1111/j.1600-0668.2007.00481.x 17880631
Wang L Chen Q Applications of a coupled multizone-CFD model to calculate airflow and contaminant dispersion in built environments for emergency management HVAC&R Res 2008 14 6 925 939 10.1080/10789669.2008.10391047
Wang L Chen Q Evaluation of some assumptions used in multizone airflow network models Build Environ 2008 43 10 1671 1677 10.1016/j.buildenv.2007.10.010
Wang LL Dols WS Chen Q Using CFD capabilities of CONTAM 3.0 for simulating airflow and contaminant transport in and around buildings HVAC&R Res 2010 16 749 763 10.1080/10789669.2010.10390932
Wang W Yoneda M Determination of the optimal penetration factor for evaluating the invasion process of aerosols from a confined source space to an uncontaminated area Sci Total Environ 2020 740 140113 10.1016/j.scitotenv.2020.140113 32559547
Wang Y Wu S Yang Y Yang X Song H Cao Z Huang Y Evaporation and movement of fine droplets in non-uniform temperature and humidity field Build Environ 2019 150 75 87 10.1016/j.buildenv.2019.01.003
Watanabe T Bartrand TA Weir MH Omura T Haas CN Development of a dose-response model for SARS coronavirus Risk Anal 2010 30 7 1129 1138 10.1111/j.1539-6924.2010.01427.x 20497390
Weber DJ, Rutala WA, Sickbert-Bennett EE, Kanamori H, Anderson D (2019) Continuous room decontamination technologies. Am J Infect Control 47s:A72–a78. 10.1016/j.ajic.2019.03.016
Weber TP Stilianakis NI Inactivation of influenza A viruses in the environment and modes of transmission: a critical review J Infect 2008 57 5 361 373 10.1016/j.jinf.2008.08.013 18848358
Webster RG Govorkova EA Continuing challenges in influenza Ann N Y Acad Sci 2014 1323 1 115 139 10.1111/nyas.12462 24891213
Wei J Li Y Enhanced spread of expiratory droplets by turbulence in a cough jet Build Environ 2015 93 86 96 10.1016/j.buildenv.2015.06.018
Wei J, Li Y (2016a) Airborne spread of infectious agents in the indoor environment. Am J Infect Control 44(9, Supplement):S102-S108. 10.1016/j.ajic.2016.06.003
Wei J Li Y Airborne spread of infectious agents in the indoor environment Am J Infect Control 2016 44 9 Suppl S102 108 10.1016/j.ajic.2016.06.003 27590694
Weiser JN The battle with the host over microbial size Curr Opin Microbiol 2013 16 1 59 62 10.1016/j.mib.2013.01.001 23395472
Weiss P Murdoch DR Clinical course and mortality risk of severe COVID-19 Lancet 2020 395 10229 1014 1015 10.1016/s0140-6736(20)30633-4 32197108
Wells WF (1934) On air-borne infection: Study II. Droplets and droplet nuclei. Am J Epidemiol 20(3):611–618. 10.1093/oxfordjournals.aje.a118097
White JF Bernstein DI Key pollen allergens in North America Ann Allergy Asthma Immunol 2003 91 5 425 435 10.1016/s1081-1206(10)61509-8 14692424
White JK Nielsen JL Madsen AM Microbial species and biodiversity in settling dust within and between pig farms Environ Res 2019 171 558 567 10.1016/j.envres.2019.01.008 30771719
Wiegand I Hilpert K Hancock RE Agar and broth dilution methods to determine the minimal inhibitory concentration (MIC) of antimicrobial substances Nat Protoc 2008 3 2 163 175 10.1038/nprot.2007.521 18274517
Wilson JM Platts-Mills TAE Home environmental interventions for house dust mite J Allergy Clin Immunol Pract 2018 6 1 1 7 10.1016/j.jaip.2017.10.003 29310755
Wu W Lin Z Experimental study of the influence of a moving manikin on temperature profile and carbon dioxide distribution under three air distribution methods Build Environ 2015 87 142 153 10.1016/j.buildenv.2015.01.027
Wu X, Fang L, Olesen BW, Zhao J (2014) Air distribution and ventilation effectiveness in a room with floor/ceiling heating and mixing/displacement ventilation. In: Vol. 261 LNEE. Lecture Notes in Electrical Engineering, p 59–67
Wu X Fang L Olesen BW Zhao J Wang F Air distribution in a multi-occupant room with mixing or displacement ventilation with or without floor or ceiling heating Sci Technol Built Environ 2015 21 8 1109 1116 10.1080/23744731.2015.1090255
Wu X Gao J Wang H Fang L Olesen BW Indoor thermal environment and air distribution in a floor-ceiling heating room with mixing or displacement ventilation Sci Technol Built Environ 2019 25 3 346 355 10.1080/23744731.2018.1527138
Wu X Wang H Gao J Wang F Influence of sensible cooling load on indoor air distribution in an office room with ceiling cooling and mixing ventilation Indoor Built Environ 2020 10.1177/1420326X20924819
Wu Y Niu J Liu X Air infiltration induced inter-unit dispersion and infectious risk assessment in a high-rise residential building Build Simul 2018 11 1 193 202 10.1007/s12273-017-0388-6 32218902
Wu Y Tung TCW Niu J Experimental analysis of driving forces and impact factors of horizontal inter-unit airborne dispersion in a residential building Build Environ 2019 151 88 96 10.1016/j.buildenv.2019.01.028
Xiao S Li Y Lei H Lin C-H Norris SL Yang X Zhao P Characterizing dynamic transmission of contaminants on a surface touch network Build Environ 2018 129 107 116 10.1016/j.buildenv.2017.12.015
Xie X Li Y Chwang ATY Ho PL Seto WH How far droplets can move in indoor environments – revisiting the Wells evaporation–falling curve Indoor Air 2007 17 3 211 225 10.1111/j.1600-0668.2007.00469.x 17542834
Xie X, Li Y, Sun H, Liu L (2009) Exhaled droplets due to talking and coughing. J R Soc Interface 6 Suppl 6(Suppl 6):S703–714.10.1098/rsif.2009.0388.focus
Yan Y Li X Ito K Numerical investigation of indoor particulate contaminant transport using the Eulerian-Eulerian and Eulerian-Lagrangian two-phase flow models Exp Comput Multiph Flow 2020 2 1 31 40 10.1007/s42757-019-0016-z 32289121
Yan Y Li X Tu J Thermal effect of human body on cough droplets evaporation and dispersion in an enclosed space Build Environ 2019 148 96 106 10.1016/j.buildenv.2018.10.039 32287988
Yang C Yang X Zhao B Person to person droplets transmission characteristics in unidirectional ventilated protective isolation room: the impact of initial droplet size Build Simul 2016 9 5 597 606 10.1007/s12273-016-0290-7 32218912
Yang J Sekhar SC Cheong KWD Raphael B Performance evaluation of a novel personalized ventilation-personalized exhaust system for airborne infection control Indoor Air 2015 25 2 176 187 10.1111/ina.12127 24810200
Yang JH CFD analysis of the inhaled-air quality for the inpatients in a four-bed sickroom J Asian Archit Build Eng 2013 12 1 109 116 10.3130/jaabe.12.109
Yang Y Wang Y Song B Fan J Cao Y Stability and accuracy of numerical investigation of droplet motion under local ventilation airflow Build Environ 2018 140 32 42 10.1016/j.buildenv.2018.05.023
Yasmeen R Ali Z Afzal N Safdar S Nasir ZA Characterization of bioaerosols in controlled environment broiler houses at different stages of growth JAPS J Anim Plant Sci 2020 30 1 81 91 10.36899/JAPS.2020.1.0010
Ye J Qian H Ma J Zhou R Zheng X Using air curtains to reduce short-range infection risk in consulting ward: a numerical investigation Build Simul 2020 10.1007/s12273-020-0649-7
Yeoh GH, Tu J (2010) Computational techniques for multiphase flows. Butterworth - Heinemann, Amsterdam
Yezli S Otter JA Minimum infective dose of the major human respiratory and enteric viruses transmitted through food and the environment Food Environ Virol 2011 3 1 1 30 10.1007/s12560-011-9056-7 35255645
Yin D Gao Q Zhu H Li J Public perception of urban companion animals during the COVID-19 outbreak in China Health Place 2020 65 102399 10.1016/j.healthplace.2020.102399 32736203
Yin Y Gupta JK Zhang X Liu J Chen Q Distributions of respiratory contaminants from a patient with different postures and exhaling modes in a single-bed inpatient room Build Environ 2011 46 1 75 81 10.1016/j.buildenv.2010.07.003 32288010
Yin Y Wunderink RG MERS, SARS and other coronaviruses as causes of pneumonia Respirology 2018 23 2 130 137 10.1111/resp.13196 29052924
Yin Y, Xu W, Gupta J, Guity A, Marmion P, Manning A, Gulick B, Zhang X, Chen Q (2009) Experimental study on displacement and mixing ventilation systems for a patient ward. HVAC R Res 15(6):1175–1191. 10.1080/10789669.2009.10390885
Yoo JH Choi NY Bae YM Lee JS Lee SY Development of a selective agar plate for the detection of Campylobacter spp. in fresh produce Int J Food Microbiol 2014 189 67 74 10.1016/j.ijfoodmicro.2014.07.032 25126968
Yoo S-J Ito K Assessment of transient inhalation exposure using in silico human model integrated with PBPK-CFD hybrid analysis Sustain Cities Soc 2018 40 317 325 10.1016/j.scs.2018.04.023
You S Wan MP Particle concentration dynamics in the ventilation duct after an artificial release: For countering potential bioterrorist attack J Hazard Mater 2014 267 183 193 10.1016/j.jhazmat.2013.12.058 24447859
Yu ITS, Li Y, Wong TW, Tam W, Chan AT, Lee JHW, Leung DYC, Ho T (2004) Evidence of airborne transmission of the severe acute respiratory syndrome virus. N Eng J Med 350(17):1731–1739. 10.1056/NEJMoa032867
Zayas G Chiang MC Wong E MacDonald F Lange CF Senthilselvan A King M Cough aerosol in healthy participants: fundamental knowledge to optimize droplet-spread infectious respiratory disease management BMC Pulm Med 2012 12 1 11 10.1186/1471-2466-12-11 22436202
Zhai ZJ, Metzger ID (2019) Insights on critical parameters and conditions for personalized ventilation. Sustain Cities Soc 48:101584. 10.1016/j.scs.2019.101584
Zhai ZJ Zhang Z Zhang W Chen QY Evaluation of various turbulence models in predicting airflow and turbulence in enclosed environments by CFD: Part 1—summary of prevalent turbulence models HVAC&R Res 2007 13 6 853 870 10.1080/10789669.2007.10391459
Zhang B, Guo G, Zhu C, Ji Z, Lin C-H (2020a) Transport and trajectory of cough-induced bimodal aerosol in an air-conditioned space. Indoor Built Enviro 30(9):1546–1567. 10.1177/1420326x20941166
Zhang C, Pomianowski M, Heiselberg PK, Yu T (2020b) A review of integrated radiant heating/cooling with ventilation systems- thermal comfort and indoor air quality. Energy Build 223:110094. 10.1016/j.enbuild.2020.110094
Zhang N Chen W Chan PT Yen HL Tang JW Li Y Close contact behavior in indoor environment and transmission of respiratory infection Indoor Air 2020 30 4 645 661 10.1111/ina.12673 32259319
Zhang R Li Y Zhang AL Wang Y Molina MJ Identifying airborne transmission as the dominant route for the spread of COVID-19 Proc Natl Acad Sci 2020 117 26 14857 10.1073/pnas.2009637117 32527856
Zhang Y Feng G Bi Y Cai Y Zhang Z Cao G Distribution of droplet aerosols generated by mouth coughing and nose breathing in an air-conditioned room Sustain Cities Soc 2019 51 101721 10.1016/j.scs.2019.101721
Zhang Z Chen Q Comparison of the Eulerian and Lagrangian methods for predicting particle transport in enclosed spaces Atmos Environ 2007 41 25 5236 5248 10.1016/j.atmosenv.2006.05.086
Zhang Z Zhang W Zhai ZJ Chen QY Evaluation of various turbulence models in predicting airflow and turbulence in enclosed environments by CFD: Part 2—comparison with experimental data from literature HVAC&R Res 2007 13 6 871 886 10.1080/10789669.2007.10391460
Zhao B Zhang Z Li X Numerical study of the transport of droplets or particles generated by respiratory system indoors Build Environ 2005 40 8 1032 1039 10.1016/j.buildenv.2004.09.018 32287997
Zheng XH, Qian H, Liu L (2011) Numerical study on a new personalized ventilation system application in cross infection prevention. J Cent South Univ 42(12):3905–39114307
Zhou Q, Gao N, Qian H (2014) CFD study on the wind-induced transmission of gaseous pollutants between flats in multistory residential buildings. Paper presented at the Indoor Air 2014, Hong Kong, China
Zhou Y Deng Y Wu P Cao S-J The effects of ventilation and floor heating systems on the dispersion and deposition of fine particles in an enclosed environment Build Environ 2017 125 192 205 10.1016/j.buildenv.2017.08.049
Zhu S, Jenkins S, Addo K, Heidarinejad M, Romo SA, Layne A, Ehizibolo J, Dalgo D, Mattise NW, Hong F, Adenaiye OO, Bueno de Mesquita JP, Albert BJ, Washington-Lewis R, German J, Tai S, Youssefi S, Milton DK, Srebric J (2020) Ventilation and laboratory confirmed acute respiratory infection (ARI) rates in college residence halls in College Park, Maryland. Environ Int 137:105537. 10.1016/j.envint.2020.105537
Zhu S Kato S Yang J-H Study on transport characteristics of saliva droplets produced by coughing in a calm indoor environment Build Environ 2006 41 12 1691 1702 10.1016/j.buildenv.2005.06.024
| 36467894 | PMC9703444 | NO-CC CODE | 2022-11-29 23:21:09 | no | Air Qual Atmos Health. 2022 Nov 28;:1-57 | utf-8 | Air Qual Atmos Health | 2,022 | 10.1007/s11869-022-01286-w | oa_other |
==== Front
Hematol Transfus Cell Ther
Hematol Transfus Cell Ther
Hematology, Transfusion and Cell Therapy
2531-1379
2531-1387
Published by Elsevier Editora Ltda.
S2531-1379(22)01248-2
10.1016/j.htct.2022.09.1133
Article
RAZÕES HEMATOLÓGICAS NA COVID-19: DIFERENÇAS ENTRE PACIENTES COM E SEM NECESSIDADE DE VENTILAÇÃO MECÂNICA INVASIVA
Fernandes NF a
Costa IF a
Ciceri ACM a
Pereira KN ab
Carvalho JAM a
Paniz C a
a Universidade Federal de Santa Maria (UFSM), Santa Maria, RS, Brasil
b Hospital Universitário de Santa Maria, Universidade Federal de Santa Maria (HUSM-UFSM), Santa Maria, RS, Brasil
15 10 2022
10 2022
15 10 2022
44 S660S660
Copyright © 2022 Published by Elsevier Editora Ltda.
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Objetivos
Investigar as diferenças na Razão Neutrófilo-Linfócito (NLR), Razão Neutrófilo-Linfócito derivada (d-NLR), Razão Linfócito-Monócito (LMR), Razão Plaquetas-Linfócitos (PLR), Razão Neutrófilos-Plaquetas (NPR) e Índice de Inflamação Sistêmica (SII) entre pacientes COVID-19 com e sem necessidade de ventilação mecânica invasiva e um grupo controle saudável.
Material e Métodos
Foram incluídos pacientes com diagnóstico de COVID-19 admitidos no Hospital Universitário de Santa Maria (HUSM) no período de 1°de março de 2020 a 31 de março de 2021. Foram excluídos pacientes com falta de informações nos prontuários, com câncer, internados por acidentes graves, com lúpus, apendicite, transferidos para outros hospitais, gestantes e menores de 18 anos, resultando em 212 pacientes. Foi incluído um grupo controle com 198 indivíduos saudáveis. Os dados do primeiro hemograma após internação foram obtidos do prontuário eletrônico do hospital. A partir desses dados, calculou-se: Razão Neutrófilo-Linfócito (NLR), razão neutrófilo-linfócito derivada (d-NLR, divisão do total de neutrófilos por leucócitos menos neutrófilos totais), Razão Linfócito-Monócito (LMR), Razão Plaquetas-Linfócitos (PLR), razão neutrófilos-plaquetas (NPR) e índice de inflamação sistêmica (SII, multiplicação de plaquetas por neutrófilos totais seguida de divisão por linfócitos totais). Os pacientes foram estratificados em com necessidade de Ventilação Mecânica Invasiva (VMI) (n=129) e sem Necessidade de Ventilação Mecânica Invasiva (NVMI) (n=83). O estudo foi aprovado pelo Comitê de Ética em Pesquisa da UFSM (CAAE 30917320.5.0000.5346).
Resultados
As razões calculadas apresentaram as seguintes medianas e intervalos interquartis para grupo controle, NIMV e IMV respectivamente: NLR 1,75 (1,42‒2,24), 6,58 (3,29‒12,7), 14,5 (8,10‒23,2); d-NLR 1,28 (1,07‒1,64), 4,00 (2,23‒7,33), 8,09 (4,81‒12,89); LMR 3,89 (3,22‒5,02), 2,00 (1,25‒3,07), 1,53 (1,00‒2,73); PLR 101 (84‒119), 227 (159‒400), 310 (207‒441); NPR 0,02 (0,01‒0,02), 0,02 (0,02‒0,04), 0,04 (0,03‒0,06); SII 387 (302‒522), 1748 (739,6‒2705), 3115 (1429‒5769). Todas as razões calculadas apresentaram resultados significativamente diferentes (p<0,001) entre o grupo controle e o grupo de pacientes com COVID-19. NLR, d-NLR, NPR e SII apresentaram diferença significativa entre grupo controle, VMI e NVMI.
Discussão
O hemograma é um exame realizado rotineiramente na maioria dos laboratórios, portanto as razões hematológicas representam parâmetros alternativos que não agregam custo e não requerem análises adicionais. Estas razões já demonstraram diferença em pacientes com outras doenças e no caso da COVID-19 para outros parâmetros, como mortalidade, por exemplo. Nosso trabalho demonstrou que existe diferença significativa entre pacientes com e sem necessidade de intubação, o que ainda não havia sido descrito anteriormente.
Conclusão
Embora mais estudos sejam necessários, os índices hematológicos abordados neste trabalho são potenciais auxiliares na tomada de decisão clínica quanto à necessidade de ventilação mecânica invasiva.
==== Body
pmc
| 0 | PMC9703626 | NO-CC CODE | 2022-11-29 23:21:11 | no | Hematol Transfus Cell Ther. 2022 Oct 15; 44:S660 | utf-8 | Hematol Transfus Cell Ther | 2,022 | 10.1016/j.htct.2022.09.1133 | oa_other |
==== Front
Hematol Transfus Cell Ther
Hematol Transfus Cell Ther
Hematology, Transfusion and Cell Therapy
2531-1379
2531-1387
Published by Elsevier Editora Ltda.
S2531-1379(22)01284-6
10.1016/j.htct.2022.09.1169
Article
AVALIAÇÃO DA EXPRESSÃO DE STATS 1 E 3 EM PACIENTES COM COVID-19
Rocha BGA
Cunha ACCH
Pereira LQ
Vito FB
Carneiro ACDM
Tanaka SCSV
Souza HM
Universidade Federal do Triângulo Mineiro (UFTM), Uberaba, MG, Brasil
15 10 2022
10 2022
15 10 2022
44 S680S681
Copyright © 2022 Published by Elsevier Editora Ltda.
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Objetivos
Avaliar a expressão dos genes STAT 1 e 3 em pacientes com COVID-19 em relação ao desfecho, gravidade e comorbidades.
Materiais e Métodos
Participaram desse estudo 89 pacientes diagnosticados com COVID-19, atendidos em hospitais do município de Uberaba – MG, no período de maio a outubro de 2020. Os pacientes foram agrupados quanto ao desfecho (alta ou óbito), gravidade da doença (leve, moderada ou grave) e presença ou ausência de comorbidades. A expressão relativa de STATs 1 e 3 foi realizada por PCR em tempo real. O gene ACTB foi utilizado como controle endógeno. Para análise estatística das variáveis numéricas foi aplicado o teste de normalidade Kolmogorov-Smirnov e as comparações estatísticas entre dois grupos foram realizadas pelo teste Mann-Whitney. Nas análises superiores a dois grupos foi aplicado o teste Kruskal-Wallis. A significância estatística foi definida como p<0,05.
Resultados
A expressão de STAT1 não foi significante nos pacientes com COVID-19 em relação as variáveis analisadas (p>0,05). No entanto, a expressão de STAT3 foi estatisticamente significativa em relação a gravidade (p=0,03), sendo mais expressa nos pacientes moderados, em relação aos casos leves (p=0,04). Não houve diferença na expressão de STAT3 em relação ao desfecho (p=0,64) e presença de comorbidades (p=0,09).
Discussão
A sinalização da via JAK/STAT é responsável por regular diversas funções imunes e a inibição da sinalização desta via tem sido sugerida na infecção pelo SARS-CoV-2, visto que a mesma está relacionada a inflamação excessiva desencadeada pela tempestade de citocinas. Em nosso estudo não foi observada associação em relação ao desfecho, gravidade e presença de comorbidades e a expressão de STAT1. No entanto, um estudo realizado na Alemanha demonstrou maior expressão desta proteína em pacientes com COVID-19. Além disso, analises do fluido broncoalveolar de macrófagos de pacientes com COVID-19 grave apresentaram aumento da expressão de fatores de transcrição, incluindo STAT1, bem como fatores reguladores de interferon, sugerindo um microambiente inflamatório nesse grupo de pacientes. Nesse sentido, a expressão de STAT1 demonstrou variações em relação ao pulmão e sangue periférico. No presente estudo foi observado o aumento da expressão de STAT3 nos casos moderados, mas não nos casos graves, podendo indicar uma função reguladora para STAT3, visto que a maioria dos indivíduos se recuperaram. Recentemente, foi relatado o envolvimento de STAT3 na patogênese da COVID-19 como um potencializador da resposta inflamatória.
Conclusão
O presente estudo conclui que não há diferença na expressão de STAT1 em relação ao desfecho, gravidade e presença de comorbidades. A expressão de STAT3 foi maior em indivíduos com doença moderada, indicando um possível papel deste gene no mecanismo de controle para progressão da doença.
==== Body
pmc
| 0 | PMC9703831 | NO-CC CODE | 2022-11-29 23:21:14 | no | Hematol Transfus Cell Ther. 2022 Oct 15; 44:S680-S681 | utf-8 | Hematol Transfus Cell Ther | 2,022 | 10.1016/j.htct.2022.09.1169 | oa_other |
==== Front
Ann Epidemiol
Ann Epidemiol
Annals of Epidemiology
1047-2797
1873-2585
Elsevier Inc.
S1047-2797(22)00300-3
10.1016/j.annepidem.2022.11.007
Brief Communication
Estimating the impact of the COVID-19 pandemic on rising trends in drug overdose mortality in the United States, 2018-2021
Lee Hyunjung PhD, MS, MPP, MBA a⁎
Singh Gopal K. PhD, MS, MSc, DPS b
a Department of Public Policy and Public Affairs, John McCormack Graduate School of Policy and Global Studies, University of Massachusetts Boston, Boston, MA
b The Center for Global Health and Health Policy, Global Health and Education Projects, Inc., Riverdale, MD
⁎ Corresponding author: Department of Public Policy and Public Affairs, John McCormack Graduate School of Policy and Global Studies, University of Massachusetts Boston, 100 William T Morrissey Blvd, Boston, MA 02125.
28 11 2022
1 2023
28 11 2022
77 8589
25 8 2022
21 11 2022
23 11 2022
© 2022 Elsevier Inc. All rights reserved.
2022
Elsevier Inc.
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Purpose
During the COVID-19 pandemic, social and economic disruption such as social isolation, job and income losses, and increased psychological distress, may have contributed to the increase in drug-overdose mortality. This study aims to measure the impact of the pandemic on monthly trends in drug-overdose mortality in the United States.
Methods
We used the 2018–2020 final and 2021 provisional monthly deaths from the National Vital Statistics System and monthly population estimates from the Census Bureau to compute monthly mortality rates by age, sex, and race/ethnicity. We use log-linear regression models to estimate monthly percent increases in mortality rates from January 2018 through November 2021.
Results
The age-adjusted drug-overdose mortality rate among individuals aged older than or equal to 15 years increased by 30% between 2019 (70,459 deaths) and 2020 (91,536 deaths). During January 2018–November 2021, the monthly drug-overdose mortality rate increased by 2.05% per month for Blacks, 2.25% for American Indians/Alaska Natives, 1.96% for Hispanics, 1.33% for Asian/Pacific Islanders, and 0.96% for non-Hispanic Whites. Average monthly increases in mortality were most marked among those aged 15–24 and 35–44 years.
Conclusions
The COVID-19 pandemic had a substantial impact on the rising trends in drug-overdose mortality during the peak months in 2020 and 2021.
Keywords
Drug overdose mortality
COVID-19 pandemic
Monthly trend
Race/ethnicity
Age-specific
==== Body
pmcIntroduction
Since 1999, more than 932,000 people have died due to drug overdose in the United States [1]. The age-adjusted death rate involving synthetic opioid (other than methadone), which largely consist of illicitly manufactured fentanyl, increased more than eleven fold, from 1.0 per 100,000 population in 2013 to 11.4 in 2019 [2]. In 2018, in 28 states and DC, fentanyl was detected in 73.9% of opioid-involved overdose deaths [3]. The COVID-19 pandemic, declared on March 11, 2020 [4], and subsequent disruptions in treatment, combined with increased psychological distress and social and economic stressors including social isolation and unemployment, may have fueled the opioid epidemic and rise in drug overdose mortality [5], [6], [7], [8]. During the pandemic, drug overdose [9], [10], [11], [12] and opioid-related overdose deaths [13], [14], [15], [16], [17], [18], [19], [20] have significantly increased. Median emergency department (ED) visits for all drug overdoses in the United States increased from 13,371 (711.1/100,000 ED visits) in 2019–15,604 (940.2/100,000 ED visits) in 2020 [12]. In 2020, 91,799 drug overdose deaths occurred, compared with 70,630 drug overdose deaths in 2019 [1,21]. The age-adjusted drug-overdose mortality rate increased by 31% between 2019 (21.6/100,000 population) and 2020 (28.3/100,000 population) [1].
There had been differences in drug overdose mortality by sex, age group, and race and ethnicity before the pandemic [21], [22], [23], [24]. The drug overdose mortality rate increased from 8.2 per 100,000 in 1999 to 29.6 in 2019 for males, and from 3.9 in 1999 to 13.7 in 2019 for females [21]. In 2019, adults aged 35–44 had the highest rate of drug overdose deaths (40.5/100,000), followed by adults aged 45–54, 25–34, 55–64, 15–24, and 65 and over [21]. Age-adjusted drug overdose mortality was highest among non-Hispanic Whites (25.9/100,000), followed by Blacks (23.7/100,000), American Indian and Alaska Natives (AIANs) (19.2/100,000), Hispanics (12.7/100,000), and lowest among Asian and Pacific Islanders (APIs, 4.1/100,000) in 2019 [22]. The average rate of increase in drug overdose mortality during 1999–2017 was the fastest for non-Hispanic Whites (7.6%/y), followed by AIANs (6.1%), APIs (5.9%), Blacks (3.6%), and Hispanics (3.3%) [23].
During the COVID-19 pandemic, drug overdose mortality increased more for males, than females [14,15]. There have been heterogeneous findings on differential effect of COVID-19 by age and race and ethnicity [[13], [14], [15],20]. Given disparities in social determinants of health, the pandemic might have disproportionately affected drug overdose mortality rates among racial and ethnic minorities [23,25]. One study found that racial and ethnic minorities experienced higher increases in drug overdose mortality than non-Hispanic Whites [20]. However, another study found that opioid overdose deaths continuously increased only for Blacks during March 2020 to March 2021 in Massachusetts, while it remained stable for AIANs, Asians, and Hispanics [16]. The heterogeneity of results among studies on the COVID-19 effect by age and race/ethnicity might arise from study period and regional differences.
This study extends previous analyses by estimating monthly percent changes in drug overdose mortality rates per 1 million population for U.S. adults aged 15 years and older during January 2018 through November 2021 by age, sex, and race/ethnicity. We also identified and analyzed peak periods in monthly drug overdose mortality during the pandemic.
Methods
The 2018–2020 final and 2021 provisional monthly deaths by age, sex, race/ethnicity, and cause of death were obtained from the National Vital Statistics System (NVSS)’s mortality files. NVSS data are collected and disseminated by the CDC's National Center for Health Statistics (NCHS) through the registration system, in which individual States and independent registration areas including the District of Columbia, New York City, and five territories are responsible for the registration of vital events - births, deaths, marriages, divorces, and fetal deaths [26]. Recent annual mortality datasets have included about 2.5 million records, which is based on the 2003 revision of the U.S. Standard Certificates and Reports [27]. Since final annual mortality data for a given year are typically released 11 months after the end of the calendar year [28], we used provisional mortality data from January to November, 2021, which were the latest data available from the NVSS [19,20,28]. The 2018–2020 monthly population estimates by age, sex, and race/ethnicity were obtained from the Census Bureau [29]. Monthly population estimates for 2021 were derived based on the annual growth rate of the population from 2020 to 2021.
We used deaths records for individuals aged 15 and older in the NVSS from January 2018 to November 2021. Drug overdose mortality rates were calculated by dividing the number of drug overdose deaths by the corresponding population and reported as deaths per 1,000,000 population. Drug overdose deaths were classified using the International Classification of Diseases, 10th Revision (ICD-10) underlying cause-of-death codes: X40-44 (unintentional), X60-64 (suicide), X85 (homicide), Y10-Y14 (undetermined intent) [21].
We estimated monthly trends in drug overdose mortality to give researchers and policymakers an early indication of shifts in mortality trends and to provide actionable information sooner than final mortality data [28]. We used log-linear regression models to estimate monthly percent increases in mortality rates from January 2018 through November 2021 by modeling the logarithm of the mortality rates as a linear function of time (month), which yielded monthly exponential rates of change in mortality rates. The estimated monthly percent changes can be used for public health surveillance to assess trends in the rate of change, providing how fast drug mortality rates increase or decrease. We computed age-adjusted drug overdose mortality rates per 1,000,000 population by month for the overall population, males, and females. Monthly drug overdose mortality rates were age adjusted by the direct method using the age distribution of the 2000 U.S. Standard Population. Standard errors of age-adjusted mortality rates were computed, and sex-, age-, race/ethnicity-specific mortality rate ratios were calculated. Statistical significance in group differences in monthly trend by sex, age group, and race/ethnicity were tested using a Hausman test after the seemingly unrelated estimation, suest, the estimations from all subgroups to be pooled together [30] for statistical significance. All analyses were conducted by Stata 17 [31].
Results
The drug-overdose deaths among individuals aged older than or equal to 15 years increased by 30% between 2019 (70,459 deaths) and 2020 (91,536 deaths). Figure 1 shows the monthly trend in drug overdose mortality per million population by age, sex, and race/ethnicity from January 2018 through November 2021. Average monthly drug overdose mortality among adult aged 35–44 was the highest, followed by the rates among adults aged 45–54, 25–34, and 55–64; individuals aged 15–24 and 65+ had lower rates than other groups (Fig. 1A). Average monthly drug mortality rates were higher for males than for females (Fig. 1B). Between February 2020 and May 2020, during the first peak of drug overdose mortality shown in the figures, the age-adjusted drug-overdose mortality rate increased by 48% overall, 52% for males, and 39% for females, while during the second peak of drug overdose mortality between October/November 2020 and March/April 2021, the increase in the drug overdose mortality rate was higher for females (34%) than for males (29%). During the first peak, drug-overdose mortality rates increased most rapidly for individuals aged 15-24 (63%), followed by individuals aged 25–34 (58%), 35–44 (52%), 45–54 (41%), 55–64 (36%), and 65+ (17%).Fig. 1 (A–C), Monthly drug overdose mortality rates per 1 million population among adults aged 15+ years by age, sex, and Race/Ethnicity, United States, January 2018 through November 2021. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)
Fig 1
Average monthly drug overdose mortality rates were highest for non-Hispanic Whites before the pandemic, followed by Blacks, AIANs, Hispanics, and APIs (Fig. 1C). During the pandemic, racial and ethnic minorities experienced a marked increase in drug overdose mortality. Between February 2020 and May 2020, the drug-overdose mortality rate increased by 52% for Blacks, 51% for AIANs, 49% for APIs, compared with 45% for non-Hispanic Whites and Hispanics. During the second peak between October/November 2020 and March/April 2021, API (53%) and AIANs (50%) showed higher increases in mortality, while other racial/ethnic groups experienced a 25%–40% increase in drug overdose mortality.
Table 1 shows estimated monthly percent changes in drug overdose mortality rates per one million population for U.S. adults aged 15 or older, using log-linear regression analysis. During January 2018–November 2021, the average monthly rate of increase in age-adjusted drug overdose mortality was 1.28% for adults aged 15 or older, 1.40% for males and 1.01% for females. Adults aged 15–24 (1.47%) and 35–44 (1.44%) had a faster average monthly increase in drug overdose mortality than other age groups. We found marked racial/ethnic disparities in the pace of monthly increases in drug overdose mortality. The monthly drug-overdose mortality rate increased by 2.05% per month for Blacks, 2.25% for AIANs, 1.96% for Hispanics, 1.33% for APIs, and 0.96% for non-Hispanic Whites.Table 1 Estimated monthly percent changes (EMPC) in age, sex, and race-specific drug overdose mortality rates per 1 million population for U.S. adults aged 15 years and older, January 2018 through November 2021
Table 1 Age 15+* Age 15–24 Age 25–34 Age 35–44 Age 45–54 Age 55–64 Age 65 or older Male* Female* NHW Black AIAN API Hispanic
EMPC (95% CI) 1.278 (1.084, 1.473) 1.468 (1.148, 1.790) 1.126 (0.897, 1.355) 1.442 (1.248, 1.637) 1.168 (0.970, 1.367) 1.258 (1.094, 1.422) 1.122 (0.957, 1.286) 1.401 (1.201, 1.602) 1.014 (0.829, 1.200) 0.961 (0.779, 1.144) 2.045 (1.829, 2.262) 2.251 (1.928, 2.576) 1.329 (0.978, 1.681) 1.955 (1.726, 2.185)
Slope† (SE) 0.013 (0.001) 0.015 (0.002) 0.011 (0.001) 0.014 (0.001) 0.012 (0.001) 0.013 (0.001) 0.011 (0.001) 0.014 (0.001) 0.010 (0.001) 0.010 (0.001) 0.020 (0.001) 0.022 (0.002) 0.013 (0.002) 0.019 (0.001)
Intercept 2.976 2.072 3.277 3.333 3.279 3.054 1.747 3.269 2.565 3.112 2.936 2.593 1.182 2.308
R-Square 0.789 0.645 0.676 0.826 0.749 0.836 0.801 0.808 0.721 0.705 0.886 0.808 0.552 0.863
% Increase (Feb-May 2020) 43.13 62.79 58.23 52.36 40.98 36.31 16.55 52.18 39.01 45.01 52.46 50.99 49.20 45.26
Rate ratio NA 3.79 3.52 3.16 2.48 2.19 1.00 1.00 0.75 1.00 1.17 1.13 1.09 1.01
⁎ For overall adults aged 15+, males, and females, we calculated monthly age-adjusted drug overdose mortality rate using the 2000 US standard population
† All slopes were statistically significant at P < .001 using log-linear regression
Notes. AIAN = American Indian and Alaska Native; CI = confidence interval; EMPC=EXP (slope)*100-100; NHW = Non-Hispanic White; SE = standard error.
Discussion
In this study, we found that COVID-19 disproportionately increased monthly drug overdose deaths among males, younger age groups, and racial and ethnic minorities. We also found that the rate of increase in drug overdose mortality during the pandemic was highest among the youngest group, that is, individuals aged 15–24. Our findings on age-group differential effects of the pandemic were consistent with the previous finding on the increase in opioid overdose deaths in Nevada that was mostly attributed to deaths among individuals aged 10–29 [13]. However, our findings differed from previous findings that there was no significant difference in the impact of COVID-19 on opioid overdose deaths by age group during 2017 and 2020 in Ohio [15], or that adults aged 30–40 and 50–60 were most affected in Milwaukee [14].
Our finding regarding higher rates of increase in drug overdose mortality among males compared to females are consistent with previous findings [14,15]. Our findings based on racial and ethnic stratified models are consistent with the previous studies which showed that Blacks, Hispanics, and AIANs experienced higher increases in drug overdose mortality than non-Hispanic Whites during the pandemic [15,20]. For example, a recent study of annual trends found that drug overdose mortality increased for non-Hispanic Blacks from 24.7 per 100,000 population in 2019–36.8 in 2020 (a 48.8% increase), for AIANs from 28.9 in 2019 to 41.4 in 2020 (a 43.3% increase), and for Hispanics from 12.4 in 2019 to 17.3 in 2020 (a 40.1% increase), compared with rates for non-Hispanic Whites from 25.0 in 2019 to 31.6 in 2020 (a 26.3% increase) [20].
Drug overdose mortality, in our study, showed the greatest increase from February to May 2020 and the greatest decrease from May 2020 to October/November 2020. The decreasing trend might be explained by implementation of temporary policies by federal and state governments to improve access to substance use disorder treatment. For example, the Centers for Medicare & Medicaid Services allowed telehealth reimbursement for Medicare beneficiaries under the 1135 waiver authority and Coronavirus Preparedness and Response Supplemental Appropriations Act on March 6, 2020 [32]. Some states required telehealth visits with Medicaid or private insurance to be paid at the same rate as analogous in-person visits, allowed for behavioral telehealth with waving requirements on prior in-person contact, relaxed state privacy laws, and allowed NPs and PA to obtain buprenorphine prescribing waivers [33,34]. Further studies are needed to estimate the impact of each intervention on access to care and drug mortality changes.
In our study, drug overdose mortality remained the highest among Blacks since August 2019 and the mortality rates among AIANs surpassed and remained higher than Whites since March 2021. Racial/ethnic disparities in drug overdose deaths might be explained by existing socioeconomic and health disparities among racial/ethnic groups and structural racism, in conjunction with COVID-19 [23,25]. For example, non-Hispanic Blacks and AIANs were more likely to experience job-related income loss, food insecurity, and depression than non-Hispanic Whites during the pandemic [7,35,36]. In particular, Blacks were less likely to have mitigation effects on mental health through government interventions such as stimulus fund or unemployment insurance [35]. Given higher rates of illicit drug use among Blacks and AIANs than Whites [37]. and lack of access to medications for opioid use disorder in short-term residential treatment [38]. it is important to implement appropriate policy interventions on prevention, treatment, and recovery support services to address underlying causes of racial and ethnic health disparities and to stem the rising tide in drug overdose deaths among U.S. adults.
Increased drug overdose mortality among youths might be explained by increase in use of illicitly manufactured fentanyl due to low cost, fentanyl contamination, and high potency [2,39]. Harm reduction intervention among young adults such as fentanyl test strips might be helpful in changing overdose risk behavior [40]. During the pandemic, adolescents might have experienced increased psychological stress and substance use disorders (SUDs) through social isolation due to school closure and social distancing, and their transitional stage of life course with psychiatric vulnerability [41]. According to Substance Abuse and Mental Health Services Administration, only about 3.5% of adolescents aged 12–17 and 3.7% of youth aged 18–25 received SUD treatment at a specialty facility in 2020, compared with 7.4% of adults aged 26 or older among those who needed substance use treatment [42]. Considering that the COVID-19 pandemic might have increased SUDs, policymakers should address an existing lack of access to SUD treatment among adolescents and youth through interventions such as expanding telehealth services including medication for opioid use disorder [43]. Further studies need to focus on adolescents and young adults to evaluate the impact of the pandemic on their drug overdose, treatment, and mortality.
Limitations
This study has limitations. Provisional death counts might be underestimated relative to final counts since provisional counts are often incomplete and causes of death may be pending investigation [28,44]. Future studies might consider adopting the predicted provisional death counts, adjusted for delayed reporting [44]. and comparing with the final death counts when data are available. Moreover, there might have been misclassification of AIANs, Asians, and Hispanics on death certificates and in provisional mortality statistics, possibly leading to an underreporting of racial/ethnic disparities in mortality rates [28]. In particular, more than 40% of AIANs who self-identified as AIANs were misclassified as Whites on the death certificate, and correction for death certificate misclassification increased mortality rate ratios of AIANs to Whites [45]. Finally, although our study calculated drug overdose deaths by combining unintentional, suicide, homicide, and undetermined deaths to obtain enough deaths for monthly trends, the trends might differ for specific causes of drug overdose mortality. The proportion of unintentional deaths due to drug overdose was 87%–88% during 2018–2019 and increased to 91%–92% during 2020–2021. The proportion of suicide deaths due to drug overdose was 7% during 2018–2019 and decreased to 4%–5% during 2020–2021. Future studies might consider analyzing trends in drug overdose mortality from specific causes during the pandemic.
Conclusions
The COVID-19 pandemic had a disproportionate impact on racial and ethnic minorities and individuals in younger age groups in the context of rising trends in drug-overdose mortality during the peak months in 2020 and 2021. Drug-overdose mortality rates increased faster among racial and ethnic minorities compared to non-Hispanic Whites.
CRediT authorship contribution statement
Hyunjung Lee: Conceptualization, Methodology, Software, Formal analysis, Data curation, Visualization, Validation, Writing – original draft, Writing – review & editing. Gopal K. Singh: Conceptualization, Methodology, Validation, Writing – review & editing.
Acknowledgments
Funding: None
No IRB approval was required for this study, which is based on the secondary analysis of a public-use federal database.
No potential conflicts of interest relevant to this article were reported.
==== Refs
References
1 Centers for Disease Control and Prevention. Drug overdose deaths remain high. https://www.cdc.gov/drugoverdose/deaths/index.html. Published 2022, Accessed July 5, 2022.
2 Mattson CL Tanz LJ Quinn K Kariisa M Patel P Davis NL. Trends and geographic patterns in drug and synthetic opioid overdose deaths—United States, 2013–2019 Morb Mortal Wkly Rep 70 6 2021 202 207
3 O'Donnell J Gladden RM Goldberger BA Mattson CL Kariisa M. Notes from the field: opioid-involved overdose deaths with fentanyl or fentanyl analogs detected - 28 States and the District of Columbia, July 2016-December 2018 MMWR Morb Mortal Wkly Rep 69 10 2020 271 273 10.15585/mmwr.mm6910a4 32163382
4 World Health Organization (WHO). WHO Director-General's opening remarks at the media briefing on COVID-19 - 11 March 2020. https://www.who.int/director-general/speeches/detail/who-director-general-s-opening-remarks-at-the-media-briefing-on-covid-19-11-march-2020. Published 2020, Accessed October 1, 2022.
5 Haley DF Saitz R. The opioid epidemic during the COVID-19 pandemic JAMA 324 16 2020 1615 1617 32945831
6 Davis CS Samuels EA. Opioid policy changes during the COVID-19 pandemic-and beyond J Addict Med 14 4 2020 e4 e5 32433363
7 Singh GK Lee H Azuine RE. Growing job-related income losses, increasing social inequalities, and physical and mental health impact during the COVID-19 pandemic, United States, April–December 2020 Int J Transl Med Res Public Heal 5 2 2021 76 89
8 Lee H Singh GK. Monthly trends in self-reported health status and depression by race/ethnicity and socioeconomic status during the COVID-19 Pandemic, United States, April 2020 - May 2021 Ann Epidemiol 63 2021 52 62 10.1016/j.annepidem.2021.07.014 34358622
9 Glober N Mohler G Huynh P Arkins T O’Donnell D Carter J Impact of COVID-19 pandemic on drug overdoses in Indianapolis J Urban Heal 97 6 2020 802 807
10 Ochalek TA Cumpston KL Wills BK Gal TS Moeller FG. Nonfatal opioid overdoses at an urban emergency department during the COVID-19 pandemic JAMA 324 16 2020 1673 1674 32945832
11 Macmadu A Batthala S Gabel AMC Rosenberg M Ganguly R Yedinak JL Comparison of characteristics of deaths from drug overdose before vs during the COVID-19 pandemic in Rhode island JAMA Netw open 4 9 2021 e2125538-e2125538
12 Holland KM Jones C Vivolo-Kantor AM Idaikkadar N Zwald M Hoots B Trends in US emergency department visits for mental health, overdose, and violence outcomes before and during the COVID-19 pandemic JAMA Psychiatry 78 4 2021 372 379 33533876
13 Garcia G-GP Stringfellow EJ DiGennaro C Poellinger N Wood J Wakeman S Opioid overdose decedent characteristics during COVID-19 Ann Med 54 1 2022 1081 1088 35467475
14 Ghose R Forati AM Mantsch JR. Impact of the COVID-19 pandemic on opioid overdose deaths: a spatiotemporal analysis J Urban Heal 99 2 2022 316 327
15 Vieson J Yeh AB Lan Q Sprague JE. During the COVID-19 pandemic, opioid overdose deaths revert to previous record levels in Ohio J Addict Med 16 2 2022 e118 34172625
16 Liao C-Y Garcia G-GP DiGennaro C Jalali MS. Racial disparities in opioid overdose deaths in Massachusetts JAMA Netw Open 5 4 2022 e229081 -e229081
17 Khatri UG Pizzicato LN Viner K Racial/ethnic disparities in unintentional fatal and nonfatal emergency medical services–attended opioid overdoses during the COVID-19 pandemic in Philadelphia JAMA Netw open 4 1 2021 e2034878 -e2034878
18 Auty SG Griffith KN. Medicaid expansion and drug overdose mortality during the COVID-19 pandemic in the United States Drug Alcohol Depend 232 2022 109340
19 Friedman J Akre S. COVID-19 and the drug overdose crisis: uncovering the deadliest months in the United States, January‒July 2020 Am J Public Health 111 7 2021 1284 1291 33856885
20 Friedman J Hansen H. Evaluation of increases in drug overdose mortality rates in the US by race and ethnicity before and during the COVID-19 pandemic JAMA Psychiatry 79 4 2022 379 381 35234815
21 Hedegaard H Miniño AM Warner M. Drug Overdose Deaths in the United States, 1999–2019 NCHS Data Brief, No 394 2020 National Center for Health Statistics Hyattsville, MD
22 Centers for Disease Control and Prevention. National Vital Statistics System, Mortality 1999-2020 on CDC WONDER Online Database, released in 2021, Data are from the Multiple Cause of Death Files, 1999-2020, as compiled from data provided by the 57 vital statistics jurisdictions through the Vital St. http://wonder.cdc.gov/ucd-icd10.html. Published 2021. Accessed May 10, 2022.
23 Singh GK Kim IE Girmay M Perry C Daus GP Vedamuthu IP Opioid epidemic in the United States: empirical trends, and a literature review of social determinants and epidemiological, pain management, and treatment patterns Int J Matern Child Heal AIDS 8 2 2019 89 100
24 Cano M Sparks CS. Drug overdose mortality by race/ethnicity across US-born and immigrant populations Drug Alcohol Depend 232 2022 109309
25 US Department of Health and Human Services. Health Resources and Services Administration. Office of Health Equity. Health Equity Report 2019-2020: Special Feature on Housing and Health Inequalities. Rockville, Maryland; 2020.
26 National Center for Health Statistics. About the National Vital Statistics System. https://www.cdc.gov/nchs/nvss/about_nvss.htm. Published 2016. Accessed May 2, 2022.
27 Ventura SJ. The U.S. National Vital Statistics System: Transitioning Into the 21st Century, 1990-2017 Vital Health Stat 1 62 2018 1 84 30248018
28 Ahmad F Cisewski JA Anderson RN. Provisional mortality data—United States, 2021 MMWR Morb Mortal Wkly Rep 71 2022 597 600 10.15585/mmwr.mm7117e1 35482572
29 US Census Bureau. 2019 Monthly National Population Estimates by Age, Sex, Race, Hispanic Origin, and Population Universe for the United States: April 1, 2010 to December 1, 2020 (NC-EST2019-ALLDATA). https://www.census.gov/data/tables/time-series/demo/popest/2010s-national-detail.html. Published 2022, Accessed April 18, 2022.
30 Fiebig DG. Seemingly unrelated regression Baltagi BH A Companion to Theoretical Econometrics 2001 Blackwell Malden, MA 101 121
31 StataCorp. Stata Statistical Software: Release 17. 2021.
32 Centers for Medicare & Medicaid Services. Medicare telemedicine health care provider fact sheet.
33 Pessar SC Boustead A Ge Y Smart R Pacula RL. Assessment of state and federal health policies for opioid use disorder treatment during the COVID-19 pandemic and beyond JAMA Heal forum 2 11 2021 e213833 10.1001/jamahealthforum.2021.3833
34 Henry BF Campbell A Hunt T Johnson JK Mandavia AD Chaple M COVID-19 related substance use services policy changes: Policymaker perspectives on policy development & implementation J Subst Abuse Treat 133 2022 108550 10.1016/j.jsat.2021.108550
35 Lee H Singh GK. Food insecurity-related interventions and mental health among US adults during the COVID-19 Pandemic, April 2020 through August 2021 Public Health Rep 137 6 2022 1187 1197 10.1177/00333549221110294 35993183
36 Lee H Singh GK. Racial and ethnic disparities in the mental health impact of the COVID-19 pandemic in the United States: analysis of 3 million Americans from the National Health Interview Survey and Household Pulse Survey Int J Transl Med Res Public Heal 6 1 2022 e417
37 Center for Behavioral Health Statistics and Quality. Racial/ethnic differences in substance use, substance use disorders, and substance use treatment utilization among people aged 12 or older (2015-2019). Rockville, MD; 2021. https://www.samhsa.gov/data/. Accessed October 9, 2022.
38 Stahler GJ Mennis J Baron DA. Racial/ethnic disparities in the use of medications for opioid use disorder (MOUD) and their effects on residential drug treatment outcomes in the US Drug Alcohol Depend 226 2021 108849
39 US Drug Enforcement Administration. Facts about Fentanyl.
40 Krieger MS Yedinak JL Buxton JA Lysyshyn M Bernstein E Rich JD High willingness to use rapid fentanyl test strips among young adults who use drugs Harm Reduct J 15 1 2018 7 10.1186/s12954-018-0213-2 29422052
41 Sarvey D Welsh JW. Adolescent substance use: Challenges and opportunities related to COVID-19 J Subst Abuse Treat 122 2021 108212
42 Substance Abuse and Mental Health Services Administration. Key Substance Use and Mental Health Indicators in the United States: Results from the 2020 National Survey on Drug Use and Health (HHS Publication No. PEP21-07-01-003, NSDUH Series H-56). Rockville, MD; 2021.
43 Davis CS Samuels EA. Continuing increased access to buprenorphine in the United States via telemedicine after COVID-19 Int J Drug Policy 93 2021 102905 10.1016/j.drugpo.2020.102905
44 Ahmad F, Rossen L, Sutton P. Provisional drug overdose death counts. https://www.cdc.gov/nchs/nvss/vsrr/drug-overdose-data.htm. Published 2021, Accessed May 4, 2022.
45 Arias E Heron M Hakes J. The validity of race and hispanic-origin reporting on death certificates in the United States: an update Vital Health Stat 2 172 2016 1 21
| 36455852 | PMC9703855 | NO-CC CODE | 2022-12-16 23:18:16 | no | Ann Epidemiol. 2023 Jan 28; 77:85-89 | utf-8 | Ann Epidemiol | 2,022 | 10.1016/j.annepidem.2022.11.007 | oa_other |
==== Front
J Diabetes Complications
J Diabetes Complications
Journal of Diabetes and Its Complications
1056-8727
1873-460X
Elsevier Inc.
S1056-8727(22)00275-6
10.1016/j.jdiacomp.2022.108363
108363
Article
Effect of the first and second COVID-19 associated lockdown on the metabolic control of patients with type 2 diabetes in Greece
Papachristoforou Eleftheria b⁎
Liatis Stavros b
Psoma Ourania a
Kountouri Aikaterini c
Lambadiari Vaia c
Tsimihodimos Vasilis a
a Department of Internal Medicine, School of Medicine, University of Ioannina, Ioannina, Greece
b First Department of Propaedeutic Internal Medicine, National and Kapodistrian University of Athens Medical School, Athens, Greece
c 2nd Department of Internal Medicine, Research Institute and Diabetes Center, National and Kapodistrian University of Athens Medical School, Attikon University Hospital, Athens, Greece
⁎ Corresponding author.
28 11 2022
28 11 2022
10836330 9 2022
15 11 2022
25 11 2022
© 2022 Elsevier Inc. All rights reserved.
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
The aim of this study was to assess the effect of the COVID-19 lockdown periods on the metabolic control of patients with type 2 diabetes (T2D) in three academic diabetes centers in Greece. There was a slight improvement in BMI, blood pressure and lipid values while the remaining parameters remained stable.
Keywords
Covid-19
lockdown
type 2 diabetes
==== Body
pmc1 Introduction
Coronavirus disease 19 (Covid-19) results from the very contagious severe acute respiratory syndrome coronavirus 2 (SARS-CoV 2) and was classified as a pandemic in a short period of time.1 Governments from many countries implemented several times the measure of lockdown, with movement restrictions and social distancing, in order to limit the spread of the disease and to protect their health systems.2 The implementation of lockdown has affected routine health care activities and everyday life. All outpatient activities were cancelled and there was also difficulty in accessibility to medications.3
We have previously reported, among others, that the first lockdown, implemented during the first wave of the pandemic, had neutral or even beneficial effects on metabolic control of patients with T2D.4., 5., 6.
Nevertheless, the repeated use of restrictive measures may have had an adverse impact on physical and psychological health of people with chronic metabolic disorders. Τhe aim of this observational study was to retrospectively examine the effect of the first and second COVID-19 associated lockdown on the metabolic control of patients with T2D in three academic centers in Greece.
2 Materials and methods
2.1 Study design
Patients with T2D receiving regular follow-up (every 3–6 months) at the Οutpatient Diabetes Clinics of three academic centers in Greece, were enrolled. Clinical and laboratory data were retrospectively collected from the available clinical files. Demographics and clinical data were collected at three time points:1) on the last patient's visit before (up to 3 months) the first lockdown, which occurred between March 16th and April 30th 2020, 2) on the first visit after the first lockdown (up to 3 months) and 3) on the first visit after (up to 3 months) the second lockdown which occurred between November 5th and May 14th, 2021. The protocol was aligned with the principles of the Declaration of Helsinki and was approved by the hospital's Ethics Committee.
2.2 Statistical analysis
Continuous variables with normal distribution are presented as mean ± one-standard deviation, whereas those lacking normality are presented as median ± interquartile range. Qualitative variables are presented as absolute and relative frequencies (%). Analysis of variance for repeated measures and Bonferroni adjustment was used for the comparisons between the studied parameters at the three time points. Data were analyzed using the Statistical Package SPSS, version 21.0 (SPSS Inc., Chicago, IL).
3 Results
The present analysis included342 participants with available data (mean age 62 ± 9.2 years, 62 % males, mean duration of diabetes 14.64 ± 9.52 years). Concerning the antidiabetic treatment of the participants, 85.9 % were receiving metformin, 30.3 % SGLT-2 inhibitors, 41.6 % GLP-1 receptor agonists and 53.2 % were receiving insulin (Reviewer 2, Comment 1). Most patients had good glycemic control (HbA1c: 6.9 ± 1.3 %), blood pressure (SBP and DBP 138 ± 18 and 80 ± 11 mmHg, respectively) and lipid profile (LDL-C 83 ± 30 mg/dl) at baseline. After the first lockdown, a small but statistically significant reduction in BMI and total cholesterol values was observed. There was a non-significant further decline in BMI after the second lockdown. These findings persisted after adjustment for baseline treatment with GLP-1 receptor agonists (p = 0.005 for BMI) and/or SGLT-2 inhibitors (P = 0.006 for BMI) (Reviewer 2, Comment 2). After the second lockdown, systolic blood pressure, total cholesterol, triglycerides and LDL-cholesterol values, were significantly lower than at baseline, whereas only total cholesterol showed a significant decrease compared to post-first lockdown period (Table 1 ).Table 1 Effect of lockdowns on clinical and laboratory variables in patients with type 2 diabetes.
Table 1 Baseline After 1st lockdown After 2nd lockdown P
BMI (kg/m2) 30.6 ± 5.8 30.3 ± 5.6a 30 ± 5.7a 0.006
Systolic blood pressure (mmHg) 138 ± 19 137 ± 16 135 ± 17a 0.019
Diastolic blood pressure (mmHg) 80 ± 12 82 ± 10 79 ± 10 0.260
Fasting plasma glucose (mg/dl) 127 ± 42 123 ± 35 125 ± 35 0.255
HbA1c (%) 6.9 ± 1.3 6.7 ± 0.9 6.7 ± 1 0.168
Total cholesterol (mg/dl) 160 ± 35 154 ± 33a 150 ± 31a, b <0.001
Triglycerides (mg/dl) 138 ± 74 132 ± 67 129 ± 63a 0.011
HDL-cholesterol (mg/dl) 49 ± 13 49 ± 13 49 ± 13 0.516
LDL-cholesterol (mg/dl) 82 ± 29 80 ± 30 76 ± 26a <0.001
Alanine aminotransferase (IU/l) 23 ± 14 22 ± 13 23 ± 15 0.739
Aspartate aminotransferase (IU/l) 20 ± 9 20 ± 9 22 ± 10 0.105
Gamma-gloutamyltransferase (IU/l) 23 ± 12 22 ± 12 21 ± 13 0.513
a Compared to baseline.
b Compared to post-first lockdown values.
4 Discussion
In this study, we found that during two consecutive lockdowns, patients with T2D exhibited small, statistically significant but clinically negligible reductions in ΒΜΙ, systolic blood pressure and lipid values. It might be argued, however, that even non-deterioration of complication-related risk factors in type 2 diabetes is a clinically meaningful observation. There are conflicting data in the literature concerning the impact of covid-19 lockdown on metabolic parameters of patients with T2D. In a meta-analysis of 36 observational studies, including patients with type 1 and type 2 diabetes, it was shown that during the first lockdown period, there was an improvement in mean glucose and glucose variability in patients with type 1 diabetes, whereas in people with type 2 diabetes there were no significant changes in glycemic control.7 In another recent systematic review and meta-analysis, including only patients with type 2 diabetes, it was demonstrated that during the first lockdown period the levels of HbA1c significantly increased.8 Ιn another study by B. Hansel et al. regarding the impact of the Covid-19 lockdown on overweight/obese patients with T2D, a reduction in body weight was demonstrated in a significant proportion of the participants.9
It has been suggested that during lockdown periods daily activities slowed down and many patients had the opportunity to work from home. Home confinement may have led to better eating behavior and a chance for more leisure-time exercise.10 Such lifestyle favorable changes may explain the results of the present study. Additionally, the potential improvement in treatment compliance and the psychological support from other family members could have a beneficial effect on metabolic control of patients with type 2 diabetes (Reviewer 2, Comment 3).
Limitations of our study include a) the fact that it was a retrospective study. Clinical and laboratory data were retrospectively collected from the available medical files; b) the hypolipidemic and antihypertensive treatment of the participants were not included in our database and thus we cannot exclude that the observed improvement in lipid values and systolic blood pressure readings could be due to treatment intensification; c) the socioeconomic status and the profession of the participants could have an impact on the results, but this remains to be elucidated within larger cohorts. (Reviewer 1, Comment 2, Reviewer 2, Comment 4).
In conclusion, in a sample of patients withT2D followed in academic outpatient diabetes clinics, consecutive lockdowns related to the COVID-19 pandemic did not seem to deteriorate metabolic control, whereas some slight improvement was noticed in BMI and lipid profile.
CRediT authorship contribution statement
All authors have read and agreed to the published version of the manuscript.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Declaration of competing interest
The authors declare no conflict of interest.
==== Refs
References
1. World Health Organization WHO coronavirus (COVID-19) dashboard | WHO coronavirus disease (COVID-19) dashboard https://covid-19.who.int/
2. Desvars-Larrive A. Dervic E. Haug N. Niederkrotenthaler T. Chen J. Di Natale A. A structured open dataset of government interventions in response to COVID-19 Sci Data 7 2020 285 10.1038/s41597-020-00609-9 32855430
3. Singhai K. Swami M.K. Nebhinani N. Psychological adaptive difficulties and their management during COVID-19 pandemic in people with diabetes mellitus Diabetes Metab Syndr 14 2020 1603 1605 10.1016/j.dsx.2020.08.025 32862099
4. D'Onofrio L. Pieralice S. Maddaloni E. Effects of the COVID-19 lockdown on glycaemic control in subjects with type 2 diabetes: the glycalock study Diabetes ObesMetab 23 2021 1624 1630 10.1111/dom.14380
5. Fernández E. Cortazar A. Bellido V. Impact of COVID-19 lockdown on glycemic control in patients with type 1 diabetes Diabetes Res Clin Pract 166 2020 108348 10.1016/j.diabres.2020.108348 Aug
6. Psoma O. Papachristoforou E. Kountouri A. Effect of COVID-19- associated lockdown on the metabolic control of patients with type 2 diabetes J Diabetes Complications 34 2020 10.1016/j.jdiacomp.2020.107756 Dec
7. Silverii G.A. Poggi C.D. Dicembrini I. Monami M. Mannucci E. Glucose control in diabetes during home confinement for the first pandemic wave of COVID-19: a meta-analysis of observational studies Acta Diabetol. 58 2021 1603 1611 10.1007/s00592-021-01754-2 Dec 34159476
8. Ojo O. Wang X.-H. Ojo O.O. Orjih E. The effects of COVID-19 lockdown on glycaemic control and lipid profile in patients with type 2 diabetes: a systematic review and meta-analysis Int J Environ Res Public Health 19 2022 1095 10.3390/ijerph19031095 Jan 19 35162117
9. Hansel B. Potier L. Chalopin S. Larger E. The COVID-19 lockdown as an opportunity to change lifestyle and body weight in people with overweight/obesity and diabetes: results from the national French COVIDIAB cohort NutrMetabCardiovascDis 31 2021 2605 2611 10.1016/j.numecd.2021.05.031
10. Ahola A.J. Mutter S. Forsblom C. Harjutsalo V. Groop P.-H. Meal timing, meal frequency, and breakfast skipping in adult individuals with type 1 diabetes – associations with glycaemic control Sci Rep 9 2019 20063 10.1038/s41598-019-56541-5 31882789
| 0 | PMC9703856 | NO-CC CODE | 2022-11-29 23:21:14 | no | J Diabetes Complications. 2022 Nov 28;:108363 | utf-8 | J Diabetes Complications | 2,022 | 10.1016/j.jdiacomp.2022.108363 | oa_other |
==== Front
Int J Infect Dis
Int J Infect Dis
International Journal of Infectious Diseases
1201-9712
1878-3511
The Author(s). Published by Elsevier Ltd on behalf of International Society for Infectious Diseases.
S1201-9712(22)00619-1
10.1016/j.ijid.2022.11.027
Article
The application of a novel 5-in-1 multiplex RT-PCR assay for rapid detection of SARS-CoV-2 and differentiation between variants of concern
Chung Hsing-Yi ae
Jian Ming-Jr a
Chang Chih-Kai a
Chen Chi-Sheng a
Li Shih-Yi a
Lin Jung-Chung b
Yeh Kuo-Ming b
Yang Ya-Sung b
Chen Chien-Wen c
Hsieh Shan-Shan a
Tang Sheng-Hui a
Perng Cherng-Lih a
Hung Kuo-Sheng d
Chang Feng-Yee b
Shang Hung-Sheng a#
a Division of Clinical Pathology, Department of Pathology, Tri-Service General Hospital, National Defense Medical Center, Taipei, Taiwan, ROC
b Division of Infectious Diseases and Tropical Medicine, Department of Medicine, Tri-Service General Hospital, National Defense Medical Center, Taipei, Taiwan, ROC
c Division of Pulmonary and Critical Care Medicine, Department of Medicine, Tri-Service General Hospital, National Defense Medical Center, Taipei, Taiwan, ROC
d Center for Precision Medicine and Genomics, Tri-Service General Hospital, National Defense Medical Center, Taipei, Taiwan, ROC
e Graduate Institute of Medical Science, National Defense Medical Center, Taipei, Taiwan
# Address correspondence to Hung-Sheng Shang, Division of Clinical Pathology, Department of Pathology, Tri-Service General Hospital, National Defence Medical Center, Taipei, Taiwan, ROC, No. 161, Sec. 6, Minquan E. Rd., Neihu Dist., Division of Clinical Pathology, TSGH, NDMC Taipei City 11490, Taiwan (ROC)
28 11 2022
28 11 2022
5 6 2022
24 10 2022
23 11 2022
© 2022 The Author(s). Published by Elsevier Ltd on behalf of International Society for Infectious Diseases.
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Objectives
To rapid detect for the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and immediately distinguish whether positive samples represent variants of concern (VOCs), we established a novel 5-in-1 VOC assay.
Methods
This assay could distinguish among five VOCs: Alpha, Beta, Gamma, Delta, and Omicron, in a single reaction tube. The five variants exhibit different single nucleotide polymorphisms (SNPs) in their viral genome, which can be exploited to distinguish them. We selected target SNPs in the spike gene, including N501Y, P681R, K417N, and deletion H69/V70 for the assay.
Results
The limit of detection of each gene locus was 80 copies per PCR reaction. We observed a high consistency among the results when comparing the performance of our 5-in-1 VOC assay, whole gene sequencing, and the Roche VirSNiP SARS-CoV-2 test in retrospectively analyzing 150 clinical SARS-CoV-2 variant positive samples. The 5-in-1 VOC assay offers an alternative and rapid high-throughput test for most diagnostic laboratories in a flexible sample-to-result platform.
Conclusion
The assay can also be applied in a commercial platform with completion of the SARS-CoV-2 confirmation test and identification of its variant within 2.5 hours.
Keywords
SARS-CoV-2 variant
variant of concerns
COVID-19
High-throughput
Omicron
==== Body
pmcIntroduction
The first severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) variant Alpha was identified in England in late November 2021(Choi and Smith, 2021; Galloway et al., 2021). Mutations in the viral genome can increase transmissibility, facilitate escape from the human immune system, and/or alter biologically important phenotypes in a way that confers a fitness advantage to the virus, such as mutations in the spike (S) gene that affect antigenicity (Davies et al., 2021; Petersen et al., 2022). The SARS-CoV-2 variants are classified as variants being monitored (VBMs), variants of interest (VOIs), variants of concern (VOCs), and variants of high consequence (VOHCs) (Ramesh et al., 2021). There are two currently circulating SARS-CoV-2 VOCs (Delta and Omicron), and three previously circulating VOCs (Alpha, Beta, and Gamma) (https://www.cdc.gov/coronavirus/2019-ncov/variants/variant-classifications.html).
Viral genomic mutations leading to new variants of SARS-CoV-2 are a real challenge in tackling the global coronavirus disease (COVID-19) pandemic (Chang et al., 2022; Chen et al., 2022; Seong et al., 2021). Viruses that are mutated have a high propensity for replication errors, providing them with an advantage; several mutations further occur with each replication cycle, making such viruses more virulent and transmissible (Al-Tawfiq et al., 2022). Each characterized variant has mutations in the gene encoding the S protein (Omicron: 30 mutations; Delta: 10 mutations; Alpha: 7 mutations and 2 deletions; Beta: 9 mutations and 1 deletion; and Gamma: >10 mutations), compared to the sequence of the wild type index virus (Wuhan-Hu-1) (Kumar et al., 2022).
Since different variants will continue to emerge, identifying positive patients and monitoring their variants are important in the epidemic prevention policy. Currently, whole-genome sequencing (WGS) analysis of positive samples can be used to confirm infection with a specific variant and characterize to the variant (Shaibu et al., 2021). WGS is a relatively complex method that can take from hours to several days to generate results. Analysis of Sanger-sequencing results of the viral S gene is an alternative strategy that can be used to classify the viral lineage of positive samples(Daniels et al., 2021). However, WGS and sanger sequencing of S protein methods are time-consuming methods and require well-trained laboratory personnel and suitable instruments (Park and Kim, 2016).
Therefore, a more rapid response to identifying emerging SARS-CoV-2 variants is required (Challen et al., 2021). We developed a novel, 5-in-1 VOC assay, that can monitor key single nucleotide polymorphisms (SNPs) of spike gene in the S gene, thus identifying SARS-CoV-2 positive samples and distinguishing among the specific variants. It could offer a correct information of SARs-CoV-2 variants of concerns for the public prevention response to COVID-19 pandemic.
Materials and methods
Study design and clinical samples
We retrospectively examined the stored residual viral transport medium of nasopharyngeal/throat swab samples, collected from patients between May 2021 and March 2022. A total of 150 positive samples were individually confirmed by detecting the nucleocapsid (N1) and envelope (E) genes of SARS-CoV-2 using quantitative real-time reverse transcription PCR (RT-PCR), based on a cycle threshold (Ct) value <35, which follows the Taiwan CDC guideline (Perng et al., 2020; Rabaan et al., 2021). The positive sample need to be re-confirmed by another SARS-CoV-2 RT-PCR system. This retrospective study was registered on February 8, 2021. and approved by the Institutional Review Board of the Clinical and Genomic Research Committee at the Tri-Service General Hospital (approval no.: C202005041). The extracted RNA of positive samples were also further detected for their SARS-CoV-2 lineage by VirSNiP SARS-CoV-2 mutation assays (TIB Molbiol, Berlin, Germany) for confirmed the variant (Ong et al.,2021). Those clinical sample were put into our lab-deveolped 5-in-1 VOC assay on those sample-to-result plateform. Simultaneously, the samples also were detected four single nucleotide polymorphisms (SNPs) of SARS-CoV-2 Spike genes in one PCR reaction for distinguishing the SARS-CoV-2 variants(ECDC, 20 December 2021). We also desired a novel 5-in-1 VOC assay, which could have confirmed the sample with the SARS-CoV-2 by N2 gene (Chung et al., 2021).
Novel 5-in-1 VOC assay design
To design the 5-in-1 VOC assay, SARS-CoV-2 genome sequences were downloaded from the GISAID database. According to the database, the selected SNP loci in the S gene could be used to distinguish between Alpha, Beta, Gamma, Delta, and Omicron variants (Figure 1 ). Primers and probes were designed using the consensus sequences obtained from the sequence alignment of the SARS-CoV-2 S protein (Table 1 ). In our design, we selected four key SNPs of S protein: N501Y, P681R, K417N, and deletion H69/V70. We also tested for the presence of N2 to confirm that the samples were positive for SARS-CoV-2, following the US CDC guidelines (Lu et al., 2020). The 5-in-1 VOC reagent comprised all primers and probes along with the Luna One-Step RT-PCR Kit components (New England Biolabs). The novel 5-in-1 VOC assay was used with two sample-to-result open platforms, the BD MAX™ System and the LabTurbo™ AIO (LabTurbo, New Taipei City, Taiwan), to determine clinical efficacy.Figure 1 Detailed diagram of SNPs that were used to distinguish the Omicron, Delta, Alpha, Beta, and Gamma variants of SARS-CoV-2.
Figure 1
Table 1 Primer and probe sequences used in this study
Table 1Protein region Fluorescence SNP mutation Sequence name Primer sequence 5′->3′
Spike FAM N501Y N501Y-F TGTTACTTTCCTTTACAATCATATGGTTTC
N501Y-R GAAAGTACTACTACTCTGTATGGTTGGTAACC
N501Y-P 5’-/56-FAM/CCAACCCAC/ZEN/TTATGGTGTTGG/3lABkFQ/-3’
HEX P681R P681R-F CCCATTGGTGCAGGTATATG
P681R-R TAGTGTAGGCAATGATGGATTGA
P681R-P 5HEX/ACTCAGACT/ZEN/AATTCTCGTCG/3lABkFQ/-3’
Taxus Red 69-70 Del69-70-F TCAACTCAGGACTTGTTCTTAC
Del69-70-R TGGTAGGACAGGGTTATCAAAC
Del69-70-P 5’-/5TexRd-XN/GTCCCAGAGACATGTATAGCAT/3lABkFQ/-3’
Cy5 K417N K417N-F TGCAGATTCATTTGTAATTAGAGG
K417N-R ATAACGCAGCCTGTAAAATCATC
K417N-P 5’-/5Cy5/GCAAACTGG/TAO/AAATATTGCT/3lAbRQSp/-3’
Nucleocapsid Cy5.5 N2-F TTACAAACATTGGCCGCAAA
N2-R GCGCGACATTCCGAAGAA
N2-P 5’-/5Cy55/ACAATTTGC/TAO/CCCCAGCGCTTCAG
/3lAbRQSp/-3’
RT-PCR detection of SARS-CoV-2 VOCs
The samples were screened using the two sample-to-result platforms with our novel 5-in-1 VOC assay. Samples consisted of viral transport media (VTM) collected form patients. For the BD MAX™ System, 500 μL of sample was added to the sample buffer tube provided with the BD MAX ExK TNA-3 kit. Subsequently, a 5-in-1 VOC PCR reagent was added at position #3 of the snap-in tubes and the extracted RNA was automatically blended in the same position. RT-PCR was then carried out using the following conditions: reverse transcription at 55°C for 10 min and initial denaturation at 99°C for 2 min, followed by 45 cycles at 99°C for 10 second and 58°C for 24 sec. A positive result for the target gene was indicated by the presence of an RT-PCR amplification curve and the associated Ct value.
For the LabTurbo AIO 48 SP-qPCR automation system, RNA was extracted from 500 μL of input sample using the LabTurbo Viral DNA/RNA Mini kit (Cat. No. LVX480-500). The extracted RNA was then mixed with PCR reagent and PCR was carried out using the following conditions: reverse transcription at 55°C for 8 min and initial denaturation at 95°C for 1 min, followed by 45 cycles at 95°C for 10 sec and at 58°C for 20 sec. A positive result for the target gene was indicated by a Ct value smaller than 35.
Evaluation of analytical sensitivity of the 5-in-1 VOC assay with analytical validation using RNA controls
To the 5-in-1 VOC assay, we used RNA controls (Vircell, Spain) of SARS-CoV-2 variants for absolute quantification. The controls included: AMPLIRUN® SARS CoV-2 B.1.1.7 RNA CONTROL, AMPLIRUN® SARS CoV-2 B.1.351 (South African variant) RNA CONTROL, AMPLIRUN® SARS CoV-2 P.1 RNA CONTROL, AMPLIRUN® SARS CoV-2 DELTA (B.1.617.2) RNA CONTROL and AMPLIRUN® SARS CoV-2 OMICRON RNA CONTROL. Serial dilutions of the control samples (20, 40, 80, 160, 320, and 640 copies/PCR reaction) were prepared using nuclease-free water to assess the limit of detection (LoD), which was defined as a 95% probability of positive replicates.
Evaluation of the 5-in-1 VOC assay specificity
The specificity of the 5-in-1 VOC assay was evaluated against several common upper respiratory tract viruses, including influenza A, influenza B, rhinovirus, respiratory syncytial virus, parainfluenza virus, adenovirus, and coronavirus HKU1. Samples testing positive for these viruses were obtained from clinical samples of the Taiwan CDC viral infection contract laboratory. The samples were analyzed using the 5-in-1 VOC assay on the LabTurbo AIO open platform.
Whole-genome sequencing of SARS-CoV-2
Whole-genome sequences of the SARS-CoV-2 isolates (hcoV-19/TSGH-39 to TSGH-69) were obtained by following the Illumina TruSeq Stranded mRNA Library Prep Kit protocol, which enriched for SARS-CoV-2 cDNA using multiplex RT-PCR amplicons. The Ovation RNA-Seq System V2 (Nugen Technologies, San Carlos, CA, USA) was used to synthesize cDNA, which was then processed into a library for WGS which was performed on the NovaSeq 6000 platform (Illumina, San Diego, CA, USA) (Charre et al., 2020). Paired-end read assemblies of the whole virus genome sequence were formed using SPAdes assembler with SARS-CoV-2 isolate Wuhan-Hu-1, complete genome (NC_045512.2) to run the genome-guide assembly pipeline.
Comparison of the performance using clinical specimens
To verify the performance of 5-in-1 VOC assay, we used 150 SARS-CoV-2 variant positive samples. Those variants were classified by VirSNiP SARS-CoV-2 mutation assays (TIB Molbiol, Berlin, Germany), which used real-time RT-PCR post-melting curve analysis to detect mutations targeting specific spike protein variations. According to the analysis, there are (N501Y, P681R, H69/V70 and K417N). The retrospective samples were detected by 5-in-1 VOC assay on two sample-to-result platforms: BD MAX™ System and LabTurbo™ AIO. Thus, a result was considered positive if the amplification curve crossed the threshold line within 35 cycles (Ct value < 35) and that was defined as a positive result for SNPs mutation gene. Following the patterns of the mutation gene, the assay could distinguish the SARS-CoV-2 variant according the table 2 and figure 2 .Table 2 Results of the 5-in-1 VOC assay according to the US CDC classification of SARS-CoV-2 variants of concern (VOC)
Table 2 Pango lineage 5-in-1 VOC assay
N2 (FAM) P681R (HEX) H69/V70 (Texas Red) K417N (Cy5) N501Y (Cy5.5)
Alpha B.1.1.7 + del +
Beta B.1.351 + + +
Gamma P.1 + +
Delta B.1.617.2 + +
Omicron BA.1 B.1.1.529 + del + +
Omicron BA.2 B.1.1.529 + + +
+, positive signal in the 5-in-1 VOC assay.
Figure 2 Overview of COVID-19 VOC target position and assay designed concepts. Five probes targeted in spike (S) protein including D69/70, K417N, N501Y, and P681R were labelled with Texas Red, Cy5, Cy5.5, and HEX. One probe which targets N2 position in Nucleocapsid (N) was labelled with FAM. According to the combination of each probe, using qPCR could characterize different COVID-19 variants in one reaction. Cycle in x-axis is the number of cycles and Rn in y-axis is the reporter fluorescence signal in qPCR.
Figure 2
RESULTS
Epidemiological features of the patients with the Omicron variant
Clinical information was retrieved for 120 patients infected by the SARS-CoV-2 Omicron variant reported between December 2021 and March 2022. We identified the isolate to be the Omicron variant by using VirSNiP SARS-CoV-2 test kits. After aligning the complete genome sequences of the isolated RNAs against the RNA sequences derived from four SARS-CoV-2 variants (hcoV-19/TSGH-48, hcoV-19/TSGH-49, hcoV-19/TSGH-51, and hcoV-19/TSGH-52)—and associated metadata—hosted on the publicly available database GIASID, we identified the isolate to be the Omicron variant.
A summary of the gender, ages, and epidemiological features of the 120 patients is shown in Table 3 . The average age and medium age were 39 and 35 years old, respectively. All 120 patients had a survival rate of 100%, and 34.2% (41/120) of the Omicron-positive patients were asymptomatic. The top two symptoms associated with infected patients were fever (24.2%) and cough (44.2%). There only 1.7% of patients with the Omicron variant exhibited diarrhea and 0.8% manifested a change in taste.Table 3 Clinical features, symptoms, and outcomes of patients with COVID-19 caused by the Omicron variant of SARS-CoV-2
Table 3Characteristics SARS-CoV-2 Omicron
Total number 120
Gender
Male 61 (50.8%)
Female 59 (49.2%)
Age
<19 8 (6.7%)
19-29 31 (25.8%)
30-39 24 (20.0%)
40-49 13 (10.8%)
50-59 14 (11.7%)
60-69 11 (9.2%)
>69 19 (15.8%)
Mean 44.0
Median (± SD) 36.5 ± 21.76
Vaccination
2 doses, no., (%) 52 (43.3%)
Brand AstraZeneca 16 (13.3%)
Pfizer-BioNTech 24 (20.0%)
CoronVac 1 (0.8%)
Moderna 11 (9.2%)
3 doses, no., (%) 47 (39.2%)
Brand Pfizer-BioNTech 25 (20.8%)
Moderna 22 (18.3%)
Symptoms
No symptoms 25 (20.8%)
Fever 27 (22.5%)
Cough 48 (40.0%)
Sore throat 31 (25.8%)
Runny nose 12 (10.0%)
Stuffy nose 5 (4.2%)
Headache 3 (2.5%)
General weakness 3 (2.5%)
Body soreness 4 (3.3%)
GI upset 1 (0.8%)
Shortness of breath 13 (10.8%)
Conscious change 24 (20.0%)
Seizure attack 3 (2.5%)
WHO severity classification
Mild 60 (50.0%)
Moderate 19 (15.8%)
severe 16 (13.3%)
Omicron sub lineage
Omicron BA.1 52 (43.3%)
Omicron BA.2 68 (56.7%)
Ct value
Mean 18.86
Median (SD) 17 ± 5.98
5-in-1 VOC genotyping assay
SARS-CoV-2 Variants of concern (VOCs) or variants under monitor (VUMs) corresponded to five dominant SARS-CoV-2 variants (Delta-lineage B.1.617.2, Alpha-lineage B.1.1.7, Beta-lineage B.1.351, Gamma Beta-lineage P.1 and Omicron-lineage B.1.1.529) whose information is hosted on GIASIAD, which contains the sequence information of more than 4,700,000 isolates identified since 2019 (Figure 1). Differences based on the single nucleotide polymorphisms (SNPs) in the Spike region—Wuhan-hu-1—identified using the five-in-one assay could be used to distinguish the variants. According to the information hosted on GASID, most of the omicron variants carry the N501Y and K417N mutations, and the P681R mutation could separate Delta from other variants. Our 5-in-1 assay, could quickly separate the specimens form Alpha, Beta, Gamma, Delta and Omicron. Moreover, this assay could distinguish omicron-BA.1 and omicron sneath-BA.2 based on deletion H69/V70 (supplementary figure1).
Analytical sensitivity of 5-in-1 VOC assay
The LoD of the 5-in-1 VOC assay was determined to be 80 copies/PCR reaction SARS-CoV-2 variant RNA control by testing 20 replicates of serially diluted variant controls (Table 4 ). This LoD is the same as that obtained using the BD MAX™ and LabTurbo AIO open platforms. According to the results for each gene in the 5-in-1 VOC assay, each variant could be distinguished as depicted in Table 2. For example: when we analyzed 20 Omicron replicate samples at 80 copies RNA control per PCR reaction, we obtained a positive result for N2, N501Y, K417N, and a negative result for P681R and deletion H69/V70. This is still a good performance in the context of the other SARS-CoV-2 variant RNAs.Table 4 Assessment of the limit of detection for the 5-in-1 VOC assay on the BD MAXTM System and LabTurbo AIO System using serially diluted RNA controls.
Table 4Testing
platform Target gene/fluoresce No. of replicates testing positive/total no. of replicates tested (%) at the indicated control copies per PCR reaction
320b 160 80 40 20
BD MAXTM N501Y/FAM 20/20 (100) 20/20(100) 20/20 (100) 19/20 (95) 7/20 (35)
P681R/HEX 20/20 (100) 20/20 (100) 20/20 (100) 18/20 (90) 6/20 (30)
69-70/Texas Red 20/20 (100) 20/20(100) 20/20 (100) 18/20 (90) N.D.
K417N/Cy5 20/20 (100) 20/20(100) 20/20 (100) 17/20 (85) N.D.
N2/Cy5.5 20/20 (100) 20/20(100) 19/20 (95) 9/20 (45) N.D.
LabTurbo AIO N501Y/FAM 20/20 (100) 20/20(100) 20/20 (100) 19/20 (95) 8/20 (40)
P681R/HEX 20/20 (100) 20/20 (100) 20/20 (100) 18/20 (90) 6/20 (30)
69-70/Texas Red 20/20 (100) 20/20(100) 20/20 (100) 17/20 (85) N.D.
K417N/Cy5 20/20 (100) 20/20(100) 20/20 (100) 17/20 (85) N.D.
N2/Cy5.5 20/20 (100) 20/20(100) 19/20 (95) 14/20 (70) N.D.
aTargets included SARS-CoV-2 spike (S) gene SNPs or the N2 gene.
bTested samples included 20 replicates of SARS-CoV-2 variant controls that were serially diluted from 320 to 20 copies/PCR reaction.
cN.D., not detected.
Analytical specificity of the 5-in-1 VOC assay
The analytical specificity of the 5-in-1 VOC assay on the LabTurbo AIO open platform was determined by testing a panel of clinical samples containing the following pathogens: influenza A, influenza B, rhinovirus, enterovirus, parainfluenza virus subtype 1 to 3, adenovirus, and coronavirus HKU1, as well as COVID-19-negative samples. No cross-reactivity with other respiratory viruses was detected using the 5-in-1 VOC assay (Supplementary Table 1).
Clinical performance of the 5-in-1 VOC assay
We analyzed 150 clinically positive SARS-CoV-2 samples in this study, confirming the SARS-CoV-2 positive results of the Taiwan CDC central laboratory. We identified the samples to be the Alpha, Betta, Delta and Omicron variant by using VirSNiP SARS-CoV-2 mutation assay. The samples were retrospectively analyzed using the 5-in-1 VOC assay on both the BD MAX™ System and the LabTurbo™ AIO open platforms. The results of these analyses demonstrated 100% agreement with the results of the VirSNiP SARS-CoV-2 test kits (Figure 3 ). To confirm that the results were consistent with those of rapid detection of SARS-CoV-2 VOCs, presumptive cases were also assessed by WGS, and the lineages were confirmed using GISAID software. We selected samples including 19 Omicron variants, 3 Delta variants, 3 Beta variants, and 5 Alpha variants. Furthermore, this novel assay also distinguished Omicron BA.1 and Omicron BA.2 (supplementary table2). Hence, the 5-in-1 VOC assay shows good concordance with WGS and the VirSNiP SARS-CoV-2 test. Additionally, the assay showed no cross-reactivity that is usually observed for other upper respiratory viruses or SARS-CoV-2-negative samples (N=500).Figure 3 Clinical performance results of the 5-in-1 variant of concern (VOC) assay in 150 positive SARS-CoV-2 samples using the BDMAX and LabTurbo AIO system compared with those obtained using the VirSNiP assay.
Figure 3
Discussion
The US CDC defines a VOC as a variant exhibiting increased transmissibility and associated with different disease severities (https://www.cdc.gov/coronavirus/2019-ncov/variants/variant-classifications.html). Further, such variants are associated with a significant recalcitrance to neutralization by antibodies gained from previous infection or vaccinations and reduced efficacy of treatments and/or vaccines. Currently, a range of emergent SARS-CoV-2 variants are being identified as VOCs.
Although WGS has been the primary modality for VOC surveillance, sequencing of SARS-CoV-2–positive samples are limited by both laboratory and bioinformatics capacity. Furthermore, turnaround times using WGS may be days to weeks. We provide a PCR based algorithm for the molecular detection of VOCs could be rapidly executed, providing quick results to inform infection prevention and control and public health measures. PCR may also be more sensitive because WGS is challenging to perform on samples with low viral loads (Wang et al., 2020). We also compared the developed assay with a commercial kit (Roche's VirSNiP) and found good concordance between these two assays. However, commercial kits may not be an affordable option during the COVID-19 pandemic in developing countries. As our 5-in-1 VOC assay is lab-developed, it is economical and hence suitable assay for developing countries. Although the prevalent VOCs worldwide harbor the S gene SNP mutation N501Y, this mutation is not present in all variants. For example, according to our WGS analysis, many Delta variants do not contain this mutation (Figure 1). Our initial screening PCR targeted N501Y, but because of rising rates of Omicron prevalence, we adjusted our laboratory developed test to include K417N and deletion H69/V70 in the same PCR reaction. When SARS-CoV-2-positive samples emitted positive signals for K417N, they were considered to be samples of the Omicron variant. Furthermore, the deletion of H67/V70 could distinguish the Omicron subvariant BA.1 and BA.2. However, our lab-developed assay could not distinguish between Beta and BA.2. The Beta variant was identified in September 2020, in Africa. One year later year, in late November 2021, the Omicron variant was showed up and WHO labeled it as VOC (Flores-Vega et al., 2022). On March 15, 2022, US CDC estimated that 23% of all the then current COVID-19 cases in the USA were caused by the Omicron BA.2 subvariant. Therefore, using our less time-consuming and economical assay, the lineage a VOC belongs to can be deduced. Given the rapid emergence of new variants, ongoing surveillance is key step, and laboratories considering a PCR-based algorithm would need to adapt the algorithm with the prevalence of changes in VOCs.
In this study, we have developed a novel multiplex RT-PCR assay for use with an automatic platform that would serve as a high-throughput and easy-to use detection method for SARS-CoV-2. As the novel coronavirus has had a severe impact on human health, economy, and social life, our assay would have a significant impact in improving the life of affected individuals by enabling early detection of the virus.
Conclusions
In this study, we developed a novel multiplex RT-PCR assay for use with an automatic platform to serve as a high-throughput and easy-to use detection method for SARS-CoV-2, including distinguishing among Alpha, Beta, Gamma, Delta, and Omicron variants. As the novel coronavirus has had a severe impact on human health, economy, and social life, our assay could have a significant impact in improving the life of affected individuals by enabling early detection of the virus.
Conflicts of Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Funding source
This work was supported by the Tri-Service General Hospital, Taipei, Taiwan, R.O.C. [grant number: TSGH-D-110100]. The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.
Ethical Approval Statement
This study was approved by the Institutional Review Board of Tri-Service General Hospital (TSGHIRB number: C202005041), registered on February 8, 2021.
References
Al-Tawfiq Jaffar A., Koritala Thoyaja, Alhumaid Saad, Barry Mazin, Alshukairi Abeer N., Temsah Mohamad-Hani, et al.. Implication of the emergence of the delta (B.1.617.2) variants on vaccine effectiveness. Infection 2022;50(3):583–96.
Challen R, Brooks-Pollock E, Read JM, Dyson L, Tsaneva-Atanasova K, Danon L. Risk of mortality in patients infected with SARS-CoV-2 variant of concern 202012/1: matched cohort study. BMJ 2021;372:n579.
Chang C-K, Jian M, Jr., Chung H-Y, Lin J-C, Hsieh S-S, Tang SH, et al. Clinical Comparative Evaluation of the LabTurboTM AIO® Reverse Transcription-Polymerase Chain Reaction and World Health Organization-Recommended Assays for the Detection of Emerging SARS-CoV-2 Variants of Concern. Infection and Drug Resistance 2022;Volume 15:595-603.
Charre C, Ginevra C, Sabatier M, Regue H, Destras G, Brun S, et al. Evaluation of NGS-based approaches for SARS-CoV-2 whole genome characterisation. Virus Evol 2020;6(2):veaa075.
Chen F, Tian Y, Zhang L, Shi Y. The Role of Children in Household Transmission of COVID-19: A Systematic Review and Meta-Analysis. Int J Infect Dis 2022.
Choi JY, Smith DM. SARS-CoV-2 Variants of Concern. Yonsei Med J 2021;62(11):961-8.
Chung H-Y, Jian M, Jr., Chang C-K, Lin J-C, Yeh K-M, Yang Y-S, et al. Multicenter study evaluating one multiplex RT-PCR assay to detect SARS-CoV-2, influenza A/B, and respiratory syncytia virus using the LabTurbo AIO open platform: epidemiological features, automated sample-to-result, and high-throughput testing. Aging 2021;13(23):24931-42.
Daniels RS, Harvey R, Ermetal B, Xiang Z, Galiano M, Adams L, et al. A Sanger sequencing protocol for SARS-CoV-2 S-gene. Influenza Other Respir Viruses 2021;15(6):707-10.
Davies NG, Jarvis CI, van Zandvoort K, Clifford S, Sun FY, Funk S, et al. Increased mortality in community-tested cases of SARS-CoV-2 lineage B.1.1.7. Nature 2021;593(7858):270-4.
ECDC. Methods for the detection and characterisation of SARS-CoV-2 variants – first update. 20 December 2021.
Flores-Vega VR, Monroy-Molina JV, Jiménez-Hernández LE, Torres AG, Santos-Preciado JI, Rosales-Reyes R. SARS-CoV-2: evolution and emergence of new viral variants. Viruses 2022;14(4):653.
Galloway SE, Paul P, MacCannell DR, Johansson MA, Brooks JT, MacNeil A, et al. Emergence of SARS-CoV-2 B.1.1.7 Lineage - United States, December 29, 2020-January 12, 2021. MMWR Morb Mortal Wkly Rep 2021;70(3):95-9.
Kumar S, Thambiraja TS, Karuppanan K, Subramaniam G. Omicron and Delta variant of SARS-CoV-2: A comparative computational study of spike protein. J Med Virol 2022;94(4):1641-9.
Lu X, Wang L, Sakthivel SK, Whitaker B, Murray J, Kamili S, et al. US CDC Real-Time Reverse Transcription PCR Panel for Detection of Severe Acute Respiratory Syndrome Coronavirus 2. Emerg Infect Dis 2020;26(8).
Ong DSY, Koeleman JGM, Vaessen N, Breijer S, Paltansing S, de Man P. Rapid screening method for the detection of SARS-CoV-2 variants of concern. J Clin Virol. 2021;141:104903.
Park ST, Kim J. Trends in Next-Generation Sequencing and a New Era for Whole Genome Sequencing. Int Neurourol J 2016;20(Suppl 2):S76-83.
Perng CL, Jian MJ, Chang CK, Lin JC, Yeh KM, Chen CW, et al. Novel rapid identification of Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) by real-time RT-PCR using BD Max Open System in Taiwan. PeerJ 2020;8:e9318.
Petersen E, Ntoumi F, Hui DS, Abubakar A, Kramer LD, Obiero C, et al. Emergence of new SARS-CoV-2 Variant of Concern Omicron (B.1.1.529) - highlights Africa's research capabilities, but exposes major knowledge gaps, inequities of vaccine distribution, inadequacies in global COVID-19 response and control efforts. Int J Infect Dis 2022;114:268-72.
Rabaan Ali A., Tirupathi Raghavendra, Sule Anupam A, Aldali Jehad, Mutair Abbas Al, Alhumaid Saad, et al.. Viral Dynamics and Real-Time RT-PCR Ct Values Correlation with Disease Severity in COVID-19. Diagnostics 2021;11(6):1091. Doi: 10.3390/diagnostics11061091.
Ramesh S, Govindarajulu M, Parise RS, Neel L, Shankar T, Patel S, et al. Emerging SARS-CoV-2 Variants: A Review of Its Mutations, Its Implications and Vaccine Efficacy. Vaccines (Basel) 2021;9(10).
Seong H, Hyun HJ, Yun JG, Noh JY, Cheong HJ, Kim WJ, et al. Comparison of the second and third waves of the COVID-19 pandemic in South Korea: Importance of early public health intervention. Int J Infect Dis 2021;104:742-5.
Shaibu JO, Onwuamah CK, James AB, Okwuraiwe AP, Amoo OS, Salu OB, et al. Full length genomic sanger sequencing and phylogenetic analysis of Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) in Nigeria. PLoS One 2021;16(1):e0243271.
Wang F, Huang S, Gao R, Zhou Y, Lai C, Li Z, et al. Initial whole-genome sequencing and analysis of the host genetic contribution to COVID-19 severity and susceptibility. Cell Discov 2020;6(1):83.
Appendix Supplementary materials
Image, application 1
Image, application 2
Image, application 3
Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.ijid.2022.11.027.
| 36455809 | PMC9703862 | NO-CC CODE | 2022-11-29 23:21:14 | no | Int J Infect Dis. 2022 Nov 28; doi: 10.1016/j.ijid.2022.11.027 | utf-8 | Int J Infect Dis | 2,022 | 10.1016/j.ijid.2022.11.027 | oa_other |
==== Front
IJID Reg
IJID Reg
IJID Regions
2772-7076
The Authors. Published by Elsevier Ltd on behalf of International Society for Infectious Diseases.
S2772-7076(22)00145-X
10.1016/j.ijregi.2022.11.010
Article
Survey of healthcare worker perceptions of changes in infection control and antimicrobial stewardship practices in India and South Africa during the COVID-19 pandemic
Mbamalu Oluchi 1⁎
Surendran Surya 2#
Nampoothiri Vrinda 3
Bonaconsa Candice 1
Edathadathil Fabia 3
Zhu Nina 4
Lambert Helen 5
Tarrant Carolyn 6
Ahmad Raheelah 47
Boutall Adam 8
Brink Adrian 9
Steenkamp Ebrahim 10
Holmes Alison 4
Singh Sanjeev 3
Charani Esmita 14
Mendelson Marc 1⁎⁎
1 Division of Infectious Diseases & HIV Medicine, Department of Medicine, Groote Schuur Hospital, University of Cape Town, Cape Town, South Africa
2 Health Systems and Equity, The George Institute for Global Health, New Delhi, India
3 Department of Infection Control and Epidemiology, Amrita Institute of Medical Science, Amrita, Vishwa Vidyapeetham, Kochi, Kerala, India
4 National Institute for Health Research, Health Protection Research Unit in Healthcare Associated, Infections and Antimicrobial Resistance, Department of Medicine, Imperial College London, London, United Kingdom
5 Department of Population Health Sciences, Bristol Medical School, University of Bristol, Bristol, United Kingdom
6 Department of Health Sciences, University of Leicester, Leicester, United Kingdom
7 Division of Health Services Research and Management, School of Health Sciences, City, University of London, United Kingdom
8 Colorectal Unit, Groote Schuur Hospital, University of Cape Town, Cape Town, South Africa
9 Division of Medical Microbiology, Faculty of Health Sciences, National Health Laboratory Service, Groote Schuur Hospital, University of Cape Town, South Africa
10 Statistical Consulting Unit, Department of Statistical Sciences, University of Cape Town, South Africa.
⁎ Corresponding author: Dr. Oluchi Mbamalu, Division of Infectious Diseases & HIV Medicine, Department of Medicine, University of Cape Town, G26/68 Groote Schuur Hospital Observatory,7925, Cape Town, South Africa, Tel: +27 79 794 9346
⁎⁎ Corresponding author: Marc Mendelson, Division of Infectious Diseases & HIV Medicine, Department of Medicine, University of Cape Town, G26/68 Groote Schuur Hospital Observatory,7925, Cape Town, South Africa, Tel: +27 79 794 9346
# This research was conducted when the author was affiliated with the Department of Infection Control and Epidemiology at Amrita Institute of Medical Sciences, Amrita Vishwa Vidyapeetham, Kochi – Kerala, India.
28 11 2022
28 11 2022
© 2022 The Authors. Published by Elsevier Ltd on behalf of International Society for Infectious Diseases.
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Objective
To identify perceptions and awareness of changes in IPC and AMS practices among healthcare workers (HCWs) during the COVID-19 pandemic in India and South Africa (SA).
Method
A self-administered online survey which included participant demographics, knowledge and sources of COVID-19 infection, perceived risks and barriers, and self-efficacy. Data were analysed using descriptive statistics.
Results
321 responses (response rate: 89.2%); 131/321 (40.8%) from India and 190/321 (59.2%) from SA; male to female response rate was 3:2, with majority of respondents aged 40-49 (89/321, 27.7%) and 30-39 (87/321, 27.1%) years. Doctors comprised 47.9% (57/119) of respondents in India and 74.6% (135/181) in SA. Majority of respondents in India (93/119, 78.2%) and SA (132/181, 72.9%) were from the private and public sectors, respectively with more respondents in SA (123/174, 70.7%) than in India (38/104, 36.5%) were involved in antimicrobial prescribing. Respondents reported increased IPC practices since the pandemic and noted need for more training on case management, antibiotic and personal protective equipment (PPE) use. While they noted increased antibiotic prescribing since the pandemic; they did not generally associate their practice with such increase. A willingness to be vaccinated, when vaccination becomes available, was expressed by 203/258 (78.7%) respondents.
Conclusions
HCWs reported improved IPC practices and changes in antibiotic prescribing during the COVID-19 pandemic. Targeted education on correct use of PPE was an identified gap. Although HCWs expressed concerns about antimicrobial resistance, their self-perceived antibiotic prescribing practices seemed unchanged. Additional studies in other settings could explore how our findings fit other contexts.
Keywords
COVID-19
infection prevention
infection prevention and control (IPC) antimicrobial stewardship
healthcare worker
change
==== Body
pmcINTRODUCTION
Since the start of the COVID-19 pandemic, healthcare workers (HCWs) across the world have had to manage and care for patients suspected of or diagnosed with infection, having to adapt their practices according to the emerging evidence about the transmissibility of this virus in healthcare settings (Isilow, 2020; Lacina, 2020). Infection prevention and control (IPC) and antimicrobial stewardship (AMS) practices among HCWs continue to be of critical importance, to prevent nosocomial transmission of this viral infection and other infectious diseases, as well as to improve outcomes (Ashinyo, 2021; Harrison, 2021; Ilesanmi et al., 2021; Pelfrene et al., 2021; Chibabhai et al., 2020; Courtenay et al., 2020; Lynch et al., 2020). HCW approach and behaviour towards IPC and AMS practices play a role in the transmission and spread, as well as outcomes, of COVID-19 in healthcare settings (Suppan et al., 2020; Tartari et al., 2020). To identify appropriate strategies for behavioural interventions to optimize IPC and AMS practices in healthcare settings, it is crucial to explore the awareness and perceptions of such practices amongst HCWs.
The coronavirus disease, first identified as a pneumonia of unknown origin at the end of 2019, has since developed into a pandemic (WHO, 2020a,b). The causative organism has been identified as the severe acute respiratory syndrome coronavirus-2 (COVID-19) (Gorbalenya et al. 2020). The pandemic continues to cause significant burden to healthcare systems worldwide (WHO, 2021a). In India and . (SA), the first cases were reported on 30 January 2020 and 5 March 2020, respectively (National Institute for Communicable Diseases, 2020; Srivastava and Priyadarshni 2020). Currently, these two countries are among the hardest hit by the pandemic in their respective regions (WHO, 2021b).
The importance of education in improved HCW IPC awareness and practices, together with adequate personal protective equipment (PPE) supply, and engagement of nonclinical staff in HCW compliance with IPC measures have been described in the literature (Ashinyo, 2021; Ilesanmi et al., 2021). Current information on the level of HCW awareness of IPC and AMS practice changes is limited, especially as it relates to the COVID-19 pandemic in India and South Africa, where pandemic vulnerability has been significant (WHO, 2021b). We undertook a survey amongst HCWs in each country through an existing research collaboration on IPC practices (Singh et al., 2021; Veepanattu et al., 2020). The aim of this study was to identify perceptions and awareness of changes in IPC and AMS practices among HCWs in India and SA, in the context of the COVID-19 pandemic.
METHODS
Study design
A cross-sectional online survey on the Qualtrics platform, with data collected using a self-administered questionnaire, was performed. Some questions were mandatory, some were optional, and some were automatically included/excluded based on earlier response provided by the respondent. Voluntary response sampling was utilised for the study whose aim was to better understand perceptions of IPC or AMS practice changes in the sampled HCW populations. Any HCW who provided informed consent prior to survey commencement was eligible to participate. The study was approved by the relevant human research ethics committees at the Amrita Institute of Health Sciences, Kerala, India (Ref: IRB-AIMS-2020-232) and the University of Cape Town, South Africa (Ref: 311/2020).
Study development
A cross-sectional survey was conducted via the online platform, Qualtrics. The study report followed the STROBE guidelines (Von Elm, 2008). The research team – made up of pharmacists, physicians, nurses, social scientists, and quantitative data analysts – designed the 43-questions survey to elicit information on HCWs’ perceptions and awareness of changes in IPC and AMS practices across multiple domains. The 4-part survey covered participant demographics, pandemic knowledge and awareness of IPC practices, perceived threats and barriers, and self-efficacy. The survey was piloted within the research team and refined, before wider cascading to participants.
Study settings and participant recruitment
The study was conducted in India and South Africa. Survey participants included HCWs over the age of 18 years – e.g., doctors, pharmacists, nurses, physiotherapists, health/healthcare researchers, and health sciences trainee students – working in any sector or area of healthcare in the two countries. Participants were recruited through personal and professional networks of the researchers in both countries. Participation was voluntary and the survey was open for participation over a three-month period.
Data collection
Data collection was by means of an online self-administered questionnaire. This took place from 15 September to 15 December 2020 and coincided with the first wave of the COVID-19 pandemic in India and the start of the second wave in South Africa. Participant Information Leaflets (PIL) were available for all participants and those willing to participate were required to provide informed consent prior to commencing the survey.
Statistical analysis
Data received from participants were exported to MS Excel and cleaned. Descriptive statistics were used to report participant characteristics and survey responses. The main outcomes of interest were awareness of changes in IPC and AMS practices since the start of the pandemic, in the following domains: pandemic knowledge and awareness of changes in infection care practices; perceived threats and barriers; and self-efficacy. Responses were captured as categorical variables, reported as percentages of received feedback for each item of interest (missing data were excluded), or scaled from strongly agree to strongly disagree, where possible. Pearson's Chi-squared test was used to assess relationships between variables and a logistic regression analysis was conducted with awareness of change in IPC and AMS as the response variable. Both Pearson's Chi-squared and regression tests were conducted using R (version 3.6.2) and p < 0.05 was considered statistically significant.
RESULTS
A total of 360 responses (corresponding to 360 participants clicking on the survey link) were obtained. 354 out of 360 HCW respondents provided consent to participate in the survey (6 out of 360 respondents declined to participate). Of those who consented to participate, responses from 33 respondents were excluded as either the country of residence was not indicated, or respondents were not from either of the two countries, giving a response rate of 89.2% (321/360). Of those who participated, 59.2 % (190/321) were from SA and 40.8% (131/321) were from India (Table 1 ).Table 1 Respondent demographics
Table 1: Response (%)
Characteristics India South Africa Total
Gender
Male 52/131 (39.7) 71/190 (37.4) 123/321 (38.3)
Female 77/131 (58.8) 116/190 (61.1) 193/321 (60.1)
Prefer not to say 2/131 (1.5) 3/190 (1.6) 5/321 (1.6)
Age
20 to 29 years 50/131 (38.2) 10/190 (5.3) 60/321 (18.7)
30 to 39 years 40/131 (30.5) 47/190 (24.7) 87/321 (27.1)
40 to 49 years 25/131 (19.1) 64/190 (33.7) 89/321 (27.7)
50 to 59 years 9/131 (6.9) 43/190 (22.6) 52/321 (16.2)
60 to 69 years 5/131 (3.8) 21/190 (11.1) 26/321 (8.1)
70 years and above 2/131 (1.5) 5/190 (2.6) 7/321 (2.2)
Profession
Medical doctor 57/119 (47.9) 135/181 (74.6) 192/300 (64.0)
Pharmacist 16/119 (11.8) 17/181 (9.4) 31/300 (10.3)
Nurse 3/119 (2.5) 13/181 (7.2) 16/300 (5.3)
Surgeon 11/119 (9.2) 3/181 (1.7) 14/300 (4.7)
Researcher 14/119 (11.8) 11/181 (6.1) 25/300 (8.3)
Healthcare assistant/other qualified healthcare worker 5/119 (4.2) 3/181 (1.7) 8/300 (2.7)
Physiotherapist 0 2/181 (1.1) 2/300 (0.7)
Other 18/119 (15.1) 10/181 (5.5) 28/300 (9.3)
Identifies with multiple professional roles 11/119 (9.2) 13/181 (7.2) 24/300 (8.0)
Work setting
Rural 5/119 (4.2) 6/181 (3.3) 11/300 (3.7)
Urban 103/119 (86.6) 160/181 (88.4) 263/300 (87.7)
Both 11/119 (9.2) 15/181 (8.3) 26/300 (8.7)
Sector of primary work
Private 93/119 (78.2) 33/181 (18.2) 126/300 (42.0)
Public 13/119 (10.9) 132/181 (72.9) 145/300 (48.3)
Both 13/119 (10.9) 14/181 (7.7) 27/300 (9.0)
Other, please specify 0 2/181 (1.1) 2/300 (0.7)
Postgraduate training in infection management or antimicrobial prescribing
Yes 31/119 (26.1) 94/181 (51.9) 125/300 (41.7)
Not sure 7/119 (5.9) 8/181 (4.4) 15/300 (5.0)
No 81/119 (68.1) 79/181 (43.6) 160/300 (53.3)
Involvement in the care of suspected or confirmed COVID-19 patients
Yes 47/115 (40.9) 148/179 (82.7) 195/294 (66.3)
Not sure 2/115 (1.7) 2/179 (1.1) 4/294 (1.4)
No 66/115 (57.4) 29/179 (16.2) 95/294 (32.3)
As part of your job, do you do any of the following in relation to antimicrobials? Please tick all that apply.
Prescribe 38/104 (36.5) 123/174 (70.7) 161/278 (57.9)
Administer 12/104 (11.5) 52/174 (29.9) 64/278 (23.0)
Review 34/104 (32.7) 75/174 (43.1) 109/278 (39.2)
Teach about infection diagnosis and treatment 47/104 (45.2) 77/174 (44.3) 124/278 (44.6)
Develop antimicrobial prescribing policy and guidelines 17/104 (16.3) 38/174 (21.8) 55/278 (19.8)
Other 30/104 (28.8) 16/174 (9.2) 46/278 (16.5)
Respondent demographics
Majority of the survey respondents in SA and India were HCWs from the Western Cape province (130/190, 68.4%) and the state of Kerala (100/131, 76.3%), respectively. There were more female (193/321, 60.1%) than male (123/321, 38.3%) respondents in the survey (Table 1). More than half of total respondents (176/321, 54.8%) were in the 30 to 39 and 40 to 49 years age groups. Most respondents were doctors (192/300, 64.0%) and 28/300 (9.3%) identified as Other or with multiple professional roles (24/300, 8.0%). Among those with Other professions in India (18/119, 15.1%), 7/18 identified as Teacher/Lecturer, 6/18 as Radiographer/Radiologist, 1/18 as a Dentist and one with no further specification. In South Africa, those with Other professions (10/181, 5.5%) identified as infection control practitioners (2/10), managers (2/10), anaesthetist (1/10), clinical psychologist (1/10), occupational therapist (1/10), optometrist (1/10), retired medical doctor (1/10) and one (1/10) with no further specification. The largest proportion of HCWs provided services in urban settings (263/300, 87.7%) and in the public sector (145/300, 48.3%); more respondents in India and SA worked in the private and public sectors, respectively. Of the two people who Indicated Other (in SA) and not Public or Private work settings, one worked in a non-governmental organisation (NGO) and the other provided no further clarification.
Less than half of total respondents (125/300, 41.7%) reported postgraduate training in infection management or antimicrobial prescribing, and 195/294 (66.3%) had been involved in the care of patients with suspected or confirmed COVID-19 infection, more in SA (148/179, 82.7%) than in India (47/115, 40.9%) (Table 1). Over half of respondents (161/278, 57.9%) were prescribers, 124/278 (44.6%) teach about infection diagnosis and treatment, and 15/300 (5%) of total respondents were not sure whether they had any postgraduate training in infection management or antimicrobial prescribing. Among those unsure of postgraduate training in infection management or antimicrobial prescribing were four doctors (4/7, 57.1% – three of the doctors identified as surgeons), two pharmacists (2/7, 28.6%) and one healthcare assistant/other qualified healthcare worker (1/7, 14.3%) in India. In South Africa, 7 out of 8 (7/8, 87.5%) respondents who chose this option identified as medical doctors while 1 out of the 8 was Other with no further clarification.
Among participants who identified as pharmacists in India, most (13/14) noted prescription review as one of the tasks they performed in relation to antimicrobials while none noted antimicrobial administration. About half of pharmacist participants (8/17) in South Africa identified prescription review as one of the tasks they perform in relation to antimicrobials. Pharmacists in South Africa who indicated Other in their selections on this subject listed Other roles in relation to antimicrobial prescribing as dispensing and stewardship practices.
Self-reported changes in personal IPC practices since COVID-19
Most participants reported an increased frequency in their subjective perception of various infection prevention practices assessed (Table 2 ): use of face masks (252/258, 97.7%), hand hygiene (243/258, 94.2%), avoidance of facial contact (214/258, 82.9%), and use of gloves (187/258, 72.5%). More than half of respondents indicated that the use of aprons had increased in their daily practice (148/258, 57.4%) and 137/258 (53.1%) noted a decrease in contact with patient bedside surfaces since the pandemic.Table 2 HCW awareness of changes in care practice
Table 2: Response (%)
Practice Extent of change India SA Total
Hand hygiene Decreased 1/98 (1.0) 1/160 (0.6) 2/258 (0.8)
Has not changed 2/98 (2.0) 10/160 (6.3) 12/258 (4.7)
Increased 95/98 (96.9) 148/160 (92.5) 243/258 (94.2)
Not applicable 0 1/160 (0.6) 1/258 (0.4)
Use of gloves Decreased 2/98 (2.0) 5/160 (3.1) 7/258 (2.7)
Has not changed 2/98 (2.0) 39/160 (24.4) 41/258 (15.9)
Increased 82/98 (83.7) 105/160 (65.6) 187/258 (72.5)
Not applicable 12/98 (12.2) 11/160 (6.9) 23/258 (8.9)
Use of face masks Decreased 0 0 0
Has not changed 1/98 (1.0) 4/160 (2.5) 5/258 (1.9)
Increased 97/98 (99.0) 155/160 (96.9) 252/258 (97.7)
Not applicable 0 1/160 (0.6) 1/258 (0.4)
Use of aprons Decreased 7/98 (7.1) 3/160 (1.9) 10/258 (3.9)
Has not changed 13/98 (13.3) 38/160 (23.8) 51/258 (19.8)
Increased 59/98 (60.2) 89/160 (55.6) 148/258 (57.4)
Not applicable 19/98 (19.4) 30/160 (18.8) 49/258 (19.0)
Avoidance of facial contact Decreased 5/98 (5.1) 8/160 (5.0) 13/258 (5.0)
Has not changed 10/98 (10.2) 20/160 (12.5) 30/258 (11.6)
Increased 83/98 (84.7) 131/160 (81.9) 214/258 (82.9)
Not applicable 0 1/160 (0.6) 1/258 (0.4)
Contact with patient bedside surfaces Decreased 69/98 (70.4) 68/160 (42.5) 137/258 (53.1)
Has not changed 1/98 (1.0) 28/160 (17.5) 29/258 (11.2)
Increased 20/98 (20.4) 51/160 (31.9) 71/258 (27.5)
Not applicable 8/98 (8.2) 13/160 (8.1) 21/258 (8.1)
COVID-19 information sources
Participating HCWs accessed information about the pandemic through a variety of sources (Figures 1 a and 1b). The top sources of information were government websites and news channels/newspapers/journals including online sources. A higher percentage of respondents in SA than in India obtained pandemic-related information from non-government websites (62.0% versus 47.8%, respectively) and colleagues (80.4% versus 57.4%, respectively) while a higher percentage of respondents in India than in South Africa obtained pandemic-related information from social media (60.9% versus 36.9%), and family and friends (37.4% versus 11.7%).Figure 1 (a): Respondents’ sources of COVID-19 information from general (n = 115) and social media in India. (b) Respondents’ sources of COVID-19 information from general (n = 179) and social media in South Africa
Figure 1:
On the social media front, major sources of information across both countries where WhatsApp (80.0% and 71.2% for India and South Africa, respectively), Facebook (58.6% and 45.5% for India and South Africa, respectively) and YouTube (47.1% and 33.3% for India and South Africa, respectively). More respondents in India and South Africa noted access to information via Instagram® and Twitter®, respectively.
Concerns about COVID-19 and its management
Figure 2 shows that while most HCWs agreed that they have sufficient knowledge about the pandemic to appropriately counsel patients on infection prevention measures, a lower number agreed that they have received sufficient training for managing patients with suspected or confirmed COVID-19 infection (108/123 and 24/40 in SA and India, respectively). The need for PPE training was identified by more respondents in India (72/94, 76.6% versus 47/147, 32.0% in SA). Respondents expressed confidence in their ability to use antibiotics for patients in the context of the pandemic (31/43, 72.1% in India and 96/125, 76.8% in SA). Over a quarter of respondents (37/91, 40.7% in India and 52/148, 35.1% in SA) perceived that antibiotic use had increased in their workplace; however, a lower number of those who identified as prescribers (4/29, 13.8% in India and 17/101, 16.8% in SA) associated their own prescribing practice with this increase. Of the HCWs who responded, 77/98 (78.6%) in India and 126/160 (78.8%) in SA reported willingness to be vaccinated for COVID-19 when a vaccine becomes available.Figure 2 Healthcare worker's level of agreement and confidence with care practices
Figure 2:
In Table 3 , the effect of participant demographics on respondent's presentation for a COVID-19 test, respondent's decision to get the COVID-19 vaccine and respondent's concern about sub-optimal IPC behaviour in the workplace are presented; a p-value < 0.05 was considered statistically significant (adjusted p-values parentheses). Majority of the variables tested had no statistical significance to the participant's disposition to present for a COVID-19 test or get the COVID-19 vaccine, or to concerns about sub-optimal workplace IPC behaviour. Following adjustment of p-values, the participant's work setting and receipt of influenza vaccination in the preceding years was found to affect the decision for uptake of the COVID-19 vaccine in South Africa but not in India.Table 3 Relationship between participant demographics and selected outcomes
Table 3:Participant's presentation for a COVID-19 test
Affected by: p-value (India) p-value (South Africa)
- Gender (male/female/prefer not to say)
0.3758 0.4736
- Age
0.0200 (0.6797) 0.0038 (0.1276)
- Work setting (rural/urban/both)
0.2969 0.1615
- Work sector (pubic/private/both/other)
0.8869 0.4692
- Participant training (yes/no/not sure)
0.5106 0.7217
Participant's decision to get the COVID-19 vaccine
Affected by: p-value (India) p-value (South Africa)
- Gender (male/female/prefer not to say)
0.9413 0.3696
- Age
0.7058 0.1926
- Work setting (rural/urban/both)
0.3835 0.0196 (0.6664)
- Work sector (pubic/private/both/other)
0.6408 0.5992
- Participant training (yes/no/not sure)
0.9404 0.2385
- Positive COVID-19 test
0.3435 0.3281
- Influenza vaccination in preceding years
0.5354 4.456e-07 (1.5150e-05)
Participant's concern about sub-optimal infection prevention behaviour in work environment
Affected by: p-value (India) p-value (South Africa)
- Age
0.4853 0.7364
- Work setting (rural/urban/both)
0.5437 0.2927
- Work sector (pubic/private/both/other)
0.0535 0.1220
- Participant training (yes/no/not sure)
0.0963 0.7292
- Positive COVID-19 test
0.7439 0.0689
Antibiotic prescription preferences for different patient sub-cohorts
Survey respondents agreed that individuals infected with the COVID-19 virus were at increased risk of acquiring secondary bacterial infections, and complications from the infection will lead to increased antibiotic prescribing. For specific patient cohorts, prescribers across both countries shared some similarities and differences in tendency to prescribe antibiotics for patients in all cases, in no cases or in selected cases (Table 4 ).Table 4 Self-reported antibiotic prescription preferences for different patient sub- cohorts
Table 4: Response (%)
India South Africa Total
COVID-19 pneumonia
In all cases 12/39 (30.8) 7/117 (6.0) 19/156 (12.2)
In no cases 3/39 (7.7) 49/117 (41.9) 52/156 (33.3)
In selected cases 24/39 (61.5) 61/117 (52.1) 85/156 (54.5)
COVID-19 pneumonia requiring oxygen
In all cases 15/39 (38.5) 12/117 (10.3) 27/156 (17.3)
In no cases 4/39 (10.3) 36/117 (30.8) 40/156 (25.6)
In selected cases 20/39 (51.3) 69/117 (60.0) 89/156 (57.1)
COVID-19 pneumonia requiring hospital admission
In all cases 20/39 (51.3) 15/118 (12.7) 35/157 (22.3)
In no cases 2/39 (5.1) 34/118 (28.8) 36/157 (22.9)
In selected cases 17/39 (43.6) 69/118 (58.5) 86/157 (54.8)
COVID-19 pneumonia with elevated biomarkers (CRP, PCT)
In all cases 26/38 (68.4) 28/118 (23.7) 54/156 (34.6)
In no cases 1/38 (2.6) 19/118 (16.1) 20/156 (12.8)
In selected cases 11/38 (28.9) 71/118 (60.2) 82/156 (52.6)
COVID-19 pneumonia requiring ICU admission
In all cases 24/39 (61.5) 33/116 (28.4) 57/155 (36.8)
In no cases 1/39 (2.6) 18/116 (15.5) 19/155 (12.3)
In selected cases 14/39 (35.9) 65/116 (56.0) 79/155 (51.0)
COVID-19 pneumonia in immunocompromised patient
In all cases 27/39 (69.2) 28/118 (23.7) 55/157 (35.0)
In no cases 1/39 (2.6) 13/118 (11.0) 14/157 (8.9)
In selected cases 11/39 (28.2) 77/118 (65.3) 88/157 (56.1)
COVID-19 pneumonia patient with worsening symptoms
In all cases 27/39 (69.2) 40/118 (33.9) 67/157 (42.7)
In no cases 1/39 (2.6) 7/118 (5.9) 8/157 (5.1)
In selected cases 11/39 (28.2) 71/118 (60.2) 82/157 (52.2)
First line antibiotic regimen for COVID pneumonia (multiple answers enabled)
Azithromycin 21/57 (36.8) 10/117 (8.5) 31/174 (17.8)
Co-amoxiclav 3/57 (5.3) 26/117 (22.2) 29/174 (16.7)
Ceftriaxone 1/57 (1.8) 18/117 (15.4) 19/174 (10.9)
Multiple antibiotics 4/57 (7.0) 31/117 (26.5) 35/174 (20.1)
First-line antibiotic for COVID-related pneumonia varied, with common choices favouring azithromycin and co-amoxiclav – more in India (36.8% versus 8.5%) and South Africa 22.2% versus 5.3%), respectively. These agents were also noted in cases where respondents preferred more than one antibiotic choice. Ceftriaxone was also among the preferred first-line antibiotics for COVID-related pneumonia, mostly by respondents in SA (18/117, 15.4%) than those in India (1/57, 1.8%).
DISCUSSION
This study provides insight into HCW's awareness of changes in IPC and AMS practices in the context of the COVID-19 pandemic in India and SA. The study findings add to the body of knowledge on HCW readiness and needs for pandemic mitigation, providing an understanding of HCW perceptions which can be further explored for improved IPC and AMS practices.
Most of the respondents in India were from the private sector while in SA, most respondents were from the public sector. This reflects, to some extent, the public/private healthcare provisions in the countries (Rout et al., 2021; Maphumulo and Bhengu, 2019). HCWs reported an awareness of and improvement in IPC practices since the pandemic. They also noted increase in antibiotic prescription volumes within their institutions though, generally, did not associate their own practice with such increase.
The HCW's degree of confidence in their own IPC practices varied across the two countries. The scale-up of infection prevention measures, most notably in the use of PPE by HCWs and in patient care (Deressa et al., 2021; Nimer et al., 2021), was earlier affected by shortages and supply issues. The study presents data mostly from HCWs who practiced in specific regions in India and South Africa and so responses may not be generalizable. Nevertheless, the lack of PPE has been reported across both countries, as in many others (Burki, 2020; Cohen and van der Meulen Rodgers, 2020; Iacobucci, 2020; Mbunge, 2020; Savoia et al., 2020; Sharma et al., 2020). In line with other studies, respondents reported improved IPC practices in the context of the pandemic (Deressa et al., 2021; Nimer et al., 2021). This has been noted to be consistent with HCW perceptions of infection risk and concern over the COVID-19 infection and its related complications (Deressa et al., 2021; Nimer et al., 2021).
A willingness to be vaccinated was high among the respondents, though about a fifth provided no response to this question. Variation in COVID-19 vaccine acceptance among HCWs have been noted in various parts of the world (Biswas et al., 2021; Gadoth et al., 2021; Di Gennaro et al., 2021; Qunaibi et al., 2021; Sallam et al., 2021; Shekhar et al., 2021). Hesitation towards vaccination has been reported among non-physician HCWs, those who utilised specific social media platforms as major information sources, and those who are not positively disposed to previous or regular influenza vaccination. COVID-19 vaccine acceptance, on the other hand, has been noted among individuals of varying ages and those perceived to be at risk of infection (Biswas et al., 2021; Gadoth et al., 2021; Di Gennaro et al., 2021; Qunaibi et al., 2021; Sallam et al., 2021; Shekhar et al., 2021; Kabamba et al., 2020). The positive attitude to vaccination in our survey may be related to the cohort which were all HCWs with higher education levels, as earlier reported elsewhere (Shekhar et al., 2021). Even though vaccination was not yet under way in several countries at the time of data collection and has since expanded to many countries, with HCWs in the early recipient groups; continued evaluation of attitudes towards vaccination is necessary to inform future education drives and planning for pandemic containment and mitigation (Gadoth et al., 2021; Ledda et al., 2021). There is also a need for ongoing education and awareness of appropriate IPC and antibiotic prescribing practices, tailored to various contexts, to further contribute to control of the COVID-19 pandemic.
HCWs expressed their preferences for antibiotic prescribing in different patient sub-cohorts, depending on the patient's clinical presentation. As noted in this survey, amoxicillin/clavulanic acid and azithromycin also feature in the antibiotic choices considered by HCWs in other studies (Dudoignon et al., 2021; Townsend et al., 2020). Previous studies have reported on the limited and recommended roles for broad-spectrum antibiotic prescribing in patients (Ginsburg and Klugman, 2020; Wei et al., 2020), as well as on records of broad-spectrum antibiotic prescribing based on clinical presentation, in COVID-19 patients (Beović et al., 2020). Clinician experience is believed to have largely guided empiric antibiotic therapy in the context of COVID-19, especially at the early stages of the pandemic (Chang and Chan, 2020). It is important to longitudinally study the effect of this pandemic on antibiotic prescribing patterns and epidemiology of bacterial infectious diseases, as well as the impact of the pandemic on AMS programmes (Sieswerda et al., 2021; Beović et al., 2020; Chang and Chan, 2020). The data generated from such studies will contribute to better understanding of the pandemic impact on antimicrobial resistance and AMS as well as support the development of sustainable and evidence-based IPC and AMS practices. The importance of effective IPC behaviours, in addition to AMS practices, for pandemic control and mitigation, cannot be over-emphasised.
LIMITATIONS
This study is subject to some limitations, which need to be considered in the interpretation of its findings. Firstly, by virtue of its design as a cross-sectional study, the survey findings are subject to change over time and with interventions, especially as HCWs gain more experience with successive waves of the pandemic. Secondly, we present data from a relatively low number of HCWs during the COVID-19 pandemic in India and SA. While the survey was rolled out across the two countries and participation amongst HCWs was not limited, responses obtained were considerably more from participants in the geographic areas where the researchers worked compared to the rest of the country. These areas were at the forefront of the COVID-19 response across both countries; however, there were also other healthcare settings at the forefront whose healthcare teams were not necessarily represented in the survey responses. The limited participation may be due to HCW preoccupation with pandemic mitigation efforts which understandably takes preference. Many of the respondents work in the geographic areas where the research team practice and so responses may be more representative of practice in the respective areas rather than of each country. The responses were obtained from HCWs across two countries at different times and phases of the pandemic; this may have had some bearing on the findings. Another limitation was posed by the online format of the study and its voluntary sampling method which would have biased results in favour of those who had online access or those who inherently were more disposed to the subject or more disposed to participating. Lastly, response bias cannot be ruled out given the survey's reliance on a self-administered and self-reported questionnaire. For instance, while many respondents indicated that IPC measures had increased in their healthcare and work settings since the pandemic, this was from their subjective perceptions (perhaps when compared against a previously poor rate) and may not necessarily reflect or equate to appropriate practices as recommended by various guidelines.
Notwithstanding, this study fills a gap in the perceptions of IPC practices by presenting insight into HCW's views and awareness of IPC and AMS changes in patient care across two countries on different sides of the middle-income scale, in two continents, during the early COVID-19 infection waves. It will be useful for providing information on perceptions to and awareness of pandemic containment and mitigation measures among HCWs in the response areas, in this and future infectious disease pandemics. This is beneficial given the position of HCWs in infectious disease and pandemic control. In addition, participation in the survey was voluntary and anonymous which likely increased the likelihood of reliable responses.
CONCLUSION
HCWs reported awareness of improved IPC measures and changes in antibiotic prescribing during the COVID-19 pandemic. Targeted education on correct use of PPE was identified as a gap to be addressed during this pandemic. Although HCWs noted increased antibiotic prescribing in their work environment, their own antibiotic prescribing practices were perceived to be largely unchanged. Strategies for IPC interventions, including AMS, need to be strengthened in infectious disease pandemic response plans with context-specific interventions to make prescribers aware of their possible contributions to AMR. While these findings cannot be generalised, they highlight the need for continued IPC and AMS awareness amongst HCWs. Additional studies across various other settings are required to explore how much the findings of this research fit with those from other contexts.
Authors’ Contributions
OM conceptualised and wrote the initial protocol for the study, with additional input and revision from CB, HL, RA, S Singh and overall oversight by EC and MM. OM, S Surendran and EC coordinated the data collection with input from NZ. SSurendran, VN, and FE contributed to data capturing; OM, FE and ES contributed to the data analysis. OM wrote the first draft of the manuscript, with oversight from EC and MM. All authors contributed to subsequent revisions and approval of the final draft.
REFERENCES
Ashinyo ME, Dubik SD, Duti V, Amegah KE, Ashinyo A, Asare BA, et al. Infection prevention and control compliance among exposed healthcare workers in COVID-19 treatment centers in Ghana: a descriptive cross-sectional study. PLOS ONE. 2021;16(3 March):1-13. doi:10.1371/journal.pone.0248282
Beović B, Doušak M, Ferreira-Coimbra J, Nadrah K, Rubulotta F, Belliato M, et al. Antibiotic use in patients with COVID-19: a “snapshot” Infectious Diseases International Research Initiative (ID-IRI) survey. J Antimicrob Chemother. 2020;75(11):3386-90. doi:10.1093/jac/dkaa326
Biswas N, Mustapha T, Khubchandani J, Price JH. The nature and extent of COVID-19 vaccination hesitancy in healthcare workers. J Community Health. 2021;(0123456789). doi:10.1007/s10900-021-00984-3
Burki T. Global shortage of personal protective equipment. Lancet Infect Dis. 2020;20(7):785-6. doi:10.1016/S1473-3099(20)30501-6
Chang CY, Chan KG. Underestimation of co-infections in COVID-19 due to non-discriminatory use of antibiotics. J Infect. 2020;81(3):e29-e30. doi:10.1016/j.jinf.2020.06.077
Chibabhai V, Duse AG, Perovic O, Richards GA. Collateral damage of the COVID-19 pandemic: exacerbation of antimicrobial resistance and disruptions to antimicrobial stewardship programmes? South African Med J. 2020;110(7):572-3. doi:10.7196/SAMJ.2020.V110I7.14917
Cohen J, van der Meulen Rodgers Y. Contributing factors to personal protective equipment shortages during the COVID-19 pandemic. Prev Med (Baltim). 2020;141(October):106263. doi:10.1016/j.ypmed.2020.106263
Courtenay M, Burnett E, Castro-Sánchez E, Du Toit B, Figueiredo RM, Gallagher R, et al. Preparing nurses for COVID-19 response efforts through involvement in antimicrobial stewardship programmes. J Hosp Infect. 2020;106(1):176-8. doi:10.1016/j.jhin.2020.06.011
Deressa W, Worku A, Abebe W, Gizaw M, Amogne W. Risk perceptions and preventive practices of COVID-19 among healthcare professionals in public hospitals in Addis Ababa, Ethiopia. PLoS One. 2021;16(6):e0242471. doi:10.1371/journal.pone.0242471
Di Gennaro F, Murri R, Segala FV, Cerruti L, Abdulle A, Saracino A, et al. Attitudes towards anti-sars-cov2 vaccination among healthcare workers: Results from a national survey in Italy. Viruses. 2021;13(3):1-11. doi:10.3390/v13030371
Dudoignon E, Caméléna F, Deniau B, Habay A, Coutrot M, Ressaire Q, et al. Bacterial pneumonia in COVID-19 critically ill patients: a case series. Clin Infect Dis. 2021;72(5):905-6. doi:10.1093/cid/ciaa762
Gadoth A, Halbrook M, Martin-Blais R, Gray A, Tobin NH, Ferbas KG, et al. Cross-sectional assessment of COVID-19 vaccine acceptance among health care workers in Los Angeles. Ann Intern Med. 2021;174(6):882-5. doi:10.7326/m20-7580
Ginsburg AS, Klugman KP. COVID-19 pneumonia and the appropriate use of antibiotics. Lancet Glob Heal. 2020;8(12):e1453-e1454. doi:10.1016/S2214-109X(20)30444-7
Gorbalenya AE, Baker SC, Baric RS, de Groot RJ, Drosten C, Gulyaeva AA, et al. The species Severe acute respiratory syndrome-related coronavirus: classifying 2019-nCoV and naming it SARS-CoV-2. Nat Microbiol. 2020;5(4): 536-544. doi:10.1038/s41564-020-0695-z.
Harrison R. COVID-19: occupational health issues for health care personnel. 2022 Jan 28 [cited 2022 Nov 20]. Available from: https://www.uptodate.com/contents/covid-19-occupational-health-issues-for-health-care-workers
Iacobucci G. Covid-19: doctors still at “considerable risk” from lack of PPE, BMA warns. BMJ. 2020;368(March):m1316. doi:10.1136/bmj.m1316
Ilesanmi OS, Afolabi AA, Akande A, Raji T, Mohammed A. Infection prevention and control during COVID-19 pandemic: realities from healthcare workers in a north central state in Nigeria. Epidemiol Infect. Published online 2021 January 7. doi:10.1017/S0950268821000017.
Isilow H. S. Africa: over 390 healthcare workers claimed by virus. 2020 Dec 15 [cited 2021 Apr 15]. Available from: https://www.aa.com.tr/en/africa/safrica-over-390-healthcare-workers-claimed-by-virus/2077832#
Kabamba Nzaji M, Kabamba Ngombe L, Ngoie Mwamba G, Banza Ndala DB, Miema JM, Lungoyo CL, et al. Acceptability of vaccination against COVID-19 among healthcare workers in the Democratic Republic of the Congo. Pragmatic Obs Res. 2020;11:103-9. doi:10.2147/por.s271096
Lacina L. What's needed now to protect health workers: WHO COVID-19 Briefing. 2020 Apr 10 [cited 2021 Aug 7]. Availale from: https://www.weforum.org/agenda/2020/04/10-april-who-briefing-health-workers-covid-19-ppe-training/
Ledda C, Costantino C, Cuccia M, Maltezou HC, Rapisarda V. Attitudes of healthcare personnel towards vaccinations before and during the covid-19 pandemic. Int J Environ Res Public Health. 2021;18(5):1-11. doi:10.3390/ijerph18052703
Lynch C, Mahida N, Gray J. Antimicrobial stewardship: a COVID casualty? J Hosp Infect. 2020;106(3):401-3. doi:10.1016/j.jhin.2020.10.002
Maphumulo WT, Bhengu BR. Challenges of quality improvement in the healthcare of South Africa post-apartheid: a critical review. Curationis, 2019;42(1):1-9. doi: 10.4102/curationis.v42i1.1901.
Mbunge E. Effects of COVID-19 in South African health system and society: An explanatory study. Diabetes Metab Syndr Clin Res Rev. 2020;14(6):1809-14. doi:10.1016/j.dsx.2020.09.016
National Institute for Communicable Diseases (NICD). First case Of COVID-19: coronavirus reported in SA. 2020 Mar 5 [cited 2021 Apr 15]. Available from: https://www.nicd.ac.za/first-case-of-covid-19-coronavirus-reported-in-sa/
Nimer R, Swedan S, Kofahi H, Khabour O. Increased adherence to infection control practices among medical laboratory technicians during the COVID-19 pandemic: a self-reported survey study. Ann Glob Heal. 2021;87(1):1-10. doi:10.5334/aogh.3378
Pelfrene E, Botgros R, Cavaleri M. Antimicrobial multidrug resistance in the era of COVID-19: a forgotten plight? Antimicrob Resist Infect Control. 2021;10(1):1-6. doi:10.1186/s13756-021-00893-z
Qunaibi E, Basheti I, Soudy M, Sultan I. Hesitancy of Arab healthcare workers towards COVID-19 vaccination: a large-scale multinational study. Vaccines. 2021;9(5):1-13. doi:10.3390/vaccines9050446
Rout SK, Sahu KS, Mahapatra S. Utilization of healthcare services in public and private healthcare in India: causes and determinants. Int J Healthc Manag, 2021;14(2):509-16. doi: 10.1080/20479700.2019.1665882.
Sallam M. Covid-19 vaccine hesitancy worldwide: A concise systematic review of vaccine acceptance rates. Vaccines. 2021;9(2):1-15. doi:10.3390/vaccines9020160
Savoia E, Argentini G, Gori D, Neri E, Piltch-Loeb R, Fantini MP. Factors associated with access and use of PPE during COVID-19: a cross-sectional study of Italian physicians. PLoS One. 2020;15:1-12. doi:10.1371/journal.pone.0239024
Sharma N, Hasan Z, Velayudhan A, M. A E, Mangal DK, Gupta SD. Personal protective equipment: challenges and strategies to combat COVID-19 in India: a narrative review. J Health Manag. 2020;22(2):157-68. doi:10.1177/0972063420935540
Shekhar R, Sheikh AB, Upadhyay S, Singh M, Kottewar S, Mir H, et al. COVID-19 vaccine acceptance among healthcare workers in the United States. Vaccines. 2021;9(2):1-18. doi:10.3390/vaccines9020119
Sieswerda E, de Boer MGJ, Bonten MMJ, Boersma WG, Jonkers RE, Aleva RM, et al. Recommendations for antibacterial therapy in adults with COVID-19 – an evidence based guideline. Clin Microbiol Infect. 2021;27(1):61-6. doi:10.1016/j.cmi.2020.09.041
Singh S, Mendelson M, Surendran S, Bonaconsa C, Mbamalu O, Nampoothiri V, et al. Investigating infection management and antimicrobial stewardship in surgery: a qualitative study from India and South Africa. Clin Microbiol Infect. 2021;27(10):1455-64. doi:10.1016/j.cmi.2020.12.013
Srivastava V, Priyadarshini S. India reports first case of novel coronavirus. Nature India. 2020 Jan 20 [cited 2021 Apr 15]. Available from: https://www.natureasia.com/en/nindia/article/10.1038/nindia.2020.15
Suppan M, Catho G, Nunes TR, Sauvan V, Perez M, Graf C, et al. A serious game designed to promote safe behaviors among health care workers during the COVID-19 pandemic: development of “Escape COVID-19.” JMIR Serious Games. 2020;8(4). doi:10.2196/24986
Tartari E, Saris K, Kenters N, Marimuthu K, Widmer A, Collignon P, et al. Not sick enough to worry? “Influenza-like” symptoms and work-related behavior among healthcare workers and other professionals: results of a global survey. PLoS One. 2020;15(5):1-14. doi:10.1371/journal.pone.0232168
Townsend L, Hughes G, Kerr C, Kelly M, O'Connot R, Sweeney E, et al. Bacterial pneumonia coinfection and antimicrobial therapy duration in COVID-19 (COVID-19) infection. JAC-Antimicrobial Resist. 2020;2(3):2-5. doi:10.1093/jacamr/dlaa071
Veepanattu P, Singh S, Mendelson M, Nampoothiri V, Edathadatil F, Surendran S, et al. Building resilient and responsive research collaborations to tackle antimicrobial resistance—lessons learnt from India, South Africa, and UK. Int J Infect Dis. 2020;100(P278-282). https://doi.org/10.1016/j.ijid.2020.08.057
Von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP; STROBE Initiative. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. J Clin Epidemiol. 2008;61(4):344-9. doi:10.1016/j.jclinepi.2007.11.008
Wei W, Ortwine JK, Mang NS, Joseph C, Hall BC, Prokesch BC. Limited role for antibiotics in COVID-19: scarce evidence of bacterial coinfection. SSRN Electron J. Published online 2020. doi:10.2139/ssrn.3622388
World Health Organization (WHO). Pneumonia of unknown cause. 2020a, Jan 5 [cited 2021 Apr 15]. Available from: https://www.who.int/csr/don/05-january-2020-pneumonia-of-unkown-cause-china/en/#.XyAbsvYF-rc.mendeley
World Health Organization (WHO). WHO Director-General's opening remarks at the media briefing on COVID-19. 2020b Mar 25 [cited 2020 April 10]. Available from: https://www.who.int/dg/speeches/detail/who-director-general-s-opening-remarks-at-the-media-briefing-on-covid-19—25-march-2020
World Health Organization (WHO). COVID-19 weekly epidemiological update 22. 2021a Jan 12 [cited 2021 Apr 15]. Available from: https://www.who.int/docs/default-source/coronaviruse/situation-reports/weekly_epidemiological_update_22.pdf
World Health Organization (WHO). COVID-19 Weekly epidemiological update 54. 2021b Aug 24 [cited 2021 Aug 29]. Available from: https://www.who.int/publications/m/item/weekly-epidemiological-update-on-covid-19—24-august-2021
Appendix Supplementary materials
Image, application 1
Acknowledgements: This study is part of the ASPIRES project (Antibiotic use across Surgical Pathways – Investigating, Redesigning and Evaluating Systems) (https://www.imperial.ac.uk/arc/aspires/). ASPIRES aims to address antimicrobial resistance and improve clinical outcomes by optimising antibiotic usage along surgical pathways. This work is also based on research supported in part by the National Research Foundation of South Africa (Grant Number: 129755). The authors wish to acknowledge the funders, all survey participants, and all those who participated in the review of and provided feedback on the survey tool.
Funding Sources: This research was supported by the Economic and Social Research Council (ESRC) as part of the Antimicrobial Cross Council Initiative supported by the seven UK research councils, the Global Challenges Research Fund (GCRF) as part of the ASPIRES project (https://www.imperial.ac.uk/arc/aspires/), the National Institute for Health Research, UK Department of Health [HPRU-2012-10047] in partnership with Public Health England and the National Research Foundation of South Africa (Grant Number: 129755). The funders did not have any role in the study design and conduct, review or approval of the manuscript, or the decision to submit the manuscript for publication.
Declaration of Competing Interests: The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.ijregi.2022.11.010.
| 36466212 | PMC9703863 | NO-CC CODE | 2022-11-29 23:21:40 | no | IJID Reg. 2022 Nov 28; doi: 10.1016/j.ijregi.2022.11.010 | utf-8 | IJID Reg | 2,022 | 10.1016/j.ijregi.2022.11.010 | oa_other |
==== Front
Sustain Cities Soc
Sustain Cities Soc
Sustainable Cities and Society
2210-6707
2210-6715
Elsevier Ltd.
S2210-6707(22)00630-8
10.1016/j.scs.2022.104326
104326
Article
Towards building resilient cities to pandemics: A review of COVID-19 literature
Amirzadeh Melika a⁎
Sobhaninia Saeideh b
Buckman Stephen T. c
Sharifi Ayyoob d
a Faculty of Architecture and Urban Planning, University of Art, 24 Arghavan Alley, Laleh St., Artesh Blvd., Tehran, Iran
b Planning, Design, and the Built Environment Department, Clemson University, 511 Roper Mountain Rd, Greenville, SC 29615, United States
c Department of City Planning and Real Estate Development, Clemson University, One North Main St., Greenville, SC 29601, United States
d Graduate School of Humanities and Social Sciences and Network for Education and Research on Peace and Sustainability (NERPS), Hiroshima University, Hiroshima 739-8511, Japan
⁎ Corresponding author.
28 11 2022
2 2023
28 11 2022
89 104326104326
25 6 2022
26 11 2022
26 11 2022
© 2022 Elsevier Ltd. All rights reserved.
2022
Elsevier Ltd
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
With the global prevalence of COVID-19 disease, the concept of urban resilience against pandemics has drawn the attention of a wide range of researchers, urban planners, and policymakers. This study aims to identify the major dimensions and principles of urban resilience to pandemics through a systematic review focused on lessons learned from the COVID-19 pandemic and comparing different perspectives regarding resilient urban environments to such diseases. Based on the findings, the study proposes a conceptual framework and a series of principles of urban resilience to pandemics, consisting of four spatial levels: housing, neighborhoods, city, and the regional and national scales, and three dimensions of pandemic resilience: pandemic-related health requirements, environmental psychological principles, and general resilience principles. The findings show that resilient cities should be able to implement the pandemic-related health requirements, the psychological principles of the environment to reduce the stresses caused by the pandemic, and the general principles of resilience in the smart city context. This framework provides scholars and policymakers with a comprehensive understanding of resilience on different scales and assists them in making better-informed decisions.
Keywords
Urban resilience
Post-COVID urban planning
Anti-virus urban design
Pandemics
Resilient design principles
==== Body
pmc1 Introduction
On March 11, 2020, World Health Organization (WHO) declared COVID-19 a pandemic (World Health Organization, 2020). According to WHO, the COVID-19 pandemic has led to the infection of about 623,479,824 people and the death of 6,625,763 people worldwide (WHO, 2022). Aside from the high mortality rate, this pandemic has led to ongoing problems and widespread global disruptions that have impacted people's lives in many aspects (Shakil et al., 2020).
Like natural disasters, pandemics cause social, organizational, and economic disruptions. Therefore, it is no surprise that COVID-19 has caused significant disruptions at all levels in terms of social impacts, from national lockdowns to self-isolation, resulting in adverse effects on small businesses and the overall economy (Sakurai & Chughtai, 2020). Moreover, cities are particularly impacted by local and global connectedness, high levels of human mobility, and a high concentration of economic activities. Therefore, it is unsurprising that cities have been epicenters of the pandemic in different parts of the world (Kummitha, 2020). Consequently, there have been renewed debates over the role of urban planning and design in controlling diseases on the one hand and maintaining the viability and economy of cities on the other hand.
Until 2020, there was limited research on the role of urban planning and design in controlling pandemics. Most policymakers mainly focused on short-term solutions, such as the lockdown of cities, public transport closure, and social distancing to manage the pandemics' risks. The main reason behind this lack of contribution is little to no consideration of calamities like pandemics in such domains (Allam & Jones, 2020) since pandemics do not frequently occur, unlike other disasters and stressors. In addition, contrary to natural disasters, pandemics often directly threaten people and the economy, not the infrastructure and built environment. Therefore, the proposed solutions are more related to public economic policy and public health issues than the need to protect or rebuild infrastructure (Litman, 2020). Another reason is that pandemics are often unpredictable, and each pandemic probably needs different design strategies (WHO, 2018). Furthermore, urban planning and design are long, drawn-out processes taking years, while reactions to pandemics are often “just-in-time” reactions.
Although resilience has been widely used for several decades in various fields, such as physics, ecology, psychology, and economy, it is a relatively novel concept in urban planning and design (Sharifi & Yamagata, 2016). About two decades ago, the resilience concept gained ground within urban planning and design (Sharifi & Yamagata, 2018a). Since then, it has been increasingly used as an organizing framework to guide scientific and political discourses in many urban contexts (Sharifi & Yamagata, 2018b). However, the main focus of resilience in urban planning and design has been on the resilience of cities and their different subsystems against adverse events such as floods, earthquakes, tsunamis, and wildfires, not pandemics. But the COVID-19 pandemic showed how different characteristics of cities play critical roles before (prevention), during (reduction through segregation), and after (planning and risk management strategies for the future) pandemics (Block et al., 2020; Lai et al., 2020).
Therefore, this study seeks to identify the main dimensions that form the resilience of cities to pandemics, the spatial scales that urban planners and policymakers need to consider in planning for pandemic resiliency of cities, the measures that should be adopted to improve the resilience of cities to pandemics, the importance of health protocols in resilient cities to pandemics, and the role of environmental psychology in reducing peoples’ stress level in cities. For this purpose, the literature on this topic is reviewed, and a conceptual framework presenting the identified dimensions, and spatial levels is introduced. The presented framework and the proposed principles can be effective in post-COVID urban planning and help practitioners and decision-makers to take action toward building more resilient cities in the face of pandemics.
2 Materials and methods
To address the research objectives, relevant studies were selected based on the systematic-review framework of Moher et al., (2009). To identify the main concepts, theories, and knowledge gaps, the Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) Checklist was used (Tricco et al., 2018). Then, each section of the PRISMA-ScR Checklist was categorized using the inductive content analysis method. In the following paragraphs, each step is explained in more detail.
First, a broad search was conducted on the Web of Science (WoS) on March 25th, 2022. The search was limited to English-published studies, using the search string:
(TS=("pandemic*" OR "epidemic*" OR "corona*" OR "covid*") AND TS=("urban*" OR "city" OR "built environment") AND TS=("resilien*"))
This search returned 707 articles, and 27 more papers were also found by searching Google Scholar and screening the articles’ references. Then, the abstracts of all the 734 studies were examined to find the most relevant ones to this study. In this step, the studies that included the characteristics of cities, which are more resilient against coronavirus, or resilient urban design against pandemics, were identified. Thus, 548 studies related to the COVID-19 pandemic but not focused on the scope of this study were excluded.
The next step was extracting and categorizing the data. In this step, the remaining 183 studies were explored using the PRISMA-ScR Checklist to find information related to urban resilience to pandemics and different resilience attribute(s) and categories. The information on all studies was covered and categorized into 22 items of the PRISMA-ScR Checklist. Then, the information in each article was further sub-classified into different categories via qualitative inductive content analysis. Therefore, the information in each item was further sub-classified into different categories via qualitative inductive content analysis. Thus, the information under each section of the present study was obtained inductively as the articles were examined. An Excel spreadsheet was developed to store the extracted data. As we continued the content analysis of the studies, new data were added to the existing categories. If not relevant to existing classes, new ones were created. This process continued until all articles were covered. Therefore, the categories were refined throughout the review process, and data with similar themes were classified into the same groups until all the data was covered. This method inductively extracted new ideas from the previous literature and reduced researchers’ bias. Using the systematic review method allowed the researchers to cover the data as much as possible, compare different ideas, avoid redundancy, and classify the data with similar themes into the same categories.
As a result of a comprehensive systematic literature review and following data categorization based on inductive content analysis, the issues related to urban resilience to pandemics were classified into four categories: ‘the pandemic-related health requirements’, ‘the environmental psychological principles’, ‘the general principles of resilience’, and ‘the smart city’. In addition, four spatial scales were identified: ‘housing’, ‘neighborhood’, ‘the city’, and ‘regional and national’.
Because other studies were published since we first started our search in 2022, we considered their insights in our study even though they were not part of the systematic search. Furthermore, the study's methodology enabled us to include many relevant studies in the reviewing process. Although other relevant studies might not have been included, the number of reviewed studies was sufficient to achieve the study's objectives. The reviewing process continued until data saturation, and adding more papers would probably not alter the results. Fig. 1 shows the process of selecting related studies and their analysis.Fig. 1 Procedures for literature search and selection of articles through the research phases. Source: adapted from Moher et al. (2009).
Fig 1
3 Results and discussions
3.1 Urban resilience and resilient cities to pandemics
In 1973, Holling (1973) introduced the term "resilience" in the ecological literature in his study, Resilience and Stability of Ecological Systems, for the first time. He defined resilience as a way to understand the dynamic and nonlinear stresses absorbed in the ecosystem and the amount of perturbation that can be absorbed by the ecosystem so that it can remain stable without significant changes in its structure. Although the initial definitions of the concept are often focused on the resistance of a system or returning to the equilibrium after experiencing a shock or a sudden change (Ludwig et al., 1997; Pimm, 1991), today, resilience is considered to be a broader concept that recognizes the importance of adaptation and non-equilibrium dynamics that is not focused solely on sudden shocks or disruptions (Amirzadeh & Barakpour, 2021; 2019a).
In recent years, many studies have used the concept of resilience in "urban systems". Some researchers have described cities as complex and adaptable social-ecological systems. They argue that resilience provides a valuable perspective for ecologists, planners, and other involved actors in urban development in the face of uncertainties (Orleans Reed et al., 2013). The idea of urban resilience generally indicates the ability to adapt and respond positively to shocks and changes in an urban system (Desouza & Flanery, 2013). Meerow et al. (2016: 42–45) noted that there are six conceptual differences related to resilience definitions in previous research: "(1) definition of 'urban'; (2) understanding of system equilibrium; (3) positive vs. neutral (or negative) conceptualizations of resilience; (4) mechanisms for system change; (5) adaptation versus general adaptability; and (6) timescale of action". The concept of urban resilience is related to studying how ecological systems adapt to disruptions caused by external factors (Davic & Welsh, 2004). This concept is generally about how an urban system can withstand a wide range of disturbances (Leichenko, 2011). These urban stresses are not situated in one area but, as Buckman & Rakhimova (2020) point out, are part of an interconnected structure that includes the environment, governance, economics, and community. Thus, it is essential to see urban resilience as a multi-dimensional concept in a way that neglecting some aspects of it leads to incomplete and incorrect conclusions about this concept (Amirzadeh & Barakpour, 2019b; Buckman & Sobhaninia, 2022; Jabareen, 2013).
Despite considerable attention to urban resilience and its frequent usage, this concept has remained ambiguous, with different interpretations in policy and academic discussions about cities (Amirzadeh et al., 2022; Sobhaninia & Buckman, 2022). Even though there are various interpretations of this concept, one of the best definitions was presented by Meerow et al. (2016). They (2016: 42–45) defined urban resilience as “the ability of an urban system – and all its constituent socio-ecological and socio-technical networks across temporal and spatial scales – to maintain or rapidly return to desired functions in the face of a disturbance, to adapt to change, and to transform systems quickly that it limits current or future adaptive capacity”. However, after reviewing the literature, it can be concluded that in addition to these resilience features, resilient cities to pandemics should also have healthy and stress-free environments (Gu et al., 2020; Megahed & Ghoneim, 2020; Tokazhanov et al., 2020). Therefore, urban resilience to pandemics can be defined as the ability of an urban system to continue its desired function and provide a sanitary and stress-free environment for its citizens during different stages of pandemics.
After analyzing the literature on resilient cities to pandemics, different categories for the data were obtained, which are shown graphically in Fig. 2 . The framework in Fig. 2 includes three essential dimensions of resilient cities to pandemics: Pandemic-related Health Requirements, Environmental Psychological Principles, and General Resilience Principles in the context of a smart city. The more cities are transformed to include requirements for improving the resilience of cities to such diseases, the faster the control of the disease and the improvement of people's life quality during pandemics will be. Moreover, according to the comprehensive systematic review, resilient city features can be classified into four spatial levels: housing, neighborhood, city, and the regional and national scales. These four spatial levels are graphically shown in Fig. 2 based on their scale, with the housing having the smallest and regional and national levels having the biggest scale. It is important to note that the three dimensions of a resilient city to pandemics cover all four primary spatial levels identified through the systematic review. Furthermore, due to the constant emphasis on the importance of smart cities since the COVID-19 pandemic (Afrin et al., 2021; Harris et al., 2022; Jaiswal et al., 2020; Kunzmann, 2020; Sharifi et al., 2021), the role of smart cities cannot be ignored in times of pandemics and therefore, these dimensions are considered in the context of the smart city in the proposed framework, which will be discussed more in the following paragraphs.Fig. 2 A conceptual framework for resilient cities to pandemics.
Fig 2
A “smart city” is considered a high-tech intensive and advanced city that uses technology to link people, information, governance, economy, and city elements to create a sustainable, greener, and competitive cities with a higher quality of life (Bakıcı et al., 2013). Using smart city technologies has been considered influential in different aspects such as patient tracing (Afrin et al., 2021; Sonn et al., 2020), transportation (Gupta et al., 2020), social distancing, medical drones (Jaiswal et al., 2020), recognizing the outbreaks, determining the available resources, drone supply delivery, virtual communication, tracking patient numbers, predicting available hospitals (Inn, 2020), and monitoring facial mask practices (Rahman et al., 2020). However, smart city tools should be adapted based on pandemic disasters to ensure urban health. Allam and Jones (2020) highlighted the importance of standardization of protocols to improve smart city communication and democratization of technology to encourage equity and transparency and, eventually, more cooperation in times of disasters.
The triple dimensions, which were classified based on the literature review on pandemic-resilient cities and experiences from COVID-19, are explained in the following paragraphs.
3.1.1 Pandemic-related health requirements
In general, the design principles for health crises such as pandemics are different from other disasters since biological crises often threaten the health of communities (Litman, 2020). Pandemic-related health requirements refer to all measures that help prevent the transmission of viruses during pandemics. The experience of the COVID-19 outbreak showed that cities need to enable the implementation of health requirements related to infectious diseases to maintain the function of the urban environments. In other words, urban environments capable of implementing such measures would adapt to such a crisis quicker and better, therefore, showing a higher level of pandemic resiliency.
Although social distancing and lockdown were the key measures introduced by WHO (Salama, 2020), Megahed and Ghoneim (2020) emphasized reducing the population density since overcrowding in public areas in times of pandemics leads to unsanitary conditions and more spreading of infectious diseases. Moreover, the role of ventilation and airflow in airborne transmission of infectious disease, particularly in indoor spaces, was another health measure that was highlighted in the literature (Gao et al., 2009; Gu et al., 2020; Li et al., 2007). Smart technologies and indoor finishing materials (Megahed & Ghoneim, 2020; Tokazhanov et al., 2020; Van Doremalen et al., 2020) are other health measures mentioned in the literature.
3.1.2 Environmental psychological principles
One of the most critical consequences of pandemics is social anxiety. The level of anxiety, fear, and despair among people indicates the vulnerability of communities facing danger (Zabaniotou, 2020). Thus, the role of health psychology in responding to a pandemic and life changes should be understood to minimize the stress caused by a disease outbreak (Arden & Chilcot, 2020; Bish and Michie, 2010). Some of the psychological regulations mentioned in previous studies are proper governance and social support (Dhar et al., 2020), accessible recreational activities, online psychological support, expansion of online educational opportunities (Akat & Karatas, 2020), maintaining social relationships and connectedness even online (Thakur & Jain, 2020), and timely and adequate health information (Tee et al., 2020). However, considering the role of urban planning and design in improving the resilience of cities to pandemics, the present study focuses on the crucial role of environmental psychology in reducing people's stress level in cities. This dimension includes factors such as facilitating social interactions while maintaining social distancing (Johnson et al., 2021; Nitschke et al., 2021; Poortinga et al., 2021) and access to green and natural environment (Tokazhanov et al., 2020; Hartig et al., 2003; Velarde et al., 2007).
3.1.3 General resilience principles
A literature review on resilience shows that many researchers and institutions have provided resilience indicators. The general characteristics of resilience presented by researchers over time such as self-sufficiency, self-organization, decentralization, diversity, multi-functionality, flexibility, adaptability, modularity, connectivity, and inclusiveness (Ahern, 2011; Allan & Bryant, 2012; Dhar and Khirfan, 2016; Godschalk, 2003; Sharifi & Yamagata, 2015; Tanner et al., 2009; The Rockefeller Foundation, 2014; Toseroni et al., 2016; Tyler & Moench, 2012) are also applicable to the urban pandemic resilience.
The summary of resilient cities' requirements for pandemics is provided in Table 1 .Table 1 Summary of requirements of resilient cities to pandemics.
Table 1Category Subcategory Ref.
Pandemic-related health requirements Social distancing
Lockdown and quarantine
Reducing the population density
Indoor ventilation, air quality, temperature, and humidity
Smart technologies
Indoor finishing materials
Atalan, 2020, Guo et al., 2021, Melone and Borgo, 2020; Baser (2021); Bhadra et al. (2021); Block et al. (2020); Kadi and Khelfaoui (2020); Lee et al. (2021); Sy et al. (2021); Gao et al., 2009, Gu et al., 2020, Li et al., 2007, Megahed and Ghoneim, 2020, Tokazhanov et al., 2020, Van Doremalen et al., 2020, Wong and Li, 2020
Environmental psychological principles Maintaining social connections and facilitating social interaction
Access to green and natural environment
Hartig et al., 2003, Johnson et al., 2021, Nitschke et al., 2021, Poortinga et al., 2021, Tokazhanov et al., 2020, Velarde et al., 2007
General resilience principles Decentralization
Self-sufficiency
Adaptability
Flexibility
Diversity
Multi-functionality
Modularity
Connectivity
Redundancy Ahern (2011); Allan and Bryant (2012); Dhar and Khirfan (2016); Godschalk (2003); Sharifi and Yamagata (2015); Tanner et al. (2009); The Rockefeller Foundation, 2014; Toseroni et al. (2016); Tyler and Moench (2012)
3.2 Principles of urban resilience to pandemics
This section provides the principles of resilient cities to pandemics in four spatial levels: housing, neighborhood, city, and regional and national levels. Each level's principles also provide three subcategories (1) pandemic-related health requirements, (2) environmental psychological principles, and (3) general resilience principles. However, there were some overlaps between some principles, and some were common among two or three dimensions.
3.2.1 Housing
Historically, residential housing has been primarily designed to reflect the culture of its residents through construction, including the evolution of construction methods and approaches resulting from past disasters (Keenan, 2020). Therefore, reviewing the patterns and the housing codes is necessary to improve housing conditions to positively impact people's mental and physical health and their life quality during pandemics. According to the literature, the COVID-19 pandemic had valuable lessons for improving housing conditions during pandemics. The most important lessons are:
Pandemic-related health requirements: First, the COVID-19 pandemic emphasized the superiority of single-family housing with private open spaces, which provides the best environment and facilities for protective health measures, such as social distancing and the better use of light, fresh air, and nature (Megahed & Ghoneim, 2020). Although following this housing model may help solve the pandemic issues, it might result in urban sprawl. According to the resilience literature, urban planners and designers should keep advocating compact urban forms rather than sprawling ones due to the various merits of this form of urban development for urban resilience (Sharifi, 2019a; Sharifi, 2019b). Therefore, this study emphasizes that the best housing model for resilient cities to pandemics is multi-family housing, which involves the positive features of single-family housing, such as a private natural environment for each household and access to natural light and fresh air.
A comparison between the two types of housing is presented in Fig. 3 .Fig. 3 Comparison between the two types of housing: single-family detached housing and multi-family housing, which involves the positive features of single-family detached housing.
Fig 3
Second, such houses should benefit from new technologies and materials to provide specific protective health measures for their occupants, such as applying artificial intelligence and touchless technologies (Tokazhanov et al., 2020). In multi-story and high-rise buildings, where contact with other residents in shared spaces is unavoidable, intelligent technologies, such as touchless door entry systems, automatic doors, voice-activated elevators, and hands-free light switches, should be used in buildings to provide touchless equipment from the main entrance door to the apartment door. Such structures should also have more elevators and stairs with proper ventilation (Megahed & Ghoneim, 2020).
Third, antibacterial fabrics and materials on the surfaces. Antibacterial fabrics and finishes should cover buildings on surfaces to prevent the spread of viruses (Tokazhanov et al., 2020).
Fourth, regarding the design and layout of interior spaces, it is necessary to have more partitions between the areas so that in case of illness of any family member, it would be possible to quarantine the infected person. It is also recommended that residential housings have several separate bathrooms in case one family member gets infected (Tokazhanov et al., 2020).
Fifth, the proper ventilation and lighting of the interior of the housings are other essential factors in ensuring the health of residents (Li et al., 2007). This could be provided through both natural and artificial resources. However, natural airflow and lighting are more recommended.
Environmental psychological principles: First, due to the high stress level during pandemics and increased periods spent at home, one of the most critical principles is the inclusion of nature and airy spaces in residential environments. Natural elements such as plants, vegetation, and private green spaces help lower blood pressure and stress hormone levels and boost immunity (Hartig et al., 2003; Velarde et al., 2007).
Second, open or semi-open spaces such as courtyards, balconies, terraces, and accessible roofs can provide residents with areas to enjoy the fresh air and sunlight and engage in physical activities such as sports and games while maintaining social distancing. Such places act as buffer zones between the house and the unsafe outside (Melone & Borgo, 2020). They are an alternative to inaccessible public areas such as streets, urban squares, and parks in times of pandemics (Poortinga et al., 2021). They also satisfy the need for the “third place” to some extent (Banai, 2020). Therefore, these spaces prevent vulnerable groups' isolation and help improve social interactions among individuals.
General resilience principles: First, “adaptability” is one of the most critical features highly emphasized in the literature on resilience. In the literature post-COVID-19, the intimacy of social relationships for members of the family, who work remotely in spaces designed primarily for entertainment and domestic pursuits, was highlighted (Keenan, 2020). Lack of personal privacy and adequate housing space for work, study, and exercise can lead to a higher stress level for residents. Therefore, with the emerging need to work from home, designers should pay more attention to creating comfortable, isolated, and adaptive layouts in housing and multipurpose furniture (Tokazhanov et al., 2020).
Second, “self-sufficiency” is another feature highlighted in the literature on resilience (Ahern, 2011; Allan & Bryant, 2012; Dhar and Khirfan, 2016; Godschalk, 2003; Sharifi & Yamagata, 2015; Tanner et al., 2009; Tyler & Moench, 2012). As well as naturally filtering the air, green spaces would also provide residents with the opportunity to produce vegetables and fruits, leading to the relative self-sufficiency of households.
3.2.2 Neighborhoods
The importance of neighborhood design is heightened during the COVID-19 pandemic since residents are more willing and sometimes forced to spend more time at their houses and in their immediate neighborhoods (Miao et al., 2021). Studies show that neighborhoods with different socioeconomic features impact their residents differently during COVID-19, and not all people are at equal risk (Biggs et al., 2021; Hatef et al., 2020). Neighborhood socioeconomic characteristics, such as race, ethnicity, and income level, are associated with social vulnerability during the pandemic (Feldman & Bassett, 2020; Hatef et al., 2020).
Apart from socioeconomic features, several physio-spatial characteristics impact community resilience. The following paragraphs summarize the literature on neighborhood features that contribute to improved resiliency.
Pandemic-related health requirements: First, access to basic essential services, including living, working, commerce, healthcare, education, and entertainment facilities within a 15 min walking or cycling (Moreno et al., 2021). The concept of “15 min City”, which has been discussed frequently in the literature, emphasizes planning based on proximity to such services in a neighborhood (Allam et al., 2022; Balletto et al., 2021; Guzman et al., 2021; Pozoukidou & Chatziyiannaki, 2021).
According to the proponents of this concept, residents would experience a higher quality of life within a 15-min radius. Moreno et al. (2021) believe this model has different environmental, social, economic, and health benefits. The 15-min city implies a shift in the emphasis of planning from the neighborhoods' access to urban facilities to the proximity of urban facilities within neighborhoods (Pozoukidou & Chatziyiannaki, 2021). In the case of pandemics, the proximity of essential services would also decrease the need for communication within the cities, which was considered one of the major contributing factors to COVID-19 transmission during the pandemic (AbouKorin et al., 2021; Megahed & Ghoneim, 2020).
Second, urban green infrastructure and natural environments at different scales in neighborhoods improve air quality, provide safe spaces for different groups of residents, improve people's quality of life, and increase the possibility of social interactions among residents in times of pandemics (Jenkins, 2020).
Environmental psychological principles: Urban designers should provide a hierarchy of places, from public and semi-public to semi-private open spaces, in the design of neighborhoods to facilitate outdoor activities, allowing residents to exercise, play, and plant vegetation during pandemics. Such areas contribute to the physical health of residents by decreasing the adverse consequences of quarantine and the closure of cities on the individuals’ mental health, as well as preventing the congestion of public spaces on the scales beyond neighborhoods in the city (Lak et al., 2020).
General resilience principles: First, the COVID-19 pandemic showed that the best model for developing neighborhood structures is creating relatively independent neighborhood units/modules to provide the weekly basic needs. This idea is consistent with “self-sufficiency” and “modularity” criteria in the literature on resilience (Ahern, 2011; Allan & Bryant, 2012; Dhar and Khirfan, 2016; Godschalk, 2003; Sharifi & Yamagata, 2015; Tanner et al., 2009; Tyler & Moench, 2012). The opportunities for agricultural activities in the neighborhood can also lead to self-sufficiency in providing food for residents during these periods.
Second, the concept of traditional mixed-use neighborhoods is one of the basic requirements of resilient cities during pandemics. Providing communities with ample public facilities minimizes the need for traveling within the cities. This idea is consistent with diversity, one of the basic general principles of resilience (Ahern, 2011; Allan & Bryant, 2012; Dhar and Khirfan, 2016; Godschalk, 2003; Sharifi & Yamagata, 2015; Tyler & Moench, 2012). In addition, due to travel restrictions in cities in the first stages of the pandemic, essential services in these neighborhoods must be within walking and cycling distance from residential houses.
Fig. 4 shows the resilient neighborhood example.Fig. 4 Resilient neighborhood example.
Fig 4
3.2.3 City
Many researchers have analyzed the resilience of urban and environmental elements to pandemics such as COVID-19. These elements include the role of green spaces (Pan et al., 2021; Samuelsson et al., 2020), population density (Lee et al., 2021; Wong & Li, 2020), neighborhood and social vulnerability (Miao et al., 2021), trust in political leadership (Fernández-Prados et al., 2021), infrastructure and their adaptive functionality (Hynes et al., 2020), and information system (Sakurai & Chughtai, 2020). Thus, according to the literature, the COVID-19 pandemic has had valuable lessons for cities’ physical form and spatial structure. The most important lessons are:
Pandemic-related health requirements: First, the form of cities matters. In a study on European cities, AbouKorin et al. (2021) argued that city form was associated with the COVID-19 spread. Their study categorized cities' urban forms as linear, grid, and radial. They concluded that linear morphologies are linked to the lowest rates of infection. In contrast, cities with grid and radial forms had significantly higher infection rates during the COVID-19 pandemic.
Second, access to a green and natural environment is essential (Tokazhanov et al., 2020; Hartig et al., 2003; Velarde et al., 2007). Even though some researchers believe that a higher risk of infection accompanies more access to public green spaces as the possibility of interacting with people increases (Pan et al., 2021), many researchers found a positive relationship between green spaces and reduced risk of COVID-19 (Engemann et al., 2019; Hubbard et al., 2021; Orioli et al., 2019; Russette et al., 2021; Venter et al., 2021). Urban green space affects people's physical and mental health as well as the ecosystem (Ugolini et al., 2020). Green spaces are believed to have different impacts on health improvements (Engemann et al., 2019; Hubbard et al., 2021; Orioli et al., 2019) and are crucial health resources in times of crisis (Poortinga et al., 2021) by increasing happiness and life satisfaction, and decreased depression and loneliness in times of lockdowns (Soga et al., 2021). Killgore et al. (2020) emphasized the importance of green spaces and noted that the average resilience to COVID-19 is greater among people who can access green spaces more often. Majewska et al. (2022) declared that access to green spaces was essential to residents' quality of life in Polish towns and cities during the pandemic.
Poortinga et al. (2021) highlighted the importance of perceived public and private green space in people's health and well-being. Venter et al. (2021) reinforce the value of urban nature during and after a crisis and found a positive relationship between the lockdown in Oslo and the increasing usage of urban green infrastructure. In Italy, Ugolini et al. (2021) found an increased visit to nearby gardens and green spaces due to social distancing and other movement restrictions. Thus, crises such as COVID-19 highlight the values associated with public areas such as parks and natural environments (Keenan, 2020) since they can be accessible to those without a private garden (Poortinga et al., 2021). Therefore, plans for including green spaces and public spaces for leisure and recreation should be prioritized. Moreover, parks and green spaces should be located close to people, and accessibility should be considered for all users through various approaches, including bicycle and pedestrian connections (Slater et al., 2020).
Third, open and public spaces should be wide enough to provide social distance (Melone & Borgo, 2020). In addition, the appropriate width of the street and the general traffic flow also provide better access to medical centers and disease control, especially in times of illness (AbouKorin et al., 2021).
Environmental psychological principles: The diversity of open and semi-open urban spaces. Maintaining social connections is essential for our well-being during an unprecedented lockdown to prevent stress and fatigue (Nitschke et al., 2021). The variety and abundance of urban areas combined with parks and green spaces and their connection with pedestrian and bicycle paths in cities play an essential role in creating safe spaces for residents and the possibility of social interaction in pandemic situations (Johnson et al., 2021). Inclusive urban areas facilitate the presence of different groups, especially the elderly and sensitive groups, and prevent the isolation of people and possible mental illnesses, such as depression and anxiety.
General resilience principles: First, the “decentralization” of facilities and population (Pisano, 2020), as well as facilitating walkability and biking in cities, should be prioritized (Majewska et al., 2022; Moreno et al., 2021). Since the physical closeness between infected and non-infected people carries the highest risk, urban services, especially medical centers and hospitals, must be distributed at different scales in the city. In addition, in a pandemic, when there is a fear of public transport congestion due to the risk of getting the disease, walkability is considered one of the essential principles to preventing disruption of activities and daily life in cities (Banai, 2020). Furthermore, bicycling infrastructures and programs, especially the Bicycle Sharing System (BSS), play a vital role in meeting the transportation needs of citizens and are a viable alternative to public transportation, as they are compatible with social distancing (Chen et al., 2022; Teixeira & Lopes, 2020). Moreover, sustainable transportation options such as bicycles and facilitating walking in the city minimize air pollution, which can improve the condition of infected individuals.
Even though the decentralization of facilities and population is suggested in cities, there are contradictory views on the effects of density on the COVID-19 spread (Barak et al., 2021; Carozzi et al., 2020; Hong & Choi, 2021; Khavarian-Garmsir et al., 2021). On the one hand, some believe that population density is an effective predictor of infection (Atalan, 2020; Lee et al., 2021; Wong & Li, 2020), and COVID-19 transmission was faster in areas with higher density because of an increase in contact rate between people (Baser, 2021; Bhadra et al., 2021; Block et al., 2020; Kadi and Khelfaoui, 2020; Sy et al., 2021). On the other hand, some believe density is not significantly associated with the infection rate resulting from more social distancing guidelines and a better healthcare system (AbouKorin et al., 2021, Hamidi et al., 2020;; Gaisie et al., 2022). Majewska et al. (2022) argued that cities should have a compact structure with a high population density to reduce commuting during pandemics.
Second, the “self-sufficiency” of cities and towns is essential. Majewska et al. (2022) suggest that towns should follow a polycentric settlement network form, which as well as allocating places for living, provides jobs, access to essential frontline services within walking distance, and agriculture. Moreover, strengthening self-sufficient communities through urban farming would improve the resilience of cities to pandemics by improving food security, lowering stress, and improving the air quality in cities. Therefore, horizontal and vertical urban gardens should be flourished in urban areas (Megahed & Ghoneim, 2020).
Third, “adaptable”, “multi-functional”, or “flexible” spaces are the most critical features of resilient urban systems (Ahern, 2011; Allan & Bryant, 2012; Dhar and Khirfan, 2016; Godschalk, 2003; Sharifi & Yamagata, 2015; Tanner et al., 2009; The Rockefeller Foundation, 2014; Tyler & Moench, 2012). Flexible urban spaces, which provide different uses simultaneously, allow the city to face uncertainties and changes ahead and accept future usages that are not considered in the current situation (Dhar and Khirfan, 2016). Modifiable and adaptable spaces in the city provide the necessary uses, such as establishing temporary hospitals during pandemics.
Fourth, the “redundancy” of public facilities needs to be considered (Pisano, 2020). Redundancy means “having more options than necessary from an efficiency perspective” (Giezen et al., 2015: 169). It is one of the essential characteristics of resilient urban systems (Ahern, 2011; Godschalk, 2003; Sharifi & Yamagata, 2015; The Rockefeller Foundation, 2014). The provision of redundant services at different scales in cities not only facilitates the accessibility of services for all groups of people but also minimizes the need for traveling within the cities and the consequent congestion in certain areas, which is a critical factor in the transmission of the disease in the time of pandemics.
Fifth, some scholars also pointed out the need for a connected system of green spaces in cities to improve resilience in the face of pandemics (Eltarabily & Elghezanwy, 2020). “Connectivity” is also one of the general resilience principles in the literature (Ahern, 2011; Dhar and Khirfan, 2016).
3.2.4 Regional and national level
Due to the nature and interconnectivity of issues at the regional and national levels, it was impossible to categorize the principles of this level into the triple dimensions (pandemic-related health requirements, environmental-psychological principles, and general resilience principles). Thus, they are discussed without the triple categorization in the following paragraphs.
First, the critical role of the urban-rural interface and urban-rural linkages must be acknowledged. Mitra et al. (2021) emphasized the crucial role of urban-rural connection for the collective security of food, energy, and water during the COVID-19 pandemic. Some scholars also highlighted the importance of preventative measures focused on the urban-rural interface to reduce exposure and control the transmission of the viruses (Polo et al., 2022; Wells et al., 2020). Due to the unprecedented movement restrictions, which disrupt people's lives during a pandemic, Sukhwani and Shaw (2022) considered pandemics a crisis for human security. Thus, they believe the urban-rural linkage should be revisited from a human security perspective to protect the survival and livelihood of people living in urban and rural areas.
Second, the extent of local autonomy in decision-making and disaster management could be a key factor. Sharma et al. (2021) discuss that a centralized governance structure would not lead to a proactive response to a pandemic. Some studies argue that city and city region levels were at the front line of coordinated action and leadership on COVID-19 during the pandemic (Sharifi & Khavarian-Garmsir, 2020). Harris et al. (2022) asserted that governance at these levels is essential for engagement with the public about preparedness for and resilience to pandemics. In a study on modes of policy coordination and policy responses to COVID-19 in China and the USA, Liu et al. (2021) concluded that national leadership should be balanced with local autonomy and public engagement to achieve effective governance in crises like pandemics.
Third, the “decentralization” of infrastructure across the country is essential. The role of infrastructure, including healthcare, water, energy, transportation, and communication, in the resilience of cities to pandemics has been highlighted in the post-pandemic literature (Sharma et al., 2021; Syal, 2021). Inadequate infrastructure in different parts of the country can lead to a higher level of vulnerability in different cities and hence, the spread of the disease (Syal, 2021)
Fourth, since cities are increasingly interconnected due to globalization, one of the most important issues regarding this scale is the “connectivity” among different cities (Kummitha, 2020). This connectivity could have a detrimental effect on preventing the spread of the disease during pandemics. Hamidi et al. (2020) concluded that connectivity among different cities negatively impacts the early spread of an epidemic disease. Metropolitan areas with more economic, social, and commuting relationships are more vulnerable to infections than less connected cities.
4 Conclusion
The present study highlights the role of architects and urban planners in improving urban resilience against future pandemics. This research aimed to identify the primary dimensions that form urban resilience, the spatial scales in planning that urban planners and policymakers need to consider, and the measures required to be adopted to achieve pandemic-resilient cities. A qualitative archival method was applied to achieve these objectives, and a wide range of literature related to resilience, particularly pandemic resiliency of cities, was reviewed through a systematic review.
The literature review showed that first, the significant dimensions of resilient cities to pandemics include (1) pandemic-related health requirements, (2) environmental-psychological principles, and (3) general resilience principles. Moreover, the triple dimensions should be considered in the context of the smart city concept. Second, the spatial scales that urban planners and designers need to consider in planning for the resilience of cities to pandemics include housing, neighborhood, city, and regional and national levels. Finally, recommendations for building resilient cities to pandemics at all four levels and three dimensions were presented.
The summary of principles of resilient cities to pandemics is presented in Table 2 .Table 2 Summary of principles of resilient cities to pandemics.
Table 2Spatial levels Pandemic-related health requirements
(H) Environmental psychological principles
(P) General resilience principles
(R)
Housing (H) HH1: Multi-family housing, which involves a private natural environment for each household and private accessibility to natural lighting and fresh air
HH2: Adding artificial intelligence and touchless technological equipment
HH3: Using antibacterial fabrics and materials on the surfaces
HH4: More partitions in the layout design -Including several separate bathrooms
HH5: Proper ventilation and lighting
HH6: Different vertical access with proper ventilation facilities in multi-family housing HP1: More green spaces to increase interaction with nature
HP2: Including open or semi-open spaces in the design
HR1: Adaptive interior layout (Adaptability)
HR2: Private green spaces to produce vegetables and fruits (Self-sufficiency)
Neighborhood (N) NH1: Access to basic essential services, including living, working, commerce, healthcare, education, and entertainment facilities within a 15 min walking or cycling
NH2: Urban green infrastructure and natural environments at different scales
NP1: A hierarchy of territories, ranging from public and semi-public to semi-private open spaces, to facilitate outdoor activities NR1: Relatively independent neighborhood units/modules and providing opportunities to plant vegetables (Self-sufficiency and Modularity)
NR2: Mixed-use neighborhoods with diversity of public facilities and essential services (Diversity)
City (C) CH1: Linear morphologies
CH2: More public and private green spaces within the city limit
CH3: The appropriate width of public spaces to provide social distance and proper width of streets to facilitate better access to medical centers
CP1: Diversity of open or semi-open public spaces to prevent stress and isolation of people
CR1: Decentralization of population and facilities, as well as facilitating walkability and biking in cities (Decentralization)
CR2: Improving self-sufficiency through providing jobs, access to essential frontline services within walking distance, and urban agriculture (Self-sufficiency)
CR3: Modifiable and adaptable spaces in the city (Adaptability, Multi-functionality, and Flexibility)
CR4: Redundancy of public facilities (Redundancy)
CR5: A connected system of green spaces in cities (Connectivity)
Regional and national (R&N) R&N1: Urban-rural connection for the collective security of food, energy, and water during the pandemics, as well as preventative measures focused on the urban-rural interface to reduce the exposure and control the transmission of the viruses
R&N2: Local autonomy in decision making and disaster management
R&N3: Decentralization of infrastructure, including healthcare, water, energy, transportation, and communication across the country
R&N4: Less economic, social, and commuting relationships among different cities
While the present study identified improving the resilience of cities to pandemics should include a hierarchy of principles in four scales, including housing, neighborhood, city, and regional and national scales, some scholars stressed just one or two of the mentioned scales, such as city and architecture scales (Megahed & Ghoneim, 2020) or only housing scale (Tokazhanov et al., 2020) in their studies. However, Lak et al. (2020) pointed out the triple scales in their framework. Most strategies involved the neighborhood and city scale in their research. We argue that overemphasizing one aspect or scale and overshadowing one or two others might not result in resiliency as expected. This is mainly because studying COVID-19 merely on one spatial scale is problematic since the mobility across various scales and dynamic cross-scale interactions would lead to the transmission of the virus (Helbich et al., 2021). In addition, planners should not overlook the macro levels, such as regional and national scale, since nowadays, cities are increasingly interconnected due to globalization (Kummitha, 2020), which would negatively influence controlling the spread of viruses.
The framework introduced in this study help urban designers, planners, scholars, and policymakers have a more precise and comprehensive picture of resilient and anti-virus cities in the face of pandemics. In addition, the principles help policymakers adopt better measures to improve cities' resilience on different scales. By achieving a clearer perception of the components of resilient cities and their spatial scales, decision-makers can better focus on policies that increase cities' adaptive capacities and prevent virus spread during pandemics. As a result, the cities' economy and civil life would be less affected. Moreover, adopting such measures would also lead to higher levels of resiliency against other disasters and chronic hazards. Thus, this study suggests that researchers, practitioners, and policymakers focus on the presented framework and the principles in the four spatial scales to make better-informed decisions regarding resilience initiatives.
We recommend that future researches focus more on developing design principles, standards, and disaster management protocols for commercial zones and public spaces in case of biological disasters such as pandemics to maintain the economy and vitality of cities and minimize the risk to the health and well-being of residents. Measuring the resilience of the built environment, such as buildings, neighborhoods, and urban public spaces, against pandemics is another topic that needs to be studied in future research. Finally, there is no unanimous agreement regarding urban and population density and its relationship with spreading infectious diseases. Thus, more data is needed from different case studies to show whether or not higher or lower density can directly affect the spread of a contagious disease.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Data Availability
No data was used for the research described in the article.
==== Refs
References
AbouKorin S.A.A. Han H. Mahran M.G.N. Role of urban planning characteristics in forming pandemic resilient cities–Case study of Covid-19 impacts on European cities within England, Germany and Italy Cities 118 2021 103324 (London, England)
Afrin S. Chowdhury F.J. Rahman M. Covid-19 pandemic: Rethinking strategies for resilient urban design, perceptions, and planning Frontiers in Sustainable Cities 32 2021 10.3389/frsc.2021.668263
Ahern J. From fail-safe to safe-to-fail: Sustainability and resilience in the new urban world Landscape and Urban Planning 100 2011 341 343
Akat M. Karataş K. Psychological effects of COVID-19 pandemic on society and its reflections on education Electron c Turkish Studies 15 4 2020 10.7827/TurkishStudies.44336
Allam Z. Jones D.S. Pandemic stricken cities on lockdown. Where are our planning and design professionals [now, then and into the future]? Land use policy 97 2020 10.1016/j.landusepol.2020.104805
Allam Z. Nieuwenhuijsen M. Chabaud D. Moreno C. The 15-minute city offers a new framework for sustainability, liveability, and health The Lancet Planetary Health 6 3 2022 e181 e183 35278381
Allan P. Bryant M. Resilience as a framework for urbanism and recovery Journal of Landscape Architecture 6 2 2012 34 45
Amirzadeh M. Barakpour N. Developing a framework for community resilience to drought in Isfahan through qualitative research method and ATLAS-ti software Journal of Environmental Studies 44 4 2019 763 781 10.22059/JES.2019.269819.1007777
Amirzadeh M. Barakpour N. Evaluating the resilience of local communities in Isfahan to Zayandehrood River and Madies’ drying up Journal of Environmental Science and Technology 2019 10.22034/JEST.2019.43822.4637
Amirzadeh M. Barakpour N. Strategies for community resilience against slow-onset hazards International Journal of Disaster Risk Reduction 66 2021 102599 10.1016/j.ijdrr.2021.102599
Amirzadeh M. Sobhaninia S. Sharifi A. Urban resilience: A vague or an evolutionary concept? Sustainable Cities and Society 2022 103853 10.1016/j.scs.2022.103853
Arden M.A. Chilcot J. Health psychology and the coronavirus (COVID-19) global pandemic: A call for research British Journal of Health Psychology 25 2 2020 231 232 32227547
Atalan A. Is the lockdown important to prevent the COVID-19 pandemic? Effects on psychology, environment and economy-perspective Annals of Medicine and Surgery 56 2020 38 42 32562476
Bakıcı T. Almirall E. Wareham J. A smart city initiative: The case of Barcelona Journal of the Knowledge Economy 4 2 2013 135 148
Balletto G. Ladu M. Milesi A. Borruso G. A methodological approach on disused public properties in the 15-minute city perspective Sustainability 13 2 2021 593 10.3390/su13020593
Banai R. Pandemic and the planning of resilient cities and regions Cities 106 2020 102929 10.1016/j.cities.2020.102929 (London, England)
Barak N. Sommer U. Mualam N. Urban attributes and the spread of COVID-19: The effects of density, compliance and socio-political factors in Israel Science of The Total Environment. 2021 148626 10.1016/j.scitotenv.2021.148626
Baser O. Population density index and its use for distribution of Covid-19: A case study using Turkish data Health Policy 125 2 2021 148 154 (Amsterdam, Netherlands) 33190934
Bhadra A. Mukherjee A. Sarkar K. Impact of population density on Covid-19 infected and mortality rate in India Modeling Earth Systems and Environment 7 1 2021 623 629 33072850
Biggs E.N. Maloney P.M. Rung A.L. Peters E.S. Robinson W.T. The relationship between social vulnerability and COVID-19 incidence among Louisiana census tracts Frontiers in Public Health 8 2021 1048 10.3389/fpubh.2020.617976
Bish A. Michie S. Demographic and attitudinal determinants of protective behaviours during a pandemic: A review British Journal of Health Psychology 15 2010 797 824 10.1348/135910710×485826 20109274
Block P. Hoffman M. Raabe I.J. Dowd J.B. Rahal C. Kashyap R. Social network-based distancing strategies to flatten the COVID-19 curve in a post-lockdown world Nature Human Behaviour 4 6 2020 588 596
Buckman S. Rakhimova N. The Resilient City: Combating Uncertainty within an Urban Context Hagen B. Pijawka D. Sustainability in the 21st century: pathways, programs and policies 3rd ed 2020 Kendall Hunt Publishing 129 148 Ps
Buckman S. Sobhaninia S. The impact of sea-level flooding on the real estate development community in Charleston, SC: Results of a ULI member survey Journal of Sustainable Real Estate 14 1 2022 4 20 10.1080/19498276.2022.2095699
Carozzi F. Provenzano S. Roth S. Urban density and COVID-19 2020 Institute of Labor Economics (IZA). Bonn IZA Discussion Papers, No. 13440
Chen Y. Sun X. Deveci M. Coffman D.M. The Impact of the COVID-19 Pandemic on the Behaviour of Bike Sharing Users Sustainable Cities and Society 2022 104003 10.1016/j.scs.2022.104003
Davic R.D. Welsh H.H. On the ecological roles of salamanders Annual Review of Ecology, Evolution, and Systematic 35 2004 405 434
Desouza K.C. Flanery T.H. Designing, planning, and managing resilient cities: A conceptual framework Cities 35 2013 89 99 (London, England)
Dhar T. Khirfan L. A multi-scale and multi-dimensional framework for enhancing the resilience of urban form to climate change Urban Climate 19 2016 72 91
Dhar B.K. Ayittey F.K. Sarkar S.M. Impact of COVID-19 on Psychology among the University Students Global Challenges 4 11 2020 2000038 10.1002/gch2.202000038
Eltarabily S. Elghezanwy D. Post-pandemic cities-the impact of COVID-19 on cities and urban design Architecture Research 10 3 2020 75 84 10.5923/j.arch.20201003.02
Engemann K. Pedersen C.B. Arge L. Tsirogiannis C. Mortensen P.B. Svenning J.C. Residential green space in childhood is associated with lower risk of psychiatric disorders from adolescence into adulthood Proceedings of the National Academy of Sciences 116 11 2019 5188 5193
Feldman, J. M., & Bassett, M. T. 2020. The relationship between neighborhood poverty and COVID-19 mortality within racial/ethnic groups (Cook County, Illinois). medRxiv 2020.10. 04.20206318[Preprint]. 6 October 2020. Google Scholar.
Gupta, M., Abdelsalam, M., & Mittal, S. (2020). Enabling and enforcing social distancing measures using smart city and its infrastructures: A COVID-19 Use case. arXiv preprint arXiv:2004.09246.
Fernández-Prados J.S. Lozano-Díaz A. Muyor-Rodríguez J. Factors explaining social resilience against COVID-19: The case of Spain European Societies 23 2021 S111 S121 10.1080/14616696.2020.1818113 SUp 1
Gaisie E. Oppong-Yeboah N.Y. Cobbinah P.B. Geographies of infections: Built environment and COVID-19 pandemic in metropolitan Melbourne Sustainable Cities and Society 81 2022 103838 10.1016/j.scs.2022.103838
Gao X. Li Y. Leung G.M. Ventilation control of indoor transmission of airborne diseases in an urban community Indoor and Built Environment 18 3 2009 205 218
Giezen M. Salet W. Bertolini L. Adding value to the decision-making process of mega projects: Fostering strategic ambiguity, redundancy, and resilience Transport Policy 44 2015 169 178
Godschalk D.R. Urban hazard mitigation: creating resilient cities Natural Hazard Review 2003 136 143
Gu D. Zheng Z. Zhao P. Xie L. Xu Z. Lu X. High-efficiency simulation framework to analyze the impact of exhaust air from covid-19 temporary hospitals and its typical applications Applied Sciences: 10 11 2020 3949
Guo C. Chan S.H.T. Lin C. Zeng Y. Bo Y. Zhang Y. Physical distancing implementation, ambient temperature and Covid-19 containment: An observational study in the United States Science of the Total Environment 789 2021 147876 10.1016/j.scitotenv.2021.147876
Guzman L.A. Arellana J. Oviedo D. Aristizábal C.A.M. COVID-19, activity and mobility patterns in Bogotá. Are we ready for a ‘15-minute city’? Travel Behaviour and Society 24 2021 245 256 10.1016/j.tbs.2021.04.008
Hamidi S. Sabouri S. Ewing R. Does density aggravate the COVID-19 pandemic? Journal of the American Planning Association 86 4 2020 495 509 10.1080/01944363.2020.1777891
Harris P. Harris-Roxas B. Prior J. Morrison N. McIntyre E. Frawley J. Respiratory pandemics, urban planning and design: A multidisciplinary rapid review of the literature Cities 127 2022 103767 10.1016/j.cities.2022.103767 (London, England)
Hartig T. Evans G.W. Jamner L.D. Davis D.S. Garling T. Tracking restoration in natural and urban field settings Journal of Environmental Psychology 23 2003 109 123
Hatef E. Chang H.Y. Kitchen C. Weiner J.P. Kharrazi H. Assessing the impact of neighborhood socio-economic characteristics on COVID-19 prevalence across seven states in the United States Frontiers in public health 8 2020 10.3389/fpubh.2020.571808
Helbich M. Browning M.H.M. Kwan M.P. Time to address the spatiotemporal uncertainties in COVID-19 research: Concerns and challenges Science of the Total Environment 764 2021 142866 10.1016/j.scitotenv.2020.142866
Holling C. Resilience and stability of ecological systems Annual Review of Ecology and Systematics 4 1973 1 23
Hong S. Choi S.H. The urban characteristics of high economic resilient neighborhoods during the COVID-19 pandemic: A case of Suwon, South Korea Sustainability 13 9 2021 4679 10.3390/su13094679
Hubbard G. Daas C.D. Johnston M. Murchie P. Thompson C.W. Dixon D. Are rurality, area deprivation, access to outside space, and green space associated with mental health during the COVID-19 pandemic? a cross-sectional study (CHARIS-E) International Journal of Environmental Research and Public Health 18 8 2021 3869 10.3390/ijerph18083869 33917067
Inn T.L. Smart city technologies take on COVID-19 World health 841 2020
Jabareen Y. Planning the resilient city: concepts and strategies for coping with climate change and environmental risk Cities 31 2013 220 229 (London, England)
Jaiswal R. Agarwal A. Negi R. Smart solution for reducing the COVID-19 risk using smart city technology IET Smart Cities 2 2 2020 82 88
Jenkins A. Biotic systems as a critical urban infrastructure during crisis: Learning from the COVID-19 pandemic Cities and Health. 2020 10.1080/23748834.2020.1789821
Johnson T.F. Hordley L.A. Greenwell M.P. Evans L.C. Associations between COVID-19 transmission rates, park use, and landscape structure Science of The Total Environment 789 2021 10.1016/j.scitotenv.2021.148123
Kadi N. Khelfaoui M. Population density, a factor in the spread of COVID-19 in Algeria: Statistic study Bulletin of the National Research Centre 44 1 2020 1 7
Keenan J.M. COVID, resilience, and the built environment Environment systems and decisions. 2020 10.1007/s10669-020-09773-0
Khavarian-Garmsir A.R. Sharifi A. Moradpour N. Are high-density districts more vulnerable to the COVID-19 pandemic? Sustainable Cities and Society 70 2021 102911 10.1016/j.scs.2021.102911
Killgore W.D. Taylor E.C. Cloonan S.A. Dailey N.S. Psychological resilience during the COVID-19 lockdown Psychiatry Research 291 2020 113216 10.1016/j.psychres.2020.113216
Kummitha R.K.R. Smart technologies for fighting pandemics: The techno- and human-driven approaches in controlling the virus transmission Government Information Quarterly 2020 10.1016/j.giq.2020.101481
Kunzmann K.R. Smart cities after COVID-19: Ten narratives disP-The Planning Review 56 2 2020 20 31
Lai K.Y. Webster C. Kumari S. Sarkar C. The nature of cities and the Covid-19 pandemic Current Opinion in Environmental Sustainability 20 2020 1 5 10.1016/j.cosust.2020.08.008
Lak A. Shakouri Asl Sh Maher A. Resilient urban form to pandemics: Lessons from COVID-19 Medical Journal of The Islamic Republic of Iran 34 1 2020 502 509 10.34171/mjiri.34.71
Lee W. Kim H. Choi H.M. Heo S. Fong K.C. Yang J. Urban environments and COVID-19 in three Eastern states of the United States Science of The Total Environment 779 2021 146334 10.1016/j.scitotenv.2021.146334
Leichenko R. Climate change and urban resilience Current Opinion in Environmental Sustainability 3 2011 164 168
Li Y. Leung G.M. Tang J.W. Yang X. Chao C.Y. Lin J.Z. Role of ventilation in airborne transmission of infectious agents in the built environment-a multidisciplinary systematic review Indoor air 17 1 2007 2 18 17257148
Litman T. Pandemic-resilient community planning. practical ways to help communities prepare for, respond to, and recover from pandemics and other economic, social and environmental shocks Victoria: Victoria Transport Policy Institute 2020 www.vtpi.org
Liu Z. Guo J. Zhong W. Gui T. Multi-level governance, policy coordination and subnational responses to COVID-19: Comparing China and the US Journal of Comparative Policy Analysis: Research and Practice 23 2 2021 204 218
Hynes W. Trump B. Love P. Linkov I. Bouncing forward: A resilience approach to dealing with COVID-19 and future systemic shocks Environment Systems and Decisions 40 2020 174 184 32837818
Ludwig D. Walker B. Holling C.S. Sustainability, stability, and resilience Conservation Ecology 1 1 1997 1 21
Majewska A. Denis M. Jarecka-Bidzińska E. Jaroszewicz J. Krupowicz W. Pandemic resilient cities: Possibilities of repairing Polish towns and cities during COVID-19 pandemic Land Use Policy 113 2022 105904
Meerow S. Newell J.P. Stults M. Defining urban resilience: a review Landscape and Urban Planning 147 2016 38 49
Megahed N.A. Ghoneim E.M. Antivirus-built environment: Lessons learned from Covid-19 pandemic Sustainable Cities and Society 61 2020 102350 10.1016/j.scs.2020.102350
Melone M.R.S. Borgo S. Rethinking rules and social practices. The design of urban spaces in the post-Covid-19 lockdown TeMA-Journal of Land Use, Mobility and Environment 2020 333 341
Miao J. Zeng D. Shi Z. Can neighborhoods protect residents from mental distress during the COVID-19 pandemic? Evidence from Wuhan Chinese Sociological Review 53 1 2021 1 26
Mitra P. Shaw R. Sukhwani V. Mitra B.K. Rahman M.A. Deshkar S. Urban–rural partnership framework to enhance food–energy–water security in the post-COVID-19 era International journal of environmental research and public health 18 23 2021 12493 34886218
Moher D. Liberati A. Tetzlaff J. Altman D.G. PRISMA Group Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement Research Methods and Reporting 6 7 2009 E1000097 10.1136/bmj.b2535
Moreno C. Allam Z. Chabaud D. Gall C. Pratlong F. Introducing the “15-Minute City”: Sustainability, resilience and place identity in future post-pandemic cities Smart Cities 4 1 2021 93 111
Nitschke J.P. Forbes P.A. Ali N. Cutler J. Apps M.A. Lockwood P.L. Resilience during uncertainty? Greater social connectedness during COVID-19 lockdown is associated with reduced distress and fatigue British Journal of Health Psychology 26 2 2021 553 569 33099800
Orioli R. Antonucci C. Scortichini M. Cerza F. Marando F. Ancona C. Exposure to residential greenness as a predictor of cause-specific mortality and stroke incidence in the Rome Longitudinal Study Environmental Health Perspectives 127 2 2019 027002 10.1289/EHP2854
Orleans Reed S. Friend R. Toan V.C. Thinphanga P. Sutrato R. Singh D. Shared learning" for building urban climate resilience – experiences from Asian cities Environment and Urbanization 25 2 2013 393 412
Pan J. Bardhan R. Jin Y. Spatial distributive effects of public green space and COVID-19 infection in London Urban Forestry and Urban Greening 62 2021 127182 10.1016/j.ufug.2021.127182
Pimm S.L. The balance of nature?: Ecological issues in the conservation of species and communities 1991 University of Chicago press
Pisano C. Strategies for post-COVID cities: An insight to Paris En Commun and Milano 2020 Sustainability 12 2020 5883 10.3390/su12155883
Polo G. Soler-Tovar D. Jimenez L.V. Benavides-Ortiz E. Acosta C.M. SARS-CoV-2 transmission dynamics in the urban-rural interface Public Health 206 2022 1 4 35306192
Poortinga W. Bird N. Hallingberg B. Phillips R. Williams D. The role of perceived public and private green space in subjective health and well-being during and after the first peak of the COVID-19 outbreak Landscape and Urban Planning 211 2021 104092 10.1016/j.landurbplan.2021.104092
Pozoukidou G. Chatziyiannaki Z. 15-Minute City: Decomposing the new urban planning eutopia Sustainability 13 2 2021 928 10.3390/su13020928
Rahman M.M. Manik M.M.H. Islam M.M. Mahmud S. Kim J.H. An automated system to limit COVID-19 using facial mask detection in smart city network 2020 Proceedings of the IEEE international IOT, electronics and mechatronics conference (IEMTRONICS) 2020 IEEE 1 5 pp.
Russette H. Graham J. Holden Z. Semmens E.O. Williams E. Landguth E.L. Greenspace exposure and COVID-19 mortality in the United States: January–July 2020 Environmental Research 198 2021 111195 10.1016/j.envres.2021.111195
Sakurai M. Chughtai H. Resilience against crises: COVID-19 and lessons from natural disasters European Journal of Information Systems 29 5 2020 585 594 10.1080/0960085X.2020.1814171
Salama A.M. Coronavirus questions that will not go away: Interrogating urban and socio-spatial implications of COVID-19 measures Emerald Open Research 2 14 2020 10.35241/emeraldopenres.13561.1
Samuelsson K. Barthel S. Colding J. Macassa G. Giusti M. Urban nature as a source of resilience during social distancing amidst the coronavirus pandemic Landscape and Urban Planning 2020 10.31219/osf.io/3wx5a
Shakil M.H. Munim Z.H. Tasnia M. Sarowar S. COVID-19 and the environment: A critical review and research agenda Science of the Total Environment 745 2020 141022 10.1016/j.scitotenv.2020.141022
Sharifi A. Khavarian-Garmsir A.R. The COVID-19 pandemic: Impacts on cities and major lessons for urban planning, design, and management Science of the Total Environment 749 2020 142391 10.1016/j.scitotenv.2020.142391
Sharifi A. Yamagata Y. A Conceptual Framework for Assessment of Urban Energy Resilience Energy Procedia 75 2015 2904 2909
Sharifi A. Yamagata Y. Principles and criteria for assessing urban energy resilience : A literature review Renewable and Sustainable Energy Reviews 60 2016 1654 1677
Sharifi A. Yamagata Y. Resilience-oriented urban planning Resilience-oriented urban planning 2018 Springer Cham 3 27
Sharifi A. Yamagata Y. Resilient urban form: A conceptual framework Lecture Notes in Energy 65 2018 10.1007/978-3-319-75798-8_9
Sharifi A. Khavarian-Garmsir A.R. Kummitha R.K.R. Contributions of smart city solutions and technologies to resilience against the COVID-19 pandemic: A literature review Sustainability 13 14 2021 8018 10.3390/su13148018
Sharifi A. Resilient urban forms: A macro-scale analysis Cities 85 2019 1 14 (London, England)
Sharifi A. Urban form resilience: A meso-scale analysis Cities 93 2019 238 252 (London, England)
Sharma A. Borah S.B. Moses A.C. Responses to COVID-19: The role of governance, healthcare infrastructure, and learning from past pandemics Journal of business research 122 2021 597 607 33518844
Slater S.J. Christiana R.W. Gustat J. Recommendations for keeping parks and green space accessible for mental and physical health during COVID-19 and other pandemics Preventing Chronic Disease 17 2020 E59 10.5888/pcd17.200204 32644919
Sobhaninia S. Buckman S.T. Revisiting and adapting the Kates-Pijawka disaster recovery model: A reconfigured emphasis on anticipation, equity, and resilience International Journal of Disaster Risk Reduction 69 1 2022 102738 10.1016/j.ijdrr.2021.102738
Soga M. Evans M.J. Tsuchiya K. Fukano Y. A room with a green view: The importance of nearby nature for mental health during the COVID-19 pandemic Ecological Applications 31 2 2021 E02248 10.1002/eap.2248
Sonn J.W. Kang M. Choi Y. Smart city technologies for pandemic control without lockdown International Journal of Urban Sciences 24 2 2020 149 151 10.1080/12265934.2020.1764207
Sukhwani V. Shaw R. Urban–Rural Linkages and Their Implication to Human Security in Pandemic Time Global pandemic and human security 2022 Springer Singapore 147 163
Sy K.T.L. White L.F. Nichols B.E. Population density and basic reproductive number of COVID-19 across United States counties PloS ONE 16 4 2021 E0249271 10.1371/journal.pone.0249271
Syal S. Learning from pandemics: Applying resilience thinking to identify priorities for planning urban settlements Journal of Urban Management 10 3 2021 205 217
Tanner T. Mitchell T. Polack E. Guenther B. Urban governance for adaptation: Assessing climate change resilience in ten Asian Cities 2009 Institute of Development Studies Brighton
Tee M.L. Tee C.A. Anlacan J.P. Aligam K.J.G. Reyes P.W.C. Kuruchittham V. Psychological impact of COVID-19 pandemic in the Philippines Journal of Affective Disorders 277 2020 379 391 32861839
Teixeira J.F. Lopes M. The link between bike sharing and subway use during the COVID-19 pandemic: The case study of New York's Citi Bike Transportation Research Interdisciplinary Perspectives 6 2020 100166 10.1016/j.trip.2020.100166
Thakur V. Jain A. COVID 2019-suicides: A global psychological pandemic Brain, Behavior, and Immunity 88 2020 952 953 10.1016/j.bbi.2020.04.062 32335196
The Rockefeller Foundation City resilience framework 2014 The Rockefeller Foundation /Arup International Development New York
Tokazhanov G. Tleuken A. Guney M. Turkyilmaz A. Karaca F. How is COVID-19 experience transforming sustainability requirements of residential buildings? A Review. Sustainability 12 2020 8732 10.3390/su12208732
Toseroni F. Romagnoli F. Marincioni F. Adapting and reacting to measure an extreme event: A methodology to measure disaster community resilience 95 2016 491 498 In: Proceedings of the international conference Environmental and climate technologies. CONNECT 2015Riga
Tricco A.C. Lillie E. Zarin W. O'Brien K.K. Colquhoun H. Levac D. PRISMA extension for scoping reviews (PRISMA-ScR): Checklist and explanation Annals of Internal Medicine 169 7 2018 467 473 30178033
Tyler S. Moench M. A framework for urban climate resilience Climate and Development 4 4 2012 311 326
Ugolini F. Massetti L. Pearlmutter D. Sanesi G. Usage of urban green space and related feelings of deprivation during the COVID-19 lockdown: Lessons learned from an Italian case study Land Use Policy 105 2021 105437 10.1016/j.landusepol.2021.105437
Van Doremalen N. Bushmaker T. Morris D.H. Holbrook M.G. Gamble A. Williamson B.N. Aerosol and surface stability of SARS-CoV-2 as compared with SARS-CoV-1 The New England Journal of Medicine 382 2020 1564 1567 32182409
Velarde M. Fry G. Tveit M. Health effects of viewing landscapes–Landscape types in environmental psychology Urban Forestry & Urban Greening 6 2007 199 212
Venter Z.S. Barton D.N. Figari H. Nowell M.S. Back to nature: Norwegians sustain increased recreational use of urban green space months after the COVID-19 outbreak Landscape and Urban Planning 214 2021 104175 10.1016/j.landurbplan.2021.104175
Wells K. Lurgi M. Collins B. Lucini B. Kao R.R. Lloyd AL. … Disease control across urban–rural gradients Journal of the Royal Society Interface 17 173 2020 20200775
Wong D.W. Li Y. Spreading of COVID-19: Density matters PloS ONE 15 12 2020 E0242398 10.1371/journal.pone.0242398
World Health Organization Managing epidemics: key facts about major deadly diseases 2018 World Health Organization Luxembourg
World Health Organization WHO coronavirus disease (COVID-19) dashboard 2020 World Health Organization https://covid19.who.int
World Health Organization WHO coronavirus disease (COVID-19) dashboard 2022 World Health Organization https://covid19.who.int
Zabaniotou A. A systemic approach to resilience and ecological sustainability during the COVID-19 pandemic: Human, societal, and ecological health as a system-wide emergent property in the Anthropocene Global Transitions 2 2020 116e126 10.1016/j.glt.2020.06.002 32835203
| 36467253 | PMC9703866 | NO-CC CODE | 2022-12-02 23:16:59 | no | Sustain Cities Soc. 2023 Feb 28; 89:104326 | utf-8 | Sustain Cities Soc | 2,022 | 10.1016/j.scs.2022.104326 | oa_other |
==== Front
Alexandria Engineering Journal
1110-0168
1110-0168
THE AUTHORS. Published by Elsevier BV on behalf of Faculty of Engineering, Alexandria University.
S1110-0168(21)00351-3
10.1016/j.aej.2021.04.104
Article
A mathematical COVID-19 model considering asymptomatic and symptomatic classes with waning immunity
Anggriani Nursanti a⁎
Ndii Meksianis Z. b
Amelia Rika a
Suryaningrat Wahyu a
Pratama Mochammad Andhika Aji a
a Department of Mathematics, Universitas Padjadjaran, Jln. Raya Bandung-Sumedang Km. 21 Jatinangor, Kab. Sumedang, 45363 Jawa Barat, Indonesia
b Department of Mathematics, Faculty of Sciences and Engineering, The University of Nusa Cendana, Kupang-NTT, Indonesia
⁎ Corresponding author at: Department of Mathematics, Universitas Padjadjaran, Jln. Raya Bandung-Sumedang Km. 21 Jatinangor, Kab. Sumedang 45363 Jawa Barat, Indonesia.
14 5 2021
1 2022
14 5 2021
61 1 113124
1 2 2021
21 4 2021
27 4 2021
© 2021 THE AUTHORS
2021
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
The spread of COVID-19 to more than 200 countries has shocked the public. Therefore, understanding the dynamics of transmission is very important. In this paper, the COVID-19 mathematical model has been formulated, analyzed, and validated using incident data from West Java Province, Indonesia. The model made considers the asymptomatic and symptomatic compartments and decreased immunity. The model is formulated in the form of a system of differential equations, where the population is divided into seven compartments, namely Susceptible Population (S0), Exposed Population (E), Asymptomatic Infection Population (IA), Symptomatic Infection Population (YS), Recovered Population (Z), Susceptible Populations previously infected (Z0), and Quarantine population (Q). The results show that there has been an outbreak of COVID-19 in West Java Province, Indonesia. This can be seen from the basic reproduction number of this model, which is 3.180126127 (R0>1). Also, the numerical simulation results show that waning immunity can increase the occurrence of outbreaks; and a period of isolation can slow down the process of spreading COVID-19. So if a strict social distancing policy is enforced like a quarantine, the outbreak will lessen.
Keywords
COVID-19
Basic Reproduction Ratio
Waning immunity
Asymptomatic
Previous infection
Parameter estimation
==== Body
pmc1 Introduction
The spread of COVID-19 has shocked society and currently has transmitted to more than 200 countries [1]. As of 04 April 2021, there are 130,998,190 confirmed cases, 2,853,280 death, and 105,447,782 recovered individuals [2]. It has caused severe economic and social loss. The disease has been transmitted from human to human via droplets [3]. Infected individuals may show symptomps such as fever, cough, sore throat, rhinorrhea, myalgia or fatigue, phlegm, and headache [3], [4], [5] with the body temperature of 39°C or above [5]. Individuals who are infected by COVID-19 can show symptoms (symptomatic) or cannot show symptoms (asymptomatic) but both types of individuals can transmit disease [3]. The incubation period has been estimated between two fourteen days [6].
Research showed that there is possibility for infected individuals to be reinfected by COVID-19. Currently it has been found that several recovered individuals have been re-infected by COVID-19 and this can cause death from fatal heart failure [7]. Of the 111 recovered patients, 5% of China and 10% of South Korea tested positive for COVID-19 [8]. This situation contradicts the fact that after a person catches the virus and then recovers, the individual will forman antibody that prevents the same virus from attacking twice. Research showed that reinfected individuals have experienced viral replication but did not neutralize antibodies, which implies that it is unlikely that long-term protective immunity will occur in people with COVID-19 after the first infection [9]. The virus’s immune response can be reduced within four months to one year after infection [10]. The genetic basis of the innate immune response affects the severity of COVID-19, it can also lead to more severe reinfection depending on antibodies generated against the bound virus but cannot neutralize the same strain [10]. The reinfection COVID-19 case has a more severe impact [11]. The reinfection occurs due to the decrease in the individual’s immunity [12]. Understanding the effects of waning immunity is important.
Mathematical models can be used to understand the complex phenomena such as population dynamics problem [13], [14], [15], [16] and disease transmission dynamics [17], [18], [19], [20], [21]. A compartment-based epidemic model in the form of system of (fractional or integer) differential equations has been formulated to understand disease transmission dynamics, where the human population is divided into different stages according to their status to the diseases [22], [23], [24], [25], [26], [27], [28], [29], [30], [31], [32], [33], [34], [35], [36]. A mathematical SEIR model is mostly used as a basis for the model’s development for COVID-19 transmission [37], [38]. The SEIR model has been extended to include quarantine compartment [39], to include symptomatic and asymptomatic classes [40]. The models are mostly used to investigate the disease transmission dynamics in several countries or provinces such as Indonesia [41], Hubei has been researched [42], Pakistan [43]. In this paper, a modified SEIR model considering symptomatic and asymptomatic cases from [44] has been formulated. The work focuses on studying the effects of waning immunity. The model is validated against data of COVID-19 incidence from West Java Province. The basic reproduction number is calculated, and a global sensitivity analysis is performed. The model is then used to determine he effects of waning immunity or reduced immunity to an increase in the number of infected individuals.
The remainder of this paper is organized as follows. In Section 2, the construction of the SEIR compartmental model. Next, the model’s mathematical properties, such as the equilibrium points, Basic Reproduction Number, and the existence of backward bifurcation, are detailed in Section 3. In Section 4, we explain the real-world problem using the incidence data of West Java Province, Indonesia. A discussion on the Basic Reproduction Number and the sensitivity analysis results are provided in Section 5. Finally, some conclusions are presented in Section 6.
2 Model Formulation
We developed model of transmission of COVID-19 by considering asymptomatic and symptomatic compartments and decreased immunity. The total population is divided into Susceptible population (S0), Exposed population (E), Asymptomatic infected population (IA), Symptomatic infected population (YS), Recovered population (Z), Susceptible that previously infected (Z0), Quarantine population (Q). The total number of population at time t is given by:N(t)=S0(t)+E(t)+IA(t)+YS(t)+Z(t)+Z0(t)+Q(t).
The assumption used in the formulation of a mathematical model for the spread of the COVID-19 disease is that individuals with symptoms will undergo hospitalization or quarantine. Deaths experienced by latent, symptomatic, asymptomatic, and quarantine individuals are caused by disease [45]. This means that the death of the three individuals is a combination of natural death and death due to disease. We assume that the μ1 parameter contained in compartments E,Ia,Ys is a death caused by COVID-19 plus a natural death factor. People who have decreased immunity can catch COVID-19 again with a high severity [11]. Hence, the second person infected will develop symptoms and be hospitalized. The model is represented by the diagrams shown in Fig. 1 , with the description of the parameters given in Table 1 .Fig. 1 Interaction Diagram of Populations.
Table 1 Parameters Description.
Parameters Descriptions
Λ The recruitment rate of susceptible population
β1 The probability of transmission from asymptomatic infected people
β2 The probability rate of transmission from symptomatic infected people
μ The natural mortality rate
μ1 Natural death rate plus COVID-19 death rate
α The probability of exposed people become infected
p The proportion of exposed people become infected
κ The rate of asymptomatic infected people become infected symptomatic
q The rate of quarantine
γ1 The natural recovery rate of infected asymptomatic people
γ2 The natural recovery rate of infected symptomatic people
δ The recovery rate of quarantine people
ξ The probability rate of recovered people become susceptible (waning immune)
So, based on the interaction diagram above, the COVID-19 spread mathematics model constructed as follows:(1) dS0dt=Λ-(β1IAS0+β2YSS0)-μS0
(2) dEdt=(β1IAS0+β2YSS0)-αE-μ1E
(3) dIAdt=pαE-κIA-γ1IA-μ1IA
(4) dYSdt=(1-p)αE-qYS-γ2YS+β1IAZ0+β2YsZ0+κIA-μ1YS
(5) dZdt=γ1IA+γ2YS+δQ-ξZ-μZ
(6) dZ0dt=ξZ-β1IAZ0-β2YsZ0-μZ0
(7) dQdt=qYS-δQ-μ1Q,
with S0(0)⩾0,E(0)⩾0,IA(0)⩾0,YS(0)⩾0,Z(0)⩾0,Z0(0)⩾0,Q(0)⩾0as the initial conditions.
3 Mathematical Analysis
Lemma 3.1 If the initial values S0(0)>0,E(0)>0,IA(0)>0,YS(0)>0,Z(0)>0,Z0(0)>0, and Q(0)>0, the solution ofS0(t),E(t),IA(t),YS(t),Z(t),Z0(t),Q(t),
of system Eqs. (1), (2), (3), (4), (5), (6), (7) are positif for all t>0.
Proof Assume thatX(t)=min{S0(t),E(t),IA(t),YS(t),Z(t),Z0(t),Q(t).},∀t>0.
Clearly, X(0)>0.
Assuming that there exist a t1>0 such that.
X(t1)=0 and X(t)>0, for all t∈[0,t1),
If X(t1)=S0(t1), then E(t)⩾0,IA(t)⩾0,YS(t)⩾0,Z(t)⩾0,Z0(t)⩾0 for all t∈[0,t1].
From the equation of model (1), we can obtaindS0dt⩾-β1IAS0-β2YSS0-μS0,t∈[0,t1].
Thus, we haveS0(t)⩾S0(0)exp-∫0t1β1IAS0+β2YSS0+μS0dt,
which will be positive since exponential functions and initial solutions S0(0)are non-negative. Thus, S0(t)>0 for all t⩾0.
Similarly, we can also prove thatE(t)>0,IA(t)>0,YS(t)>0,Z(t)>0,Z0(t)>0,Q(t)>0.
Lemma 3.2 All solution of system Eqs. (1), (2), (3), (4), (5), (6), (7) are bounded for all t∈[0,t0]
Proof Since N(t)=S0(t)+E(t)+IA(t)+YS(t)+Z(t)+Z0(t)+Q(t).
We get:dNdt=Λ-μ(S0+Z+Z0)-μ1(E+IA+YS+Q).
Assume that μ=μ1, to simplify the analysis process.
Then:dNdt=Λ-μN.
Thus we have0⩽limt→∞supN(t)⩽Λμ,
so all solutions of system Eqs. (1), (2), (3), (4), (5), (6), (7) are ultimately bounded for all t∈[0,t0].
3.1 Non-endemic Equilibrium Point
The non-endemic equilibrium point of the COVID-19 disease model is obtained by setting IA=0,E=0,YS=0, and substituting it into Eqs. (1), (2), (3), (4), (5), (6), (7) to obtain:(8) P0=(S00,E0,IA0,YS0,Z0,Z00,Q0)=Λμ,0,0,0,0,0,0)
3.2 Stability of Non-endemic Equilibrium Point
Theorem 3.3 The non-endemic equilibrium point of system Eqs. (1), (2), (3), (4), (5), (6), (7) is locally asymptotically stable whenever it exists.
Proof By following Diekmann (2000) [46] substituting P0 from 8 into the Jacobian matrix for the non-endemic equilibrium point is obtained:J(P0)=-μ0-βΛμ-βΛμ0000-α-μβΛμβΛμ0000pα-κ-γ1-μ100000(1-p)ακ-q-γ2-μ100000γ1γ2-ξ-μ0δ0000ξ-μ0000q00-δ-μ1.
The characteristics of the polynomial is(9) P(λ)=1μ((λ+μ)P1(λ))=0,
P1(λ)=a0λ6+a1λ5+a2λ4+a3λ3+a4λ2+a5λ+a6.
From the polynomial (P(λ)) we get λ1=-μ and for λi with i=2,3,…,7 will be negative if aj>0 where j=0,1,2,…,6,R0<1,a1a2>a0a3,a1(a2a3+a0a5)>a12a4+a0a32, and a1a2a4>a0(a1a6+a2a5). Since the coefficients in the characteristic equation P1(λ) are complex, we proceed to analyze the coefficient values numerically with β1=β2. The results of the numerical analysis obtained (see Appendix A.), show that for λi with i=2,3,…,7 negative. Because λj with j=1,2,3,…,7 is negative, it can be concluded that the non-endemic equilibrium point of the system (1–7) is locally stable, so Theorem 3.3 is proven.
3.3 Basic Reproduction Ratio
The Basic Reproduction Ratio (R0) is an important number in epidemiology, which is defined as the number of secondary infections caused by one primary infection in a population. We use the next-generation method to determine R0, the value of R0 an be obtained by finding the dominant eigenvalue FV-1. Where F and V are Jacobian matrices of f (newly infected matrices) and v (exiting matrices) that are evaluated at the disease-free equilibrium point (P0) from 8. From the models (1–7) are obtained:F=0Λβ1μΛβ2μ000000,V=-α-μ00αp-κ-γ1-μ0(1-p)ακ-q-γ2-μ1
andFV-1=-αp-1β2-pβ1μ1+γ1p-κ-γ1β2-pβ1γ2+qΛμμ1+αμ1+κ+γ1γ2+q+μ1Λβ1γ2+q+μ1+β2κμγ2+q+μ1μ1+κ+γ1Λβ2μγ2+q+μ1000000.
The eigenvalues of (FV-1) are:λ1=-αp-1β2-pβ1μ1+γ1p-κ-γ1β2-pβ1γ2+qΛμμ1+αμ1+κ+γ1γ2+q+μ1,λ2,3=0.
Following the method described by Castilo Chavez et al. (2002) [47] the basic reproduction number in the COVID-19 is:R0=ξ(FV-1)=-αp-1β2-pβ1μ1+γ1p-κ-γ1β2-pβ1γ2+qΛμμ1+αμ1+κ+γ1γ2+q+μ1.
Under certain conditions where the probability of transmission from infected people same as from asymptomatic infected people hold β1=β2 and the natural recovery rate of infected people asymptomatic and symptomatic γ1=γ2=γ. It obtained the reproduction number for this condition symbolized by R0β. Where:R0β=Λαβ(pq+μ1+γ+κ)μ(μ1+γ+κ)(μ1+α)(μ1+γ+q).
3.4 Endemic Equilibrium Points
Theorem 3.4 An endemic equilibrium point of system
P1=(S0∗,E∗,IA∗,YS∗,Z∗,Z0∗,Q∗) will exist if G>0 and H>0 or G<0 and H<0.
Proof The endemic point of this disease is endemic in certain areas for a certain period, which releases the COVID-19 in the population. It is indicated by the presence of compartments exposed to virus transmission E∗,Ia∗,YS∗ at steady state. By calculating model (1), (5), (6), (7) and setting the right hand side zero we obtained:S0∗=Λβ(IA∗+YS∗)+μ,IA∗=αEpγ+κ+μ1,YS∗=(1-p)αE∗+βIA∗Z0∗q+γ-βZ0∗,Z∗=γ(IA∗+YS∗)+δQ∗(μ+ξ),Z0∗=ξZ∗(β(IA∗+YS∗)+μ,Q∗=qYS∗δ+μ1.
By substituting S0∗,IA∗,YS∗,Z0∗,Z0∗,Q∗ to equation (2.2) and set the right hand side equal to zero, obtained:(10) A2E2+A1E+A0=0,
which this polynomial have to roots E=0 or E=E∗ which can be written byE∗=GH,
where G and H written on Appendix B.
Because the denomerator of H always positif, the steady state E∗ will exsist if R0β>1 and pq>0, see attachment for the proof. The system of Eqs. (1), (2), (3), (4), (5), (6), (7) will have an endemic equilibrium point if G>0 and H>0 or G<0 and H<0. This condition indicates that the system of Eqs. (1), (2), (3), (4), (5), (6), (7) has a unique endemic equilibrium point.
3.5 Stability of Endemic Equilibrium Point
Theorem 3.5 The endemic equilibrium point of the system (P1) is locally asymptotically stable whenever it exists.
Proof Following method on the proof of Theorem 3.5, Based on the method of proof of Theorem 2, by substituting P1 is obtained characteristic polynomial(11) Q(λ)=a0λ7+a1λ6+a2λ5+a3λ4+a4λ3+a5λ2+a6λ+a7.
From the polynomial (Q(λ)) we get λi with i=1,2,3,…,7 will be negative if aj>0 where j=0,1,2,…,7,a1a2>a0a3,a1(a2a3+a0a5)>a12a4+a0a32,a1a2(a3a4+a0a7)>a0a3(a1a6+a2a5), and a2a5>a0a7. Since the coefficients in the characteristic equation Q(λ) are complex, we proceed to analyze the coefficient values numerically. The results of the numerical analysis obtained can be seen in the Appendix C. It satisfies the Routh-Hurwitz’s criteria so that the endemic point is locally asymptotically stable whenever it exists. These results will remain consistent using the parameter values in Table 2 .Table 2 Parameters Values.
Parameter Value Unit Source
Λ 107365×65 people×day-1 Estimated
β1 1.727 ×10-7 (people×day)-1 Fitting
β2 7.478 ×10-8 (people×day)-1 Fitting
μ 1365×65 day-1 Estimated
μ1 0.082 day-1 Fitting
α 15.2 day-1 [52]
p 0.2 N/A Assumed
κ 0.19 day-1 [53]
q 1.026 ×10-6 day-1 Fitting
γ1 110 day-1 [52]
γ2 114 day-1 Assumed
δ 0.1 day-1 [54]
ξ 0.02 day-1 Assumed
3.6 Global Stability of The Equilibria
Theorem 3.6 The non-endemic equilibrium point (P0) is globally asymptotically stable ifβ1Λμ<μ1andβ2Λμ<μ1.
Proof Refer to global proving by Tewa et al. (2009)[48], let
P0=(S00,E0,IA0,YS0,Z0,Z00,Q0)=Λμ,0,0,0,0,0,0 is the non-endemic equilibrium point of system Eqs. (1), (2), (3), (4), (5), (6), (7).
Define the Lyapunov functionV(t)=S0-S0∗-S0∗lnS0S0∗+E+IA+YS+Z+Z0+Q.
Differentiating with respect to time yieldsdVdt=(S0-S0∗)dS0dtS0∗+dEdt+dIAdt+dYSdt+Zdt+Z0dt+QdtdVdt=(S0-S0∗)(Λ-μS0)S0∗-(S0-S0∗)(β1IA+β2YS)+β1IAS0+β2YSS0-αE-μ1E+pαE-κIA-γIA-μ1IA+(1-p)αE-qYS-γYS+β1IAZ0+β2YSZ0+κIA-μ1YS+γIA+γYS+δQ-ξZ-μZ+ξZ-β1IAZ0-β2YSZ0-μZ0+qYS-δQ-μ1QdVdt=(S0-S0∗)(Λ-μS0)S0-(S0-S0∗)(β1IA+β2YS)+β1IAS0+β2YSS0-μ1(E+IA+YS+Q)-μ(Z+Z0)dVdt=Λ(S0-S0∗)(1S0-1S0∗)+(β1S0∗-μ1)IA+(β2S0∗-μ1)YS-μ1(E+Q)-μ(Z+Z0)dVdt=Λ(S0-S0∗)(1S0-1S0∗)+(β1Λμ-μ1)IA+(β2Λμ-μ1)YS-μ1(E+Q)-μ(Z+Z0).
The value of dVdt will be negative ifβ1Λμ<μ1andβ2Λμ<μ1.
By following LaSalle’s extension on Lyapunov’s method [49], disease-free equilibrium P0 is globally asymptotically stable.
This concludes the proof.
4 Sensitivity Analysis
This section presents a global sensitivity analysis of the model. We use the combination of Latin Hypercube Sampling (LHS) and Partial Rank Correlation Coefficient (PRCC) to determine the most influential parameters of the model [50]. LHS is stratified sampling without replacement. The parameter distribution is divided into equation probability intervals and then is sampled. Each parameter interval is sampled once and the entire range of each parameter is explored. A matrix is then generated which consists of N rows for the number of samples and k columns for the number of varied parameters. The model solution is then generated using the combination of parameters (each row). The outcome of interest is the increasing number of infected individuals. The result of sensitivity analysis is given in Fig. 2 .Fig. 2 PRCC over time when we measure against the increasing number of infected individuals.
It showed that the waning immunity (ξ) is one of the influential parameters. When the value of waning immunity increases, the number of infected individuals also increases. This means that waning immunity would contribute to the increasing number of infected individuals. Therefore, an analysis of effects of waning immunity is of importance. The parameters k,p are also influential and has positive relationship. This means that the rate of asymptomatic become infected and the proportion of exposed individuals become infected contributes to an increasing number of infected individuals. When these parameter values increases, the number of infected individuals decreases.
5 A Case Study
In this section, we estimated the parameters β1,β2, and γ against data of West Java, Indonesia. The data are obtained from the website https://pikobar.jabarprov.go.id/table-case/. We estimate the parameter values by minimizing the sum of squared error. The parameters b and a are estimated against the data for the first 30 days. It is sufficient since the aim is to obtain the general insights of the values of parameters β1,β2 and q in the early period of the outbreak. The other parameter values are obtained from literature and are given in Table 2.
The lsqnonlin built-in function in MATLAB is used for the parameter estimation.
We minimize the sum of squared error as(12) SE=∑t=1nQt-gt(x)2.
where Qt is the number of active cases of Q up to day t, respectively, while gt(x) is the number of active cases for Q up to day t from the model’s solution, respectively. The transmission rate, β0 and β1, the quarantine rate q are then estimated using the “lsqnonlin” built-in function in MATLAB. The case fatality rate is estimated using the linear regression method.
The initial conditions used Table 3 . The initial conditions for susceptible individuals are an approximate total population in West Java. The fitted values of β1,β2 and q. The values are then used in the numerical simulation. The plot of model’s solution and data is given in Fig. 3 . With these parameter values, the reproduction number for West Java R0=3.180126127. This means that an outbreak happens and the control needs to be implemented to minimize the risk of infections.Table 3 Initial Values of each Compartments.
Compartment S(0) E(0) Ia(0) Ys(0) Z(0) Z0(0) Q(0)
Initial Values 107 100 100 100 100 100 5
Fig. 3 Fitting Parameter from Confirmed Cases (a) and Cumulative Death (b).
6 Numerical Simulation
This numerical simulation is designed to support the results of the analysis discussed in the previous section. We set the parameter by curve fitting from actual case of COVID-19 in West Java Province, Indonesia. We applied Runge–Kutta-Fehlberg (RKF) method in MAPLE software, to solve the ordinary differential equations of model Eqs. (1), (2), (3), (4), (5), (6), (7) using the parameters in Table 2, Table 3. RKF method is one of the most popular numerical approach because it is quite accurate, stable, to program [51].
Fig. 4 show the endemic incidence where the susceptible population (S0) decreases as a result of transmission from the symptomatic and asymptomatic infected population. Hereafter, this increases the latent population (E), the asymptomatic infected population (IA), the symptomatic infected population (YS), the recovered population (Z), the susceptibility to previously infected populations (Z0), and the quarantine population (Q).Fig. 4 Dynamical Population of each Compartment: (a) Population of S,Z0,Z(b) Population of E,Ys, & Ia.
However, after the 20th day, the latent (E) and asymptomatic infected population (IA) decreased, this is because the latent population and the asymptomatic infected population became the removed population. While symptomatic human populations (YS) have declined due to an increase in quarantined populations (Q), the recovered (Z) and susceptible that previously infected populations (Z0) have consequently increased.
Fig. 5 shows the number of quarantine population Q and the cumulative population of quarantine. The number of quarantine compartment populations increases as the asymptomatic infection increases. The peak occurs at 30 days where the number reaches 70 and after that decreases. At the end of the 400th day, the number of cumulative quarantine reaches 3200.Fig. 5 Dynamical Population of (a) Active Quarantine (b) Cumulative Quarantine.
6.1 The Effect of Waning Immunity
In this section, we simulate the sensitivity analysis for the effect of parameter ξ, related to waning immunity issue, which describes the probability rate of recovered people become susceptible, and the probability rate of susceptible people that previously infected become asymptomatic infected, respectively. Using the parameters and initial values in Table 2, Table 3, except for ξ, we choose ξ=0.001,0.01,0.1,1.
Fig. 6, Fig. 7 show the effect of increasing the value of ξ and τ. In these simulations the peak time of disease spread do not change, but at the time after the peak has been passed, the more value of ξ and τ multiply the number of Asymptomatic infected population (YS).Fig. 6 Simulation of The effect of Waning Immunity (ξ) on (a)Symptomatic Infected Population (YS) and (b) Quarantine Population (Q).
Fig. 7 Simulation of The effect of Waning Immunity ξ with respect to time (t) for each Compartments (a) E, (b) IA, (c) YS, and (d) Z.
The effect of changes in the value of the probability rate of recovered people become susceptible (ξ) on the E,IA,YS, and Z compartments is shown in Fig. 7. The changes value of the parameter ξ did not have a significant impact on the number of compartments E and IA. The number of populations E(t) and IA(t) is relatively unchanged for every ξ∈[0.001,0.1]. This means that changes in the reinfected parameter value do not really affect the number population Exposed (E) and asymptomatic infected population (IA).
However, the higher the value of the parameter ξ, the higher the population of YS after passing the peak of the spread, and the lower the population Z. When this parameter is greater, the recovered population (Z) decreases due to the loss of immunity and returns to the susceptible population that previously infected (Z0). Where population Z0 can be re-infected to become asymptomatic infected population (YS).
6.2 The Effect of Quarantine
Fig. 8 show that with increase the value of quarantine parameter (q), the peak size and the final size of symptomatic infected population (YS) is decrease. This show that Increasing the intensity of quarantine policy may press the spread of COVID-19. Fig. 9 show that changes in the value of quarantine parameters to the population of each compartments E,IA,YS, and Z are presented in 3-dimensional changes in time. When the value of the quarantine parameter q is increased, within the range [0.01,1], the number of infected populations can be reduced. This is indicated by the reduction in the peak value of Exposed (E), Asymptomatic Infected (IA), and Symptomatic Infected (YS), in the change in the value of q. Meanwhile the population in the quarantine compartment (Q) is increasing from population of Symptomatic Infected which did Quarantine.Fig. 8 Dynamical population of Symptomatic Infected Population (YS) in changes of quarantine parameter (q).
Fig. 9 Simulation of The effect of quarantine parameter (q) with respect to time (t) for each Compartments (a) E, (b) IA, (c) YS, and (d) Q.
7 Discussion and Conclusion
We have formulated a mathematical model of COVID-19 transmission by considering infected individuals with symptoms and asymptomatic, as well as decreased immunity, validated with data from West Java Province, Indonesia. The compartment-based model is formulated as a system of differential equations, where the population is divided into Susceptible Populations (S0), Exposed Populations (E), Asymptomatic Infection Populations (IA), Symptomatic Infection Populations (YS), Recovered Populations (Z), Susceptible Populations previously infected (Z0), and Quarantine Population (Q). Then the model is analyzed mathematically, the results show that there are two equilibrium points, namely a disease-free equilibrium point and an endemic equilibrium point. Besides, with the next-generation matrix method, the Basic Reproduction Number (R0) for West Java Province is obtained of 3.180126127. This means that West Java Province is affected by the COVID-19 outbreak and controls are needed to minimize the risk of transmitting COVID-19. Stability and sensitivity are analyzed to determine the parameters that influence the spread of COVID-19. The simulation results show that the factor of decreasing immunity can affect the spread of COVID-19. This is because when the increase in immunity decreases, the infected population increases. Meanwhile, reinfection has no significant effect on the number of exposed and infected asymptomatic populations and the isolation period can slow the spread of COVID-19 in West Java Province, Indonesia. The results obtained can be used as a reference for the early prevention of the spread of COVID-19 in West Java.
8 Authors’s Contributions
N Anggriani, R Amelia, & W Suryaningrat contributed to the study design, model formulation, model analysis, and numerical simulation. MZ Ndii designed sensitivity analysis and performed the case study including parameter estimation. MAA Pratama complete and verify the analysis. All authors have read and agreed to the published version of the manuscript.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Appendix A Proof of Numerical Analysis of Theorem 3.3
Coefficient polynomial (P1(λ)):a0=1,a1=1.030393017,a2=0.3892774217,a3=0.06516694579,a4=0.004442996019,a5=0.00006597971860,&a6=2.773132296×10-9
Value of:a1a2=0.4011087370,a0a3=0.06516694579,a1(a2a3+a0a5)=0.4683242878,a12a4+a0a32=0.008963903101,a1a2a4=0.001782124518,a0(a1a6+a2a5)=0.00002568727211.
Appendix B Characteristic Polynomial of Exposed Compartment
G=K+(((-μ12γ+((-q-γ)δ-κγ-γ2)μ1+(-γ2+(-κ-q)γ-κq)δ)μ+βΛμ12+(δβΛ+Λβγ+(κ+qp)Λβ)μ1+(Λβγ+(2qp+κ)Λβ)δ)ξ+(βΛμ12+(δβΛ+Λβγ+(κ+qp)Λβ)μ1+(Λβγ+(κ+qp)Λβ)δ)μ)α+(-μ13γ+((-q-γ)δ-κγ-γ2)μ12+(-γ2+(-κ-q)γ-κq)δμ1)μξ,
whereK=((μ1γ+(q+γ)δ)2(κ+γ+μ1)2(α+μ1)2ξ2+2(κ+γ+μ1)(δ+μ1)βΛα(α+μ1)(-μ12γ+((-γ+(2p-1)q)δ-γ(κ+γ+qp))μ1+δ(q+γ)(qp-κ-γ))ξ+β2α2Λ2(δ+μ1)2(κ+γ+qp+μ1)2)μ2+2ξβΛα((κ+γ+μ1)(-μ13γ+((-2γ+(2p-1)q)δ-γ(κ+γ+qp))μ12+((-γ+(2p-1)q)δ-2γ2+((-p-1)q-2κ)γ+q(qp-κ))δμ1-δ2(κ+γ)(q+γ))(α+μ1)ξ+(μ12+(κ+γ+δ+qp)μ1+δ(κ+γ))(δ+μ1)βΛα(κ+γ+qp+μ1))μ+ξ2(μ12+(κ+γ+δ+qp)μ1+δ(κ+γ))2β2Λ2α2,
andH={2αβδpqξ(μ1+α)}.
By substituting the parameter value from Table 2 we have endemic equilibrium point:E=1536.477836,IA=158.8583370,Q=0.01936422900,S0=679.0681919,YS=3434.980193,Z=17931.49917,Z0=577.7849236.
So, the non-endemic equilibrium point is stable if it exists, because it satisfies Routh-Hurwit’s criterion. These results remain consistent using the parameter values in the Table 2, except β1=β2=0.7477942169036×10-10.
Appendix C Proof of Numerical Analysis of Endemic Equilibrum Point
Coefficient polynomial (Q(λ)):a0=1.000000000,a1=2.171964310,a2=1.885854850,a3=0.84368388,a4=0.20819336,a5=0.0279316,a6=0.001468969,a7=0.00001178844740.
Value of:a1a2=4.096009428,a0a3=0.8436838800,a1(a2a3+a0a5)=3.516403565,a12a4+a0a32=1.693939876,a1a2(a3a4+a0a7)=0.7195098093,a0a3(a1a6+a2a5)=0.04713281469,a2a5=0.05267494333,a0a7=0.00001178844740.
Acknowledgments
The work was supported by Universitas Padjadjaran, with contract number 1735/UN6.3.1/LT/2020 through Hibah Riset Data Pustaka dan Daring.
==== Refs
References
1 Li Q. Guan X. Wu P. Wang X. Zhou L. Tong Y. Feng Z. Early transmission dynamics in Wuhan, China, of novel coronavirus-infected pneumonia New England journal of medicine. 382 2020 1199 1207 31995857
2 Worldometer, COVID-19 coronavirus pandemic, (2020). Avaliable from: https://www.worldometers.info/coronavirus/.
3 Chan J.F.W. Yuan S. Kok K.H. Yo K.K.W. Chu H. A familial cluster of pneumonia associated with the 2019 novel coronavirus indicating person-to-person transmission: a study of a family cluster Lancet 395 2020 2019 514 523
4 Chen N. Zhou M. Dong X. Qu J. Gong F. Epidemiological and clinical characteristics of 99 cases of 2019 novel coronavirus pneumonia in Wuhan, China: a descriptive study Lancet 395 2020 2019 507 513
5 Huang C. Wang Y. Li X. Ren L. Zhao J. Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China Lancet 395 2020 2019 497 506
6 Centers for Disease Control and Prevention, Situation Summary COVID-19, 2020. Retrieved March 23, 2020 (https://www.cdc.gov/coronavirus/2019-nCoV/index.html).
7 B. Shayak, M.M. Sharma, M. Gaur, & A.K. Mishra. Impact of reproduction number on multiwave spreading dynamics of COVID-19 with temporary immunity: a mathematical model, International Journal of Infectious Diseases (2021). Preprint at doi: 10.1016/j.ijid.2021.01.018.
8 Law S.K. Leung A.W.N. Xu C. Is reinfection possible after recovery from COVID-19? Hong Kong Medical Journal 26 2020 264 265 32536614
9 P. Selhorst, S.V. Ierssel, J. Michiels, J. Mariën, K. Bartholomeeusen, E. Dirinck, S. Vandamme, H. Jansens,& K.K. Ariën.Symptomatic SARS-CoV-2 reinfection of a health care worker in a Belgian nosocomial outbreak despite primary neutralizing antibody response, Clinical Infectious Diseases, (2021). Preprint at doi.org/10.1093/cid/ciaa1850
10 Galanti M. Shaman J. Direct Observation of Repeated Infections with Endemic Coronaviruses J. Infect. Dis. 3 2020 409 415
11 Ledford H. Coronavirus Re-infections: Three Questions Scientists Are Asking Nature Research Journals 7824 2020 168 169
12 Giannitsarou C. Kissler S. Toxvaerd F. Waning Immunity and the Second Wave: Some Projections for SARS-CoV-2 CEPR Centre for Economic Policy Research 2020
13 Kumar S. Kumar R. Cattani C. Samet B. Chaotic behaviour of fractional predator-prey dynamical system Chaos, Solitons & Fractals 135 2020 109811
14 Kumar S. Ghosh S. Samet B. Goufo E.F.D. An analysis for heat equations arises in diffusion process using new Yang-Abdel-Aty-Cattani fractional operator Mathematical Methods in the Applied Sciences 2020 1 19
15 Kumar S. Kumar R. Agarwal R.P. Samet B. A study of fractional Lotka-Volterra population model using Haar wavelet and Adams-Bashforth-Moulton methods Mathematical Methods in the Applied Sciences 2020 1 15
16 Kumar S. Kumar A. Samet B. Dutta H. A study on fractional host-parasitoid population dynamical model to describe insect species Numerical Methods for Partial Differential Equations 2020 1 20
17 Anggriani N. Tasman H. Ndii M.Z. Supriatna A.K. Soewono E. Siregar E. The effect of reinfection with the same serotype on dengue transmission dynamics Appl. Math. Comput. 349 2019 62 80
18 Ndii M.Z. Amarti Z. Wiraningsih E.D. Supriatna A.K. Rabies epidemic model with uncertainty in parameters: crisp and fuzzy approaches IOP Conference Series: Materials Science and Engineering 332 2018 012031
19 S. Kumar, A. Kumar, B. Samet, J.F. Gmez-Aguilar,& M.S. Osman, A chaos study of tumor and effector cells in fractional tumor-immune model for cancer treatment, Chaos, Solitons & Fractals, 141 (2020), 110321.
20 Ghanbari B. Kumar S. Kumar R. A study of behaviour for immune and tumor cells in immunogenetic tumour model with non-singular fractional derivative Chaos, Solitons & Fractals 133 2020 109619
21 Ndii M.Z. Supriatna A.K. Stochastic mathematical models in epidemiology Information 20 2017 6185 6196
22 Alrabaiah H. Mo A. Safi M.H. DarAssi B. Al-Hdaibat S. Ullah M.A. Khan S.A.A.Sh.ah. Optimal control analysis of hepatitis B virus with treatment and vaccination Results in Physics 19 2020 103599
23 Oud M.A.A. Ali A. Alrabaiah H. Ullah S. Khan M.A. Islam S. A fractional order mathematical model forCOVID-19 dynamics with quarantine, isolation, and environmental viral load Advances in Difference Equations 106 2021 1 19
24 Z. Liu, P. Magal, O. Seydi, & G. Webb., Predicting the Cumulative Number of Cases for the COVID-19 Epidemic in China from Early Data. Preprint at doi: 10.20944/preprints202002.0365.v1 (2020a).
25 Z. Liu, P. Magal, O. Seydi, & G. Webb., Understanding Unreported Cases in the COVID-19 Epidemic Outbreak in Wuhan, China, and the Importance of Major Public Health Interventions, Biology 9(2020b).
26 Pierre M. Webb G. The Parameter Identification Problem for SIR Epidemic Models: Identifying Unreported Cases J. Math. Biol. 77 2018
27 K. Biswas, A. Khaleque, & P. Sen, COVID-19 spread: Reproduction of data and prediction using a SIR model on Euclidean network, pp. 4–7, 2020, [Online]. Available: http://arxiv.org/abs/2003.07063.
28 K. Biswas & P. Sen, Space-time dependence of corona virus (COVID-19) outbreak, no. 1, pp. 5-7, 2020, [Online]. Available: https://arxiv.org/abs/2003.03149.
29 Khan M.A. Atangana A. Modeling the dynamics of novel coronavirus(2019-nCov) with fractional derivative Alexandria Engineering Journal 59 2020 2379 2389
30 Khan M.A. Atangana A. Alzahrani E. The dynamics of COVID-19 with quarantined and isolation Advances in Difference Equations 1 2020 1 22
31 Alqarni M.S. Alghamdi M. Muhammad T. Alshomrani A.S. Khan M.A. Mathematical modeling for novel coronavirus (COVID-19) and control Numerical Methods for Partial Differential Equations 2020
32 Chu Y. Ali A. Khan M.A. Islam S. Ullah S. Dynamics of fractional order COVID-19 model with a case study of Saudi Arabia Results in Physics 21 2021 103787 33552881
33 S. Kumar, R. Kumar, S. Momani,& Samir Hadid, A study on fractional COVID-19 disease model by using Hermite wavelets, Mathematical Methods in the Applied Sciences, (2021), 1-17.
34 Khan M.A. Ullah S. Kumar S. A robust study on 2019-nCOV outbreaks through non-singular derivative The European Physical Journal Plus 136 2021 1 20
35 Safare K.M. Betageri V.S. Prakasha D.G. Veeresha P. Kumar S. A mathematical analysis of ongoing outbreak COVID-19 in India through nonsingular derivative Numerical Methods for Partial Differential Equations 37 2021 1282 1298
36 Kumar S. Chauhan R.P. Momani S. & Samir Hadid, Numerical investigations on COVID-19 modelthrough singular and non-singularfractional operators Numerical Methods for Partial Differential Equations 2020 1 27
37 Lin Q. Zhao S. Daozhou G. Lou Y. Yang S. A conceptual model for the coronavirus disease 2019 (COVID-19) outbreak in Wuhan, China with individual reaction and governmental action International Journal of Infectious Diseases. 93 2020 2019 211 216
38 Ivorra B. Ferrandez M.R. Perez M.V. Ramos A.M. Mathematical modeling of the spread of the coronavirus disease 2019 (COVID-19) taking into account the undetected infections. The case of China Communications in Nonlinear Science and Numerical Simulation. 88 2020
39 Tang B. Xia F. Tang S. Bragazzi N.L. Li Q. The effectiveness of quarantine and isolation determine the trend of the COVID-19 epidemic in the final phase of the current outbreak in China International Journal of Infectious Diseases 96 2020 288 293 32437933
40 J.P. Arcede, R.L.C. Anan, C.Q. Mentuda, & Y. Mammer., Accounting for Symptomatic and Asymptomatic in a SEIR-type model of COVID-19. Preprint at https://arxiv.org/abs/2004.01805 (2020).
41 Ndii M.Z. Mage A.R. Messakh J.J. Djahi B.S. Optimal vaccination strategy for dengue transmission in Kupang city Indonesia. Heliyon 6 2020 e05345 33204872
42 H. Sun, Y. Qiu, H. Yan, Y. Huang, Y. Zhu, et al., Tracking and Predicting COVID-19 Epidemic in China Mainland. Preprint at https://www.medrxiv.org/content/10.1101/2020.02.17.20024257v1 (2020).
43 Ullah S. Khan M.A. Modeling the impact of non-pharmaceutical interventions on the dynamics of novel coronavirus with optimal control analysis with a case study Chaos, Solitons & Fractals 139 2020 110075 32834618
44 Anggriani N. Supriatna A.K. Soewono E. A Critical Protection Level Derived from Dengue Infection Mathematical Model Considering Asymptomatic and Symptomatic Classes Journal of Physics Conference Series. 423 2013 1 10
45 Kontis V. Bennett J.E. Rashid T. Parks R.M. Pearson-Stuttard J. Guillot M. Asaria P. Zhou B. Battaglini M. Corsetti G. McKee M. Di Cesare M. Mathers C.D. Ezzati M. Magnitude, demographics and dynamics of the effect of the first wave of the COVID-19 pandemic on all-cause mortality in 21 industrialized countries Nat. Med. 26 2020 1919 1928 33057181
46 Diekmann O. Heesterbeek J. Mathematical Epidemiology of Infectious Diseases Wiley Series in Mathematical and Computational Biology 2000
47 C. Chavez, C., Z. Feng, W. Huang., On the computation of R0 and its role in global stability. In: Math ApproachesEmerg Reemerg Infect Dis Intro. IMA, 125 (2002), 229-250.
48 Tewa J.J. Dimi J.L. Bowong S. Lyapunov functions for a dengue disease transmission model Chaos, Solitons, & Fractals 39 2009 936 941
49 LaSalle J.P. The stability of dynamical systems 1976 SIAM Philadelphia
50 Marino S. Hogue I.B. Ray C.J. Kirschner D.E. A methodology for performing global uncertainty and sensitivity analysis in systems biology J. Theor. Biol. 254 2008 178 198 18572196
51 Islam A.M. A comparative study on numerical solutions of initial value problems (IVP) for ordinary differential equations (ODE) with Euler and Runge Kutta Methods American Journal of Computational and Applied Mathematics 5 2015 393 404
52 Kucharski A.J. Russell T.W. Diamond C. Liu Y. Edmunds J. Funk S. Eggo R.M. Sun F. Jit M. Munday J.D. Davies N. Gimma A. van Zandvoort K. Gibbs H. Hellewell J. Jarvis C.I. Clifford S. Quilty B.J. Bosse N.I. Abbott S. Klepac P. Flasche S. Early dynamics of transmission and control of COVID-19: a mathematical modelling study Lancet. Infect. Dis 20 2020 553 558 32171059
53 Guan W.J. Ni Z.Y. Hu Y. Liang W.H. Ou C.Q. He J.X. Clinical Characteristics of Coronavirus Disease 2019 in China New Engl. J. Med. 382 2020 2019 1708 1720
54 D. Aldila, S.H.A. Khoshnaw, E. Safitri, Y.R. Anwar, A.R.Q. Bakry, B.M. Samiadji, D.A. Anugerah, M.F. Farhan GH, I.D. Ayulani, S.N. Salim, A mathematical study on the spread of COVID-19 considering social distancing and rapid assessment: The case of Jakarta, Indonesia, Chaos, Solitons and Fractals, 139 (2020), 110042.
| 0 | PMC9703878 | NO-CC CODE | 2022-11-29 23:21:14 | no | 2022 Jan 14; 61(1):113-124 | utf-8 | null | null | null | oa_other |
==== Front
Hematol Transfus Cell Ther
Hematol Transfus Cell Ther
Hematology, Transfusion and Cell Therapy
2531-1379
2531-1387
Published by Elsevier Editora Ltda.
S2531-1379(22)01249-4
10.1016/j.htct.2022.09.1134
Article
GRAVE TROMBOCITOPENIA IMUNOLÓGICA SECUNDÁRIA À VACINA ANTI-COVID-19 ‒ PFIZER
Araujo AA a
Barbosa FM b
Pizza M b
Borsato ML b
Bruniera P b
Luporini SM ab
a Hospital Municipal Infantil Menino Jesus (HMIMJ), Instituto de Responsabilidade Social Sírio Libanês (IRSSL), São Paulo, SP, Brasil
b Irmandade da Santa Casa de Misericórdia de São Paulo, São Paulo, SP, Brasil
15 10 2022
10 2022
15 10 2022
44 S660S661
Copyright © 2022 Published by Elsevier Editora Ltda.
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Introdução
Casos de trombocitopenia imunológica secundária à vacinação SARS-CoV-2 com vacina da Pfizer têm sido relatadas. O alarme público foi intensificado seguindo a morte do primeiro paciente por hemorragia intracraniana ocorrida na Flórida, relatada no USA Today e The New York Times. Grupos de pesquisadores de centros universitários dos USA estudaram séries de casos com baixas contagens de plaquetas duas semanas após vacinação SARS-CoV e desenvolvendo trombocitopenia imunológica.
Relato de caso
Masc, 15 anos. Entrada PSI ‒ HMIMJ em 09.2021 por sufusões hemorrágicas em pele e mucosas (sangramento nasal volumoso e hematúria macroscópica) de instalação abrupta, plaquetas 2.000 μL. primeira dose da vacina anti-COVID-19 (Pfizer) 14 dias antes. Na UTI, recebeu gamaglobulina humana, metilprednisolona por 3 dias, sem resposta. Necessitou 58 unidades plaquetas e 4 unidades CH. Mielograma – 09.2021 – S. megacariocítica – normoplásica com morfologia e maturação preservadas. Sorologia e teste COVID-19 negativos, funções hepática e renal normais, avaliação pela reumatologia, sem anormalidades. Obtido Eltrombopag (farmácia alto custo) iniciada dose 50 mg por 3 semanas aumentando para 75 mg. Mielograma (11.2021) S. megacariocítica hiperplásica, morfologia e maturação preservadas. Mantida medicação por várias semanas. Melhora no sangramento nasal embora plaquetas ≤10.000 μL. Após 50 dias na UTI foi para enfermaria e após 2 semanas alta e seguimento ambulatorial na Hematologia-ped da Santa Casa de São Paulo. Mantendo plaquetas ao redor de 8.000 μL, sem sangramentos ativos, Recebeu Vincristina 1 mg EV semanal por 4 semanas, sem sucesso. Após 10 semanas de Eltrombopag a medicação foi suspensa, iniciada Azatioprina 1.5 mg/kg/dia ‒ por 4 meses. No terceiro mês da medicação plaquetas 61.000 μL, uma semana após 212.000 e, a partir desta data contagens normais. Completou 4 meses de azatioprina em abril 2022. Resposta da Notificação de Evento Adverso à vacina PFIZER – Resposta: B1 – reação temporal consistente, mas sem evidências na literatura para se estabelecer uma relação causal. Conduta: contraindicação sem substituição do esquema.
Discussão/Conclusão
Paciente com grave trombocitopenia imunológica iniciada após 14 dias da vacina anti-COVID-19 da Pfizer. Refratariedade para gamaglobulina humana, corticoterapia. No início e no curso de Eltrombopag contagens <10.000 μL, mas sem sangramentos ativos. Resposta após 6 meses do início do quadro com azatioprina. Avaliados casos relatados de pacientes com trombocitopenia seguindo vacinação identificados pelo Vaccine Adverse Events Reporting Systems (VAERS). Concluíram que é possível que a vacina da Pfizer tenha potencial de “gatilho” para trombocitopenia imunológica de novo, embora raramente. Taxa relatada de trombocitopenia foi de 0,8 milhão doses Vacina Pfizer. Incidência anual trombocitopenia imunológica é de 3.3 milhões para cada 100.000 adultos com trombocitopenia. Assim, casos de trombocitopenia relatado ao VAERS não sugerem necessidade de preocupação com a segurança da vacina. Estudos adicionais são necessários para determinar quadros de trombocitopenia imunológica pós-vacinal coincidentes ou causais.
==== Body
pmc
| 0 | PMC9703981 | NO-CC CODE | 2022-12-01 23:19:04 | no | Hematol Transfus Cell Ther. 2022 Oct 15; 44:S660-S661 | utf-8 | Hematol Transfus Cell Ther | 2,022 | 10.1016/j.htct.2022.09.1134 | oa_other |
==== Front
Hematol Transfus Cell Ther
Hematol Transfus Cell Ther
Hematology, Transfusion and Cell Therapy
2531-1379
2531-1387
Published by Elsevier Editora Ltda.
S2531-1379(22)01247-0
10.1016/j.htct.2022.09.1132
Article
RELAÇÃO ENTRE PERFIL HEMATOLÓGICO E AS COMORBIDADES EM PACIENTES ASSINTOMÁTICOS E SINTOMÁTICOS LEVES COM COVID-19 EM SERGIPE
Rezende MS
Santos JTCD
Santos SO
Santos JD
Silva NL
Pinheiro CS
Araújo AAS
Martins-Filho PRS
Quintans-Júnior LJ
Schimieguel DM
Universidade Federal de Sergipe (UFS), São Cristóvão, SE, Brasil
15 10 2022
10 2022
15 10 2022
44 S659S660
Copyright © 2022 Published by Elsevier Editora Ltda.
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Objetivos
Um surto de pneumonia iniciou-se no final do ano de 2019, na China. O vírus causador da infecção foi identificado como um SARS-CoV-2 e a doença foi nomeada como COVID-19. Pacientes que progrediram para óbito o perfil hematológico da COVID-19 apresentavam menores índices de hemoglobina, eritrócitos e hematócrito, assim como leucocitose, linfopenia, neutrofilia e basofilia, indicadores de resposta inflamatória exacerbada. Além disso, também foi descrita a presença de trombocitopenia. Este trabalho teve como objetivo investigar a relação entre o perfil hematológico e as comorbidades preexistentes em indivíduos acometidos pela COVID-19 assintomática ou sintomática leve.
Material e Métodos
Foram selecionados indivíduos participantes do projeto de avaliação da soroprevalência do SARS-CoV-2 em Sergipe – EPISERGIPE, no qual todos os indivíduos responderam a um questionário semiestruturado. Após triagem positiva no teste rápido (Wondfo-IgG/IgM), foram coletadas amostras de sangue periférico para realização do hemograma e análise sorológica de anti-SARS-CoV-2 IgG e IgM por Imunoensaio Fluorescente (FIA) utilizando o kit comercial IchromaTM COVID-19 Ab.
Resultados
Foram analisadas amostras de 1009 participantes, dos quais 817 (81,0%) apresentaram resultados positivos e 192 (19,0%) apresentaram resultados negativos para o anti-SARS-CoV-2. Do total de indivíduos, 685 eram do sexo feminino e 324 do sexo masculino, com idade média foi de 52,5 anos. Em relação às comorbidades, 460 não relataram a presença, e 328 descreveram uma ou mais comorbidades preexistentes, sendo as mais comuns: hipertensão 215 (65,5%), diabetes 95 (29,0%) e sobrepeso 56 (17,0%). Dos 817 indivíduos que apresentaram resultados positivos, 385 (47,1%) foram assintomáticos e 433 (52,9%) sintomáticos. Dentre os sintomáticos, os sintomas mais citados foram cefaleia 229 (52,8%), tosse 198 (45,7%), perda de paladar/olfato 180 (41,6%), rinorreia 180 (41,6%) e febre 141 (32,6%). Nos 817 indivíduos com resultados positivos, a anemia foi a alteração hematológica mais frequente, 189 (23,1%) casos, destes 102 (12,5%) sintomáticos e 87 (10,6%) assintomáticos. A segunda alteração mais frequente foi leucopenia com 89 casos (38 assintomáticos e 51 sintomáticos), seguida por leucocitose 63 (31 assintomáticos e 32 sintomáticos).
Discussão
Neste estudo a anemia foi o parâmetro hematológico mais frequente nos participantes positivos para COVID-19, sintomáticos e assintomáticos, diferentemente de outros estudos em que a anemia é encontrada em estágios graves da doença, nos quais a redução da hemoglobina pode ser explicada pela diminuição da eritropoese e hemólise, em consequência dos fatores inflamatórios desencadeados pela COVID-19. Neste estudo pode-se hipotetizar ainda a pré-existência de anemia nestes pacientes antes da pandemia. A leucopenia foi a segunda alteração mais encontrada, e em outros estudos realizados com pacientes internados, esta é a alteração mais frequente, sendo um possível marcador de gravidade da doença.
Conclusão
A partir dos resultados prévios encontrados, observa-se alterações no perfil eritrocitário e leucocitário em indivíduos assintomáticos e sintomáticos leves com COVID-19, diferentemente de outros relatos da literatura, nos quais as alterações eritrocitárias foram evidentes em casos graves da doença.
==== Body
pmc
| 0 | PMC9704019 | NO-CC CODE | 2022-12-01 23:19:04 | no | Hematol Transfus Cell Ther. 2022 Oct 15; 44:S659-S660 | utf-8 | Hematol Transfus Cell Ther | 2,022 | 10.1016/j.htct.2022.09.1132 | oa_other |
==== Front
Curr Opin Virol
Curr Opin Virol
Current Opinion in Virology
1879-6257
1879-6265
Elsevier B.V.
S1879-6257(22)00097-9
10.1016/j.coviro.2022.101286
101286
Article
Role of cytokines in poxvirus host tropism and adaptation
Rahman Masmudur M
McFadden Grant
Center for Immunotherapy, Vaccines and Virotherapy, Biodesign Institute, Arizona State University, USA
22 11 2022
12 2022
22 11 2022
57 101286101286
© 2022 Elsevier B.V. All rights reserved.
2022
Elsevier B.V.
Elsevier has created a Monkeypox Information Center (https://www.elsevier.com/connect/monkeypox-information-center) in response to the declared public health emergency of international concern, with free information in English on the monkeypox virus. The Monkeypox Information Center is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its monkeypox related research that is available on the Monkeypox Information Center - including this research content - immediately available in publicly funded repositories, with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the Monkeypox Information Center remains active.
Poxviruses are a diverse family of double-stranded DNA viruses that cause mild-to-severe disease in selective hosts, including humans. Although most poxviruses are restricted to their hosts, some members can leap host species and cause zoonotic diseases and, therefore, are genuine threats to human and animal health. The recent global spread of monkeypox in humans suggests that zoonotic poxviruses can adapt to a new host, spread rapidly in the new host, and evolve to better evade host innate barriers. Unlike many other viruses, poxviruses express an extensive repertoire of self-defense proteins that play a vital role in the evasion of host innate and adaptive immune responses in their newest host species. The function of these viral immune modulators and host-specific cytokine responses can result in different host tropism and poxvirus disease progression. Here, we review the role of different cytokines that control poxvirus host tropism and adaptation.
==== Body
pmc Current Opinion in Virology 2022, 57:101286
This review comes from a themed issue on Antiviral strategies
Edited by Lieve Naesens and Bruno Canard
https://doi.org/10.1016/j.coviro.2022.101286
1879–6257/© 2022 Elsevier B.V. All rights reserved.
Introduction
Poxviruses are large double-stranded DNA (dsDNA) viruses infecting insects and various vertebrate species. They belong to the Poxviridae family of viruses and are further classified into two subfamilies: the Entomopoxvirinae, infecting insects, and the Chordopoxvirinae, infecting vertebrates. Poxviruses that infect a wide range of vertebrate species are grouped into 18 genera based originally on their serological reactions, but more recently by their genomic features [1]. Among these poxviruses, members of the genera orthopoxvirus include many of the commonly recognized human pathogens such as variola virus (VARV), the causative agent of smallpox, cowpox virus (CPXV), and monkeypox virus (MPXV). Outside the orthopoxvirus genus, examples of poxviruses with human tropism include molluscum contagiosum virus and tanapox virus (TPV). Most poxviruses have evolved within a small number of host species with which they share co-evolutionary history, however, in lab culture, they can frequently infect cells from different host species. This broader cellular infectivity, compared with more limited host specificity, is mainly due to the lack of requirement for selective receptor proteins on target cells. Unlike most other mammalian viruses, poxviruses rely on relatively ubiquitous cellular surface molecules and exploit multiple host and virus-encoded proteins required for cell binding, fusion, and entry processes 2, 3, 4. At the cellular level, since poxviruses can bind and enter most mammalian cells in vitro, tropism is largely determined by the viruses' ability to modulate diverse intracellular antiviral pathways activated in response to virus sensing and infection. However, at the host organism level, the innate antiviral pathways activated by different virus-induced cytokines play a major role in determining the poxvirus tropism 5, 6. Here, we specifically discuss the recent progress in understanding how these crucial host cytokines regulate poxvirus tropism and adaptation.
The linear dsDNA genome of poxviruses ranges from 130 to 375 thousand base pairs and encodes between 130 and 300 open-reading frames. The central region of the genome is highly conserved among poxviruses and includes many dozens of essential genes required for transcription, replication, and virion assembly. The two ends of the linear viral genome are much more variable and encode host-interactive genes that control host range, help evade host immune responses, and other functions to control cellular responses. In addition, the function of these viral genes (referred to as host range genes) is closely linked to the successful replication of poxviruses in cultured cells originating from different tissues and hosts. However, the role of many poxvirus-specific host range genes uniquely required for host tropism and evasion of host immune responses is not well characterized from different poxviruses [7].
Host-restricted and zoonotic infection of poxviruses
One of the hallmarks of poxviruses is their host-restricted infections at the organismal level in vivo. Almost every vertebrate species can be infected with a selected member of poxviruses, some of which cause disease and some cause only subclinical infections. Genome sequencing of these poxviruses has identified a few viral genes unique to that poxvirus, and functional studies of such genes suggest that they have acquired host-specific functions [8]. For example, recently identified C7L-like host range gene M159 in MYXV-Tol (myxoma virus isolate Toledo), a member of Leporipoxvirus and known to cause disease in European rabbits, is critical for the recent species leap that now causes lethal disease in hares ( Figure 1). In culture, the recombinant knockout construct of MYXV-Tol lacking M159 can no longer productively infect neither a hare cell line nor primary hare PBMCs (peripheral blood mononuclear cells) [9••]. On the other hand, MYXV-Lau (myxoma virus isolate Lausanne), which causes myxomatosis in European rabbits and lacks Tol-M159, cannot infect this hare cell line nor primary hare PBMCs. However, the construction of a recombinant MYXV-Lau expressing just the Tol-M159 gene now allowed MYXV-Lau to replicate in both immortalized and primary hare cells. Although the host cell target(s) of M159 are yet to be identified, these results suggest that even the genetic acquisition of a single viral host range protein function can dramatically alter the tropism of recipient poxvirus. Similarly, C7L-like host range proteins from other poxviruses contribute to their host and cellular tropism [10]. Functional and structural studies of these C7-like proteins demonstrated that some of them bind and antagonize host sterile alpha-motif domain-containing 9 protein to overcome type-I IFN-mediated host restriction 11, 12, 13•. Thus, poxvirus-encoded proteins known as host range factors can dictate which cells, tissues, or hosts they can productively infect.Figure 1 Myxoma virus (MYXV) species leap from rabbits to hares. MYXV-Lau causes myxomatosis only in European rabbits (Oryctolagus cuniculus). However, during a recombination event, MYXV-Lau acquired a genomic cassette with a C7-like host range gene called M159 from an unknown poxvirus. This new MYXV isolate called MYXV-Tol can now cause myxomatosis-like disease in Iberian hares (Lepus granatensis) and European rabbits.
Figure 1
Apart from the virus-encoded proteins, host-specific factors and immune functions critically impact poxvirus tropism. The very well-studied poxvirus for host specificity is ectromelia virus (ECTV), a mouse-specific orthopoxvirus with a very narrow rodent-specific host range in nature. ECTV causes high mortality in susceptible mice strains, including BALB/c, DBA/2, A/J, and C3H, whereas C57BL/6, AKR, and I29 strains are much more resistant to the disease known also as mousepox 14, 15. In ECTV-resistant mice, multiple genetic loci have been identified and are referred to as restriction factors. For example, Ly49H (also called resistance to mousepox-1, Rmp-1) maps to the natural killer gene complex (NKC) and activates NK cells to control early virus replication in C57BL/6 mice, but is lacking in BALB/c mice [16]. Cytokines such as type-I IFN, IL-12, and IL-18 play essential roles in mediating this inherent genetic resistance to mousepox. These studies revealed that host-specific cytokine responses largely contribute to the tropism of poxviruses [17].
Unlike ECTV, some other orthopoxviruses are naturally capable of leaping from their reservoir host species to cause zoonotic diseases, including MPXV, CPXV, vaccinia virus (VACV)-like and Akhmeta virus. These viruses naturally circulate in wild and domestic animals, where they may or may not induce disease, have a broader host range, and often cause disease outbreaks in humans and other animals. For example, MPXV infections have been reported in various rodents, such as mice, rats, rabbits, hamsters, woodchucks, jerboas, porcupines, prairie dogs, hedgehogs, and several nonhuman primate species 18, 19. Although humans are considered accidental hosts, MPXV has now become the major zoonotic poxvirus for humans since the eradication of smallpox [20••]. Similarly, CPXV reservoirs are exclusively rodents in nature, but many wild animals can become accidental hosts, including cats, dogs, elephants, diverse zoo animals, and nonhuman primates from where humans can acquire an infection. Thus, poxviruses that can naturally leap into multiple host species are believed to have the further potential to acquire additional host-adapted mutations or acquired novel host regulatory proteins that can antagonize cytokine responses in diverse hosts. Apart from orthopoxviruses, members of capripoxvirus and parapoxvirus genera that normally infect farm animals can also infect humans after direct contact transmission, suggesting that these viruses encode host range proteins that can modulate human cytokine responses [21].
How cytokine-mediated innate immune responses regulate poxvirus host-specific infections and tropism
Cytokines and interferons (IFNs) are extracellular signaling molecules that play a key role in mediating an early immune response against invading pathogens, including viruses, and are essential components of host defense. In most cases, cytokine(s) activate protective responses that can provide complete clearance of viral infection. These cytokines, which can be either anti-inflammatory or pro-inflammatory, eventually clear the virus-infected cells by activating diverse mechanisms, including inflammation. These critical cytokines include IFNs, tumor necrosis factor (TNF), interleukin-1 (IL-1), IL-12, IL-18, as well as multiple chemokines. Furthermore, these cytokines alone or in combination with each other, can further activate a network of downstream signaling pathways and stimulated genes such as interferon-stimulated genes (ISGs) and TNF-stimulated genes (TSGs).
Apart from host cytokines, poxviruses also encode diverse immune modulatory factors that can counteract the antiviral responses activated by different cytokines, including ISGs and TSGs, to determine the host-specific tropism and disease caused by different poxviruses. For example, VACV has been shown to be relatively resistant to IFN responses in cells from different species and also can confer resistance to IFN to other viruses such as vesicular stomatitis virus 22, 23. On the other hand, MYXV-Lau, a leporipoxvirus, exhibits resistance to IFN only in rabbit cells, but the virus is relatively sensitive to the IFN-induced antiviral state in human or mouse cells [24]. This species-specific anti-IFN function is mainly because MYXV-Lau host range factors such as dsRNA-binding vaccinia E3-like protein M029 have specialized in inhibiting restriction factors present in rabbits but not in other species that are yet to be identified [25]. For most of the cytokines that function to protect against viruses, poxviruses have co-evolved encoded proteins that dampen their functions at a different level. They encode self-defense proteins that can directly target the cytokine and prevent it from receptor binding or other interactions, proteins that function as a decoy receptor that also directly binds cytokines and thus competes with the natural receptor, and proteins that can inhibit or modulate the downstream intracellular signaling networks activated by different cytokines [7]. In most cases, whether these virus-encoded proteins targeting different cytokines and their network subsequently determine the tropism of poxviruses is yet to be studied in greater detail [24]. Several reviews have focused on this topic of how viruses counteract different cytokines, and it is beyond the scope of this mini-review 23, 26, 27, 28, 29•, 30, 31.i) Interferons:
IFNs are the key cytokines that are rapidly produced and released from the cells in response to virus infection or by sensing virus-induced ligands such as pathogen-associated molecular patterns or damage-associated molecular patterns. Subsequently, the released IFNs bind to IFN receptors on the surface of target cells to trigger signaling pathways that activate the expression of hundreds of inducible genes known as ISGs 32, 33. There are three types of IFNs, namely type-I, type-II, and type-III IFNs, which have many subtypes. Some ISGs can be upregulated by all IFNs, while others are upregulated by selective IFNs. For example, Interferon Regulatory Factor 1 (IRF1) is upregulated preferentially by IFN-alpha and not by IFN-gamma [34]. This type of selective IFN-induced response is vital for cell, tissue, or host-specific innate immune responses and generation of an antiviral state within responsive cells, thereby controlling selective virus infection and spread 35, 36. Thus, one can predict that the orchestration of the expression of host-specific ISGs can alter the cellular or host tropism of different viruses, including poxviruses. Poxvirus-related functions of some of the key ISGs, such as ISG15, protein kinase R, and 2′,5′-oligoadenylate synthetases, are well characterized 37, 38, 39. However, many more remain to be studied in greater detail. As mentioned before, in the case of poxviruses, most of our knowledge about the role of IFNs in tropism has come from studies on ECTV using different strains of mice and genetic knockouts of C57BL/6 mice [14]. In ECTV-resistant C57BL/6 mice, apart from different genetic loci, potent NK, cytotoxic T lymphocytes (CTLs), and IFNɣ responses are generated against ECTV infection at higher levels than in ECTV-susceptible BALB/c mice [15]. Further studies in C57BL/6 mice with genetic deficiencies in innate immune pathways such as TLR9–MyD88–IRF7 and STING–IRF7/NF-kB confirmed that inefficient production of type-I IFNs will increase mortality in C57BL/6 mice [40]. Crosstalk between IFN-I and NF-κB pathways also confers resistance to lethal poxvirus infection [41]. A recent study demonstrated that IFN-I response is required in a cell-type-specific manner: C57BL/6 mice lacking IFNR in NK cells and monocytes become sensitive to disease caused by ECTV [42••]. In addition, at the cellular level, different DNA sensing pathways such as cyclic GMP-AMP synthase (cGAS) and stimulator of interferon genes (STING) can trigger the production of type-I IFNs in selected cell types and thus regulate poxvirus cellular tropism 23, 43, 44•, 45. Using ECTV, it was shown that bone marrow-derived cells play a major role in cGAS-dependent IFN production and protection against ECTV [45]. However, ECTV-encoded protein Schlafen (vSlfn) has been identified as the primary inhibitor of the cGAS–STING pathway, without which the virus is severely attenuated [44•]. Apart from ECTV, using VACV and mice knocked out for various innate sensing molecules, the role of IFNs in protection against infection and pathogenesis has been extensively documented 23, 27.
Type-I IFNs also play a major role in controlling the infection of MYXV in cells derived from mice, humans, and likely other vertebrate species. For example, in mouse primary embryonic fibroblasts, virus-mediated induction of type-I IFN through ERK and IRF3 signaling pathway completely inhibits MYXV replication. Mice that are genetically lacking STAT1 and thus defective in IFN signaling became susceptible to lethal MYXV infection after intracranial injection [46]. Thus, the ERK–IFN–STAT1 pathway contributes to the species-specific protection against MYXV infection in host species outside of lagomorphs. However, MYXV has evolved rabbit-specific strategies that can inhibit this highly conserved pathway to cause lethal disease in rabbits. This rabbit-specific specificity is likely related to the evolutionary time, estimated to be at least 10 million years, that MYXV has co-evolved with South American rabbits. It is anticipated that IFN signaling contributes to the selective tropism of most, if not all, chordopoxviruses. Apart from the natural innate protection against certain poxviruses, the host can also co-evolve with the virus to acquire genetically controlled innate immunity against the selective pressure from poxviruses. This is evident from the genomic sequencing of feral rabbits from Australia and Europe, and it was found that the evolution of resistance in European rabbits to MYXV is associated with enhanced innate antiviral immunity that was acquired in just the 70 years following the first release of MYXV into wild rabbit populations in the early 1950s [47]. MYXV, on the other hand, also evolved to overcome these newly acquired innate host defenses 48, 49. Thus, the ongoing dynamics of the virus-host battle can shape and reshape the host tropism of poxviruses.
ii) TNF and TNF superfamily cytokines:
TNF is a potent pro-inflammatory cytokine that plays an essential role in the host control of many viral infections, including poxviruses [26]. TNF and the TNF ligand superfamily members bind to TNF receptor (TNFR) superfamily members to trigger downstream TNFR signaling that leads to the induction of hundreds of TSGs. TNF and the TNF signaling network directly play key roles in protection against poxvirus infections. For example, C57BL/6 mouse strains that are naturally resistant to lethal ECTV infection produce high levels of TNF and potent immune responses than susceptible BALB/c strain that produces little TNF and only weak immune responses. However, C57BL/6 mouse lacking TNFRs or TNF (TNF-/-) becomes much more susceptible to ECTV infection and disease 50, 51•. Furthermore, exogenous treatment with mouse TNF was shown to reduce ECTV replication and mortality in mice [52]. Similarly, in TNFR2-knockout C57BL/6 mice, infection with VACV resulted in higher viral loads in spleens and livers and defective viral clearance [53]. Like type-I IFN, TNF also plays a role in restriction of MYXV in primary human macrophages [54]. Treatment of human fibroblasts with IFN-β and TNF resulted in activation of a synergistic antiviral state that can completely restrict MYXV and other poxviruses such as VACV and TPV that infect humans [55]. These studies suggest that, apart from the independent action of individual cytokines, their cooperative action can also further alter the tropism of poxviruses and other viruses [6]. Indeed, the treatment of human cells with TNF plus IFN synergistically upregulates an additional set of downstream genes that neither cytokine alone will induce [55]. This strongly suggests that co-induction of multiple cytokines at the organismal level plays a key role in tropism within tissues and host organisms that is not well modeled by assessing individual cytokines in cultured cells.
iii) Other cytokines:
IL-18 is a pleiotropic pro-inflammatory cytokine belonging to the IL-1 superfamily. IL-18 plays an important regulatory role in both innate and acquired immune responses against diverse pathogens, including poxviruses [56]. IL-18 signals through membrane-bound IL18Rα and IL18Rβ, which then stimulates the production of IFNɣ from T-helper lymphocyte cells (Th1) and macrophages and also enhances the cytotoxicity of NK cells. IL-12 is a member of the heterodimeric cytokines, composed of two chains, IL12A (p35) and IL12B (p40). IL-12 signals through a heterodimeric receptor formed by IL12Rβ1 and IL12Rβ2 to activate NK and T cells to stimulate the production of antiviral cytokines such as IFNɣ and TNF [57]. IL-12 and IL-18 are important for the cell-mediated immune response against poxviruses [58]. C57BL/6 mice that are lacking either IL12p40 (IL12p40-/-) or IL-18 (IL-18-/-) or both cytokines (double-knockout IL12p40-/-IL18-/-) are becoming highly susceptible to ECTV infection [58]. In these mice, the Th1 cytokine response was diminished, but the Th2 cytokine response was enhanced. In addition, there were reduced cytotoxic NK cells and CTL responses, resulting in reduced proliferation of virus-specific CD8+ T cells compared with the wild-type mice [58]. Thus, IL12p40 and IL-18 play an important role in the activation of antiviral responses by upregulation of IFNɣ production and cell-mediated immune responses. The role of IL-12 and IL-18 in activating innate and adaptive immune responses against viruses was further tested using recombinant VACVs expressing either IL-12 or IL-18 alone or both cytokines together [59]. Expression of either IL-12 or IL-18 enhanced viral clearance from ovaries and spleen in virus-infected mice, which involved NK and T cells [59]. However, when both cytokines were expressed from the same virus, the Th1 response increased for virus clearance [59]. Thus, IL-12 and IL-18 synergistically activate the antiviral response. In another study, expression of IL-18 using VACV resulted in attenuation of virus replication but elicited improved CTL responses [60]. An IL-12-expressing MYXV construct was also attenuated in European rabbits, suggesting that these cytokines can alter the tropism of poxviruses in hosts, and profoundly regulated the viral disease manifestations [61].
Conclusions
Cytokines are the gatekeeper and among the first lines of host defense against invading pathogens, including poxviruses. On the other hand, successful viruses have demonstrated the ability to emerge, re-emerge, or persist in a host in a fashion that is linked to their ability to subvert or evade antiviral cytokine responses. The outcome of such host and virus interactions determines the overall tropism of the virus at the cellular, tissue, and host level. Among the DNA viruses, poxviruses are known to circulate in almost every vertebrate species and can cause disease in host-restricted manner. However, poxviruses are also known to leap species, occasionally re-emerge, and cause zoonotic infections. For example, MPXV was long thought to be a rodent poxvirus in Africa that only occasionally caused disease in humans as a dead-end infection with a secondary human-to-human attack rate of less than 10%. But the current worldwide MPXV epidemic revealed the potential for extended human transmission that is linked to human behavior rather than evolution of new genetic viral variants (which, of course, may still occur). But host species leaping of poxviruses can also be due to either acquiring novel genes or selecting mutations in key immune-evading proteins that allow dampening of the cytokine responses in the newly adapted host 62•, 63. For example, VARV, which caused smallpox and killed millions of people per year for centuries, may have jumped from an unknown precursor host species or reservoir thousands of years ago and then adapted to humans after the original host had gone extinct [64]. During its adaptation in humans, VARV has lost multiple genes from the most recent common ancestor, suggesting that mutation or gene loss can enhance host- specific virulence of poxviruses 65, 66••. After the successful eradication of smallpox by a very successful worldwide vaccination program, newer emerging poxviruses such as MPXV and CPXV are appearing as ‘new’ zoonotic viruses and are becoming a progressively bigger threat to human health. The current MPXV outbreak and spread in many countries in populations that are not vaccinated prove that they are still a global threat. Sequencing of the circulating MPXV clades suggests that the MPXV has acquired mutations in certain genes involved in regulating host responses, compared with the apparent precursor virus from West Africa, but it is unknown if any of these mutations are responsible for the apparent increases in human-to-human transmission 20••, 67•, 68. More functional studies can only reveal whether they newly acquired host immune regulatory functions. Understanding the role of cytokines and how poxviruses counteract them also has implications for the development of vaccines, antiviral drugs, use of poxviruses as a vaccine platform, expression vector, and oncolytic viruses for the treatment of cancers.
Funding
This work was supported by 10.13039/100000002 National Institutes of Health grant R01 AI080607 and R21 AI163910 to G.M. and M.M.R. The funders had no role in study design, data collection, and interpretation or the decision to submit the work for publication.
Conflict of interest statement
The authors declare no conflict of interest.
Data availability
No data were used for the research described in the article.
==== Refs
References and recommended reading
Papers of particular interest, published within the period of review, have been highlighted as:• of special interest
•• of outstanding interest
1 Odom M.R. Hendrickson R.C. Lefkowitz E.J. Poxvirus protein evolution: family wide assessment of possible horizontal gene transfer events Virus Res 144 2009 233 249 19464330
2 Moss B. Poxvirus cell entry: how many proteins does it take? Viruses 4 2012 688 707 22754644
3 Moss B. Membrane fusion during poxvirus entry Semin Cell Dev Biol 60 2016 89 96 27423915
4 Schmidt F.I. Bleck C.K. Mercer J. Poxvirus host cell entry Curr Opin Virol 2 2012 20 27 22440962
5 Bartee E. McFadden G. Cytokine synergy: an underappreciated contributor to innate anti-viral immunity Cytokine 63 2013 237 240 23693158
6 McFadden G. Cytokine determinants of viral tropism Nat Rev Immunol 9 2009 645 655 19696766
7 Seet B.T. Poxviruses and immune evasion Annu Rev Immunol 21 2003 377 423 12543935
8 Bratke K.A. McLysaght A. Rothenburg S. A survey of host range genes in poxvirus genomes Infect Genet Evol 14 2013 406 425 23268114
9•• Águeda-Pinto A. Identification of a novel myxoma virus C7-like host range factor that enabled a species leap from rabbits to hares mBio 13 2022 e0346121
Making different knock-out and knock-in recombinant viruses, this study demonstrates that the C7-like host range factor present in the myxoma virus isolate Toledo is responsible for the infection of hare cells and caused myxoma virus species leap from rabbits to hares.
10 Liu J. Rothenburg S. McFadden G. The poxvirus C7L host range factor superfamily Curr Opin Virol 2 2012 764 772 23103013
11 Meng X. Structural basis for antagonizing a host restriction factor by C7 family of poxvirus host-range proteins Proc Natl Acad Sci USA 112 2015 14858 14863 26578811
12 Meng X. C7L family of poxvirus host range genes inhibits antiviral activities induced by type I interferons and interferon regulatory factor 1 J Virol 86 2012 4538 4547 22345458
13• Conrad S.J. Myxoma virus lacking the host range determinant M062 stimulates cGAS-dependent type 1 interferon response and unique transcriptomic changes in human monocytes/macrophages PLoS Pathog 18 2022 e1010316
This study demonstrates that the C7L-like host range factor M062 from myxoma virus isolate Lausanne regulates the SAMD-9-dependent DNA sensing pathway.
14 Esteban D.J. Buller R.M.L. Ectromelia virus: the causative agent of mousepox J Gen Virol 86 2005 2645 2659 16186218
15 Sigal L.J. The pathogenesis and immunobiology of mousepox Adv Immunol 129 2016 251 276 26791861
16 Fang M. CD94 is essential for NK cell-mediated resistance to a lethal viral disease Immunity 34 2011 579 589 21439856
17 Cheng W.Y. Comparison of host gene expression profiles in spleen tissues of genetically susceptible and resistant mice during ECTV infection Biomed Res Int 2017 2017 6456180
18 Beer E.M. Rao V.B. A systematic review of the epidemiology of human monkeypox outbreaks and implications for outbreak strategy PLoS Negl Trop Dis 13 2019 e0007791
19 Alakunle E. Monkeypox virus in Nigeria: infection biology, epidemiology, and evolution Viruses 12 2020 11
20•• Isidro J. Phylogenomic characterization and signs of microevolution in the 2022 multi-country outbreak of monkeypox virus Nat Med 28 2022 1569 1572 35750157
Sequencing of the monkeypox virus isolates demonstrates that the monkeypox virus outbreak in 2022 is most likely from a single origin, and mutational analysis suggests potential human adaptation.
21 Silva N.I.O. Here, there, and everywhere: the wide host range and geographic distribution of zoonotic orthopoxviruses Viruses 13 2020 1
22 Whitaker-Dowling P. Youngner J.S. Vaccinia rescue of VSV from interferon-induced resistance: reversal of translation block and inhibition of protein kinase activity Virology 131 1983 128 136 6316646
23 Smith G.L. Talbot-Cooper C. Lu Y. How does vaccinia virus interfere with interferon? Adv Virus Res 100 2018 355 378 29551142
24 McFadden G. Poxvirus tropism Nat Rev Microbiol 3 2005 201 213 15738948
25 Rahman M.M. McFadden G. Myxoma virus dsRNA binding protein M029 Inhibits the Type I IFN-Induced Antiviral State in a Highly Species-Specific Fashion Viruses 9 2017 2
26 Alvarez-de Miranda F.J. TNF decoy receptors encoded by poxviruses Pathogens 10 2021 8
27 El-Jesr M. Teir M. Maluquer de Motes C. Vaccinia virus activation and antagonism of cytosolic DNA sensing Front Immunol 11 2020 568412
28 Lawler C. Brady G. Poxviral targeting of interferon regulatory factor activation Viruses 12 2020 10
29• Hernaez B. Alcamí A. Virus-encoded cytokine and chemokine decoy receptors Curr Opin Immunol 66 2020 50 56 32408109
The review discusses the current knowledge on how virus-encoded cytokine and chemokine decoy receptors regulate the host antiviral responses mediated by cytokines and chemokines.
30 Nelson C.A. Structural conservation and functional diversity of the Poxvirus Immune Evasion (PIE) domain superfamily Viruses 7 2015 4878 4898 26343707
31 Smith G.L. Vaccinia virus immune evasion: mechanisms, virulence and immunogenicity J Gen Virol 94 2013 2367 2392 23999164
32 Schneider W.M. Chevillotte M.D. Rice C.M. Interferon-stimulated genes: a complex web of host defenses Annu Rev Immunol 32 2014 513 545 24555472
33 Schoggins J.W. A diverse range of gene products are effectors of the type I interferon antiviral response Nature 472 2011 481 485 21478870
34 Galon J. IL-12 induces IFN regulating factor-1 (IRF-1) gene expression in human NK and T cells J Immunol 162 1999 7256 7262 10358173
35 Wu X. Intrinsic immunity shapes viral resistance of stem cells Cell 172 2018 423 438.e25 29249360
36 Hubel P. A protein-interaction network of interferon-stimulated genes extends the innate immune system landscape Nat Immunol 20 2019 493 502 30833792
37 Burgess H.M. Mohr I. Cellular 5′-3′ mRNA exonuclease Xrn1 controls double-stranded RNA accumulation and anti-viral responses Cell Host Microbe 17 2015 332 344 25766294
38 Liu S.W. Poxvirus decapping enzymes enhance virulence by preventing the accumulation of dsRNA and the induction of innate antiviral responses Cell Host Microbe 17 2015 320 331 25766293
39 Eduardo-Correia B. ISG15 is counteracted by vaccinia virus E3 protein and controls the proinflammatory response against viral infection J Virol 88 2014 2312 2318 24257616
40 Xu R.H. Sequential activation of two pathogen-sensing pathways required for Type I interferon expression and resistance to an acute DNA virus infection Immunity 43 2015 1148 1159 26682986
41 Rubio D. Crosstalk between the type 1 interferon and nuclear factor kappa B pathways confers resistance to a lethal virus infection Cell Host Microbe 13 2013 701 710 23768494
42•• Melo-Silva C.R. Resistance to lethal ectromelia virus infection requires Type I interferon receptor in natural killer cells and monocytes but not in adaptive immune or parenchymal cells PLoS Pathog 17 2021 e1009593
This study demonstrates the importance of a functional type I interferon signaling pathway in monocytes and natural killer cells to protect against the lethal ectromelia virus infection in C57BL/6 mice.
43 Cheng W.Y. The cGas-Sting signaling pathway is required for the innate immune response against ectromelia virus Front Immunol 9 2018 1297
44• Hernáez B. Viral cGAMP nuclease reveals the essential role of DNA sensing in protection against acute lethal virus infection Sci Adv 6 2020 38
This study demonstrates that orthopoxvirus-encoded viral Schlafen vSlfn is the primary inhibitor of antiviral innate immune responses activated by the cGAS-STING pathway, and in the absence of this protein, mouse poxvirus pathogenesis is severly reduced.
45 Wong E.B. Resistance to ectromelia virus infection requires cGAS in bone marrow-derived cells which can be bypassed with cGAMP therapy PLoS Pathog 15 2019 e1008239
46 Wang F. Disruption of Erk-dependent type I interferon induction breaks the myxoma virus species barrier Nat Immunol 5 2004 1266 1274 15502830
47 Alves J.M. Parallel adaptation of rabbit populations to myxoma virus Science 363 2019 1319 1326 30765607
48 Kerr P.J. Next step in the ongoing arms race between myxoma virus and wild rabbits in Australia is a novel disease phenotype Proc Natl Acad Sci USA 114 2017 9397 9402 28808019
49 Kerr P.J. Punctuated evolution of myxoma virus: rapid and disjunct evolution of a recent viral lineage in Australia J Virol 93 2019 8
50 Ruby J. Bluethmann H. Peschon J.J. Antiviral activity of tumor necrosis factor (TNF) is mediated via p55 and p75 TNF receptors J Exp Med 186 1997 1591 1596 9348317
51• Tuazon Kels M.J. TNF deficiency dysregulates inflammatory cytokine production, leading to lung pathology and death during respiratory poxvirus infection Proc Natl Acad Sci USA 117 2020 15935 15946 32571912
This study demonstrates the importance of a functional TNF signaling pathway, including transmembrane TNF that regulates the production of other proinflammatory cytokines to protect the host against lethal virus infection.
52 Atrasheuskaya A.V. Protective effect of exogenous recombinant mouse interferon-gamma and tumour necrosis factor-alpha on ectromelia virus infection in susceptible BALB/c mice Clin Exp Immunol 136 2004 207 214 15086382
53 Chan F.K. A role for tumor necrosis factor receptor-2 and receptor-interacting protein in programmed necrosis and antiviral responses J Biol Chem 278 2003 51613 51621 14532286
54 Wang F. RIG-I mediates the co-induction of tumor necrosis factor and type I interferon elicited by myxoma virus in primary human macrophages PLoS Pathog 4 2008 e1000099
55 Bartee E. The addition of tumor necrosis factor plus beta interferon induces a novel synergistic antiviral state against poxviruses in primary human fibroblasts J Virol 83 2009 498 511 18971273
56 Ihim S.A. Interleukin-18 cytokine in immunity, inflammation, and autoimmunity: Biological role in induction, regulation, and treatment Front Immunol 13 2022 919973
57 Hamza T. Barnett J.B. Li B. Interleukin 12 a key immunoregulatory cytokine in infection applications Int J Mol Sci 11 2010 789 806 20479986
58 Wang Y. IL-12p40 and IL-18 play pivotal roles in orchestrating the cell-mediated immune response to a poxvirus infection J Immunol 183 2009 3324 3331 19657092
59 Gherardi M.M. Ramírez J.C. Esteban M. IL-12 and IL-18 act in synergy to clear vaccinia virus infection: involvement of innate and adaptive components of the immune system J Gen Virol 84 2003 1961 1972 12867626
60 Verardi P.H. IL-18 expression results in a recombinant vaccinia virus that is highly attenuated and immunogenic J Interferon Cytokine Res 34 2014 169 178 24168450
61 Stanford M.M. Myxoma virus expressing human interleukin-12 does not induce myxomatosis in European rabbits J Virol 81 2007 12704 12708 17728229
62• Senkevich T.G. Ancient gene capture and recent gene loss shape the evolution of orthopoxvirus-host interaction genes mBio 12 2021 e0149521
Genomic analysis of orthpoxvirus accessory genes responsible for virus-host interactions suggests that most of the accessory genes were captured early in chordopoxvirus evolution in host followed by extensive gene duplication resulting in several paralogous gene families.
63 Hendrickson R.C. Orthopoxvirus genome evolution: the role of gene loss Viruses 2 2010 1933 1967 21994715
64 Thèves C. Crubézy E. Biagini P. History of smallpox and its spread in human populations Microbiol Spectr 4 2016 4
65 Alcamí A. Was smallpox a widespread mild disease? Science 369 2020 376 377 32703866
66•• Mühlemann B. Diverse variola virus (smallpox) strains were widespread in northern Europe in the Viking Age Science 369 2020 6502
A significant study demonstrated the presence of variola virus (smallpox) strains during the Viking Age and how the virus evolved to kill millions of people worldwide.
67• Wang L. Genomic annotation and molecular evolution of monkeypox virus outbreak in 2022 J Med Virol 2022 10.1002/jmv.28036
A study demonstrating that the strains of monkeypox virus 2022 outbreak are undergoing mutations from the closely related strain isolated in 2018.
68 Khosravi E. Keikha M. B.1 as a new human monkeypox sublineage that linked with the monkeypox virus (MPXV) 2022 outbreak - Correspondence Int J Surg 105 2022 106872
| 36427482 | PMC9704024 | NO-CC CODE | 2022-12-09 23:15:07 | no | Curr Opin Virol. 2022 Dec 22; 57:101286 | utf-8 | Curr Opin Virol | 2,022 | 10.1016/j.coviro.2022.101286 | oa_other |
==== Front
Infect Dis Model
Infect Dis Model
Infectious Disease Modelling
2468-2152
2468-0427
KeAi Publishing
S2468-0427(22)00091-4
10.1016/j.idm.2022.11.008
Article
Modelling the impact of timelines of testing and isolation on disease control
Li Ao
Wang Zhen
Moghadas Seyed M. ∗
Agent-Based Modelling Laboratory, York University, Toronto, Ontario, M3J 1P3, Canada
∗ Corresponding author.
28 11 2022
3 2023
28 11 2022
8 1 5871
15 10 2022
21 11 2022
© 2022 The Authors
2022
Elsevier has created a Monkeypox Information Center (https://www.elsevier.com/connect/monkeypox-information-center) in response to the declared public health emergency of international concern, with free information in English on the monkeypox virus. The Monkeypox Information Center is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its monkeypox related research that is available on the Monkeypox Information Center - including this research content - immediately available in publicly funded repositories, with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the Monkeypox Information Center remains active.
Testing and isolation remain a key component of public health responses to both persistent and emerging infectious diseases. Although the value of these measures have been demonstrated in combating recent outbreaks including the COVID-19 pandemic and monkeypox, their impact depends critically on the timelines of testing and start of isolation during the course of disease. To investigate this impact, we developed a delay differential model and incorporated age-since-symptom-onset as a parameter for delay in testing. We then used the model to compare the outcomes of reverse-transcription polymerase chain reaction (RT-PCR) and rapid antigen (RA) testing methods when isolation starts either at the time of testing or at the time of test result. Parameterizing the model with estimates of SARS-CoV-2 infection and diagnostic sensitivity of the tests, we found that the reduction of disease transmission using the RA test can be comparable to that achieved by applying the RT-PCR test. Given constraints and inevitable delays associated with sample collection and laboratory assays in RT-PCR testing post symptom onset, self-administered RA tests with short turnaround times present a viable alternative for timely isolation of infectious cases.
Keywords
Testing
Isolation
Delay equations
Simulations
Turnaround time
Handling Editor: Dr HE DAIHAI
==== Body
pmc1 Introduction
Emerging infectious diseases often pose significant and unique challenges for their controllability. Pandemic diseases are particularly concerning due to their scale and speed of spread globally (Morens et al., 2020). The COVID-19 pandemic has demonstrated these challenges, especially at the early stages of SARS-CoV-2 emergence in the absence of vaccination.
One of the key measures that have been widely used during COVID-19 is testing to identify infected cases and isolate them to reduce onward transmission (CDC, 2022b; Grassly et al., 2020; MacIntyre, 2020; Mercer & Salit, 2021; Wells et al., 2021). Prior to the widespread availability of RA tests, RT-PCR was the predominant approach to identifying SARS-CoV-2 infections (Rosenberg & Holtgrave, 2021; Wells et al., 2021, 2022b,a). However, RT-PCR requires laboratory assay, a process that often delays the test results after sample collection, which could lead to a late start of isolation (Rosenberg & Holtgrave, 2021). Even without considering resources required for, and costs associated with RT-PCR method, the scale of testing could overwhelm the system with limited capacity, further lengthening the turnaround times (Kucharski et al., 2020). While awaiting test results, the possibility of disease transmission from positive cases can be reduced by initiating isolation at the time of sample collection.
The availability of RA tests has provided an alternative means of scaling up testing capacity and shortening turnaround times from days to minutes (Wells, Pandey, Moghadas, et al., 2022). The low-cost, self-administered RA tests can mitigate the impact of disease through the immediate start of isolation for positive cases. However, diagnostic sensitivity of RA tests may be lower than RT-PCR, which could lead to higher rates of false-negatives (Wells, Pandey, Moghadas, et al., 2022).
Understanding the effect of timelines of testing and isolation on disease dynamics can help to improve testing strategies. In this study, we aimed to formulate a delay-differential model to investigate the dynamics of disease transmission encapsulating three important factors: (i) time of testing after the onset of disease symptoms; (ii) delay in turnaround times of the test results and start of isolation; and (iii) diagnostic sensitivity of the test which varies with time. We formulated our model with the natural history of COVID-19, and simulated it to illustrate the findings using parameters estimated for SARS-CoV-2 infection, while presenting relevant theoretical analyses.
2 Model formulation
To develop our model from the classical susceptible-infected-recovered (SIR) framework, we made the following assumptions:(A1) Disease transmission occurs through contacts between susceptible and infectious individuals in a mass action form.
(A2) Immunity generated from natural infection is long-lasting, hence omitting the possibility of reinfection following recovery.
(A3) The epidemic timelines are short compared to the lifespan of individuals, thus excluding the effect of demographics such as births and natural deaths.
Denoting the class of susceptible individuals by S, we haveS'(t)=-βS(t)Λ(t),E'(t)=βS(t)Λ(t)-σE(t),
where ‘′’ represents the derivative of the variables with respect to time t and Λ(t) is the force of infection which will be formulated later. Upon infection, individuals will be placed in the E class for a duration of time (referred to as latent period) before becoming infectious. During latency, the disease is not transmissible. We assumed that a proportion p of latent individuals develop symptoms, and the remaining will be asymptomatic without presenting any clinical symptoms of the disease during the course of infection. Denoting asymptomatic individuals by A, we haveA′(t)=(1−p)σE(t)−γAA(t),
where 1/σ represents the average latent period and 1/γ A is the average duration of asymptomatic infectiousness.
Infected individuals who become symptomatic will be infectious in a stage of disease prior to the onset of symptoms, referred to as pre-symptomatic. Letting P denote this stage with an average duration of 1/γ P, we have(1) P′(t)=pσE(t)−γPP(t).
For identification of infected individuals, we assume that only symptomatic cases will be tested, and isolated once the results are confirmed as positive. Let a represent the time elapsed since the onset of symptoms. We define i(t, a), v(t, a), k(t, a), and w(t, a) to be the densities of populations, with respect to age a at time t, who are not tested, tested and waiting for the results, isolated with positive results, and not isolated due to false-negative results. At any age a post symptom onset, the testing rate is defined by μ(a) with the diagnostic sensitivity q(a). The dynamics of population densities are thus governed by the following first-order partial differential equations(2) ∂i(t,a)∂t+∂i(t,a)∂a=-μ(a)i(t,a)-γi(t,a),∂v(t,a)∂t+∂v(t,a)∂a=μ(a)i(t,a)-μ(a-τ)e-γτi(t-τ,a-τ)-γv(t,a),∂k(t,a)∂t+∂k(t,a)∂a=q(a-τ)μ(a-τ)e-γτi(t-τ,a-τ)-γk(t,a),∂w(t,a)∂t+∂w(t,a)∂a=(1-q(a-τ))μ(a-τ)e-γτi(t-τ,a-τ)-γw(t,a),
where 1/γ is the average infectious period after the onset of symptoms, and τ is the time delay in receiving the test results. Let a 0 > 0 denote the minimum amount of time elapsed post symptom onset before performing a test. We define μ(a) byμ(a)=Ce−1/a/aa0≤a≤1γ0otherwise
In this formulation, [a 0, 1/γ] represents the window of opportunity for testing after the onset of symptoms. In general, testing and isolation of individuals beyond the average duration of infectiousness for symptomatic disease will not be effective in reducing transmission. The testing rate increases with a to a maximum at one (a = 1) day after the onset of symptoms and then declines. We considered this rate to reasonably reflect the reality of RT-PCR testing which may involve traveling time of an individual to a testing centre and possible waiting time for sample collection. However, for self-administered RA tests, the rate of testing can be at the maximum without any significant delay after the onset of symptoms. Thus, for RA testing, we considered:μ∗(a)=Ce−1a0≤a≤1μ(a)otherwise
where Ce −1 = max{μ(a)}. Using (2), the force of infection is expressed by(3) Λ(t)=δAA(t)+P(t)+δI∫0∞(i(t,a)+v(t,a)+w(t,a))da+δk∫0∞k(t,a)da.
where δ A, δ I, and δ k are the transmissibilities of asymptomatic, non-isolated symptomatic, and isolated symptomatic individuals relative to pre-symptomatic cases, respectively. Fig. 1 illustrates a schematic diagram of the model with testing and isolation.Fig. 1 Schematic representation of the model dynamics.
Fig. 1
To derive the equations for transmission dynamics of the model, it is biologically reasonable to assume that(4) i(t,∞)=v(t,∞)=k(t,∞)=w(t,∞)=0.
Solving (2) subject to the following boundary and initial conditions, respectively,(5) i(t,0)=γpP(t)i(0,a)=i0(a)v(t,0)=0v(0,a)=v0(a)k(t,0)=0k(0,a)=k0(a)w(t,0)=0w(0,a)=w0(a)
we obtainedi(t,a)=γPb1(a)P(t-a),v(t,a)=γPe-γaP(t-a)∫0aeγuμ(u)b1(u)du-e-γτ∫0aeγuμ(u-τ)b1(u-τ)du,k(t,a)=γPe-γ(a+τ)P(t-a)∫0aeγuq(u-τ)μ(u-τ)b1(u-τ)du,w(t,a)=γPe-γ(a+τ)P(t-a)∫0aeγu(1-q(u-τ))μ(u-τ)b1(u-τ)du,
whereb1(a)=e−∫0a(μ(u)+γ)du
Let(6) I(t)=∫0∞i(t,a)daV(t)=∫0∞v(t,a)daK(t)=∫0∞k(t,a)daW(t)=∫0∞w(t,a)da
be the number of individuals at time t who are not tested, tested, and waiting for the results, isolated with positive results, and non-isolated due to false-negative results, respectively. Therefore, we have the following differential equations:I'(t)=pσγP∫0∞b1(a)E(t-a)da-γPI(t),V'(t)=pσγP∫0∞e-γaE(t-a)(∫0aeγuμ(u)b1(u)du-e-γτ∫0aeγuμ(u-τ)b1(u-τ)du)da-γPV(t),K'(t)=pσγP∫0∞(e-γ(a+τ)E(t-a)∫0aeγuq(u-τ)μ(u-τ)b1(u-τ)du)da-γPK(t),W'(t)=pσγP∫0∞(e-γ(a+τ)E(t-a)∫0aeγu(1-q(u-τ))μ(u-τ)b1(u-τ)du)da-γPW(t).
The force of infection Λ(t) can be expressed byΛ(t)=δAA(t)+P(t)+δI(I(t)+V(t)+W(t))+δkK(t),t≥a.
Summarizing the above details, we arrived at the following model equations with both discrete and distributed delays:(7) S'(t)=-βS(t)Λ(t),E'(t)=βS(t)Λ(t)-σE(t),A'(t)=(1-p)σE(t)-γAA(t),P'(t)=pσE(t)-γPP(t),I'(t)=pσγP∫0∞b1(a)E(t-a)da-γPI(t),V'(t)=pσγP∫0∞e-γaE(t-a)(∫0aeγuμ(u)b1(u)du-e-γτ∫0aeγuμ(u-τ)b1(u-τ)du)da-γPV(t),K'(t)=pσγP∫0∞(e-γ(a+τ)E(t-a)∫0aeγuq(u-τ)μ(u-τ)b1(u-τ)du)da-γPK(t),W'(t)=pσγP∫0∞(e-γ(a+τ)E(t-a)∫0aeγu(1-q(u-τ))μ(u-τ)b1(u-τ)du)da-γPW(t).
3 Reproduction number
The basic reproduction number (R0) is a fundamental quantity defined as the expected number of secondary infections generated by a single infectious individual in an entirely susceptible population (Diekmann & Heesterbeek, 2000). According to the theory of epidemics, we would expect the disease to die out if R0<1, while it will cause a short-term outbreak or persist in the population if R0>1. A related quantity, the control reproduction number (Rc), can be defined when certain control measures (such as testing and isolation, treatment, or vaccination) are implemented, which is often used to determine whether the disease can be controlled.
To derive these quantities, we note that S(t), E(t), A(t), P(t), I(t), V(t), K(t) and W(t) are non-negative for all t > 0. The model has a family of disease equilibria: E0=(S(0),0,…,0) with any initial value of S(0) for the susceptible population. From (7), we see that S(t) is strictly decreasing in the presence of infection. Hence, the model has no endemic equilibrium point. Considering the duration and transmissibility of asymptomatic, non-isolated symptomatic, isolated symptomatic, and non-isolated false-negative cases, the total number of secondary cases generated in the presence of testing and isolation can be calculated using a previous method (Alexander et al., 2008), and is given byRc=βS(0)[(1-p)δAγA+pγP+pδIγ+p(δk-δI)∫0∞(e-γ(a+τ)∫0aeγuq(u-τ)μ(u-τ)b1(u-τ)du)da]
In the absence of testing and isolation (μ(a) ≡ 0), Rc reduces to the basic reproduction number(8) R0=βS(0)(1−p)δAγA+pγP+pδIγ.
4 Modified model
Unlike RT-PCR test results that are associated with some delay from the time of sample collection, there are several rapid antigen tests that provide results within minutes (Wells, Pandey, Moghadas, et al., 2022), allowing infected individuals with positive results to initiate isolation immediately. One possible way that may help reduce disease transmission is to initiate isolation at the time of RT-PCR testing while awaiting the test result. If the result is positive, then the individuals will continue their isolation until recovery. If the test result is negative (which may be false-negative based on the diagnostic sensitivity of the test), individuals can end their isolation and return to normal activities. In the context of our model, individuals who end their isolation due to a false-negative result can transmit the disease similar to infectious, non-isolated cases without restrictions. Similar to the model (7), we let i(t, a), v(t, a), k(t, a) and w(t, a) represent the population densities with respect to age a at time t that are not tested, tested and waiting for the results but isolated, isolated with positive result, and non-isolated due to a false negative result, respectively. DefiningI(t)=∫0∞i(t,a)da,V(t)=∫0∞v(t,a)da,K(t)=∫0∞k(t,a)da,W(t)=∫0∞w(t,a)da
and solving for densities i(t, a), v(t, a), k(t, a), and w(t, a), we obtained the same equations as (7) with the modified force of infection:(9) Λˆ(t)=δAA(t)+P(t)+δI(I(t)+W(t))+δk(V(t)+K(t)).
The control reproduction number for the modified model can then be expressed by(10) Rcˆ=βS(0)[(1-p)δAγA+pγP+pδIγ+p(δk-δI)∫0∞e-γa∫0aeγuμ(u)b1(u)duda-p(δk-δI)∫0∞(e-γ(a+τ)∫0aeγu(1-q(u-τ))μ(u-τ)b1(u-τ)du)da]
5 Parameterization and results
To illustrate the effect of timelines of testing and isolation with a delay in the test result, we simulated the model with estimates of parameters specific to SARS-CoV-2 infection. We used early estimates related to the original Wuhan-I strain, summarized in Table 1 . Assuming R0=1.5 and S(0) = 10, 000, the transmission rate β = 3.9501 × 10−5 was computed from equation (8).Table 1 Model parameters and their associated values.
Table 1Parameter Description Value Source
R0 Basic reproduction number 1.5 Assumed
β Transmission rate 3.9501 × 10−5 Computed from (8)
δA Relative transmissibility of asymptomatic infection 0.26 (Sayampanathan et al., 2021; Moghadas et al., 2020; Sayampanathan et al., 2021; Moghadas et al., 2020)
δI Relative transmissibility of non-isolated symptomatic infection 0.89 (Ferretti et al., 2020; Moghadas et al., 2020; Ferretti et al., 2020; Moghadas et al., 2020)
δk Reduction in transmissibility of isolated symptomatic infection 0.2 Assumed
1/σ Average latent period 2.2 days (Li et al., 2020; Moghadas et al., 2021; Li et al., 2020; Moghadas et al., 2021)
1/γA Average duration of asymptomatic stage 5 days (Gatto et al., 2020; Moghadas et al., 2020; Gatto et al., 2020; Moghadas et al., 2020)
1/γP Average duration of pre-symptomatic stage 2.3 days (Li et al., 2020; Moghadas et al., 2020; Li et al., 2020; Moghadas et al., 2020)
1/γ Average infectious period following the onset of symptoms 3.2 days (Li et al., 2020; Moghadas et al., 2020; Li et al., 2020; Moghadas et al., 2020)
p Proportion of infected individuals who develop symptoms 0.649 (Sah et al., 2021; Sah et al., 2021)
We set the minimum delay (a 0) in testing after the onset of symptoms to 0.1 day (i.e., 2.4 h). We then varied a 0 to a maximum of 2 days. We also varied the delay in obtaining the results of testing (τ) between 0.1667 days (i.e., 4 h) and 2 days when simulating the model for the RT-PCR test. For RA testing, given the short turnaround times, τ was set to 0.0104 (i.e., 15 min).
To determine the constant parameter C in the testing rate μ(a), we used the conversion formulae for probability of an event occurring given its rate. Assuming an 80% probability that a symptomatic individual is tested 1 day after the onset of symptoms, we calculated C by using the formulae 0.8=1−e−Ce−1 at a = 1, corresponding to the maximum value μ(1), giving C = 4.375.
To incorporate the test sensitivity, we fitted the function(11) q(t)=c1tc2e[−c3tc4+c5]
to discretized temporal diagnostic sensitivity of the RT-PCR derived from a previous study (Wells, Pandey, Moghadas, et al., 2022), giving the values c 1 = 1.722, c 2 = 5.625, c 3 = 5.127, c 4 = 0.484, and c 5 = 1.523. For the RA test, we considered the diagnostic sensitivity of Abbott-Panbio, which was expressed as the product of the diagnostic sensitivity of the RT-PCR and the temporal percent positive agreement between Abbott-Panbio and RT-PCR tests in serial testing data (Wells, Pandey, Moghadas, et al., 2022). Fitting (11) to Abbott-Panbio sensitivity estimates, we obtained c 1 = 3.393, c 2 = 7.869, c 3 = 9.963, c 4 = 0.4031, c 5 = 4.985 (Fig. 2 ). The sensitivity curves were used to determine the probability of an infected individual being identified as positive at the time of testing.Fig. 2 Fitting equation (11) to estimates of temporal diagnostic sensitivity of RT-PCT and Abbott-Panbio rapid antigen tests.
Fig. 2
5.1 Effect of delay in testing and test result on Rc
We evaluated the reduction in Rc achieved by testing and isolating infected individuals. For the RT-PCR test with isolation of positive cases starting at the time of test result, we found that R0 can be reduced by at least 23% (from 1.5 to below 1.15) if testing started within 16 h of symptom onset, and results are available within 10 h of sample collection (Fig. 3 A). As the delay in testing increases, a shorter turnaround time is required to achieve the same reduction in the reproduction number. In contrast, when isolation started at the time of sample collection and continued for those whose results are positive, turnaround time has little effect on changing the outcomes. For example, for testing within 16 h of symptom onset, R0 reduces to below 1.15 even if the test turnaround time is 2 days (Fig. 3B).Fig. 3 Control reproduction number as a function of delay in RT-PCR testing and delay in test results when isolation starts at: (A) the time of results for positive cases; (B) the time of testing.
Fig. 3
For the RA test, the delay in start of isolation is negligible due to the short (15-min) turnaround time. We found that the reduction in R0 can exceed 20% (from 1.5 to below 1.2) when symptomatic individuals performed a test within 16 h of symptom onset (Fig. 4 ). These results suggest that, despite higher sensitivity of the RT-PCR test, RA testing can be a viable alternative to costly and resource consuming RT-PCR testing.Fig. 4 Control reproduction number as a function of delay in RA testing with 15 min turnaround time.
Fig. 4
5.2 Effect of testing and isolation on incidence
We simulated the model for a 1-year time horizon to compare the incidence of infection under scenarios of RT-PCR and RA testing, when isolation starts at the time of test result. Fig. 5 A shows that the performance of RT-PCR testing with delay of 2.4 h after the onset of symptoms and turnaround time of 24 h is virtually equivalent to RA testing with delay of 24 h. Since self-administered RA tests do not involve logistical and practical constraints of the RT-PCR test, their early application could outperform RT-PCR testing with prolonged turnaround times in reducing the overall incidence (Fig. 5A).Fig. 5 Incidence of infection under different scenarios of delay in RT-PCR and RA testing with the start of isolation (A) at the time of results, and (B) at the time of testing.
Fig. 5
When isolation starts at the time of sample collection, the turnaround time of RT-PCR has a counterintuitive effect compared with RA testing (Fig. 5B). For example, RT-PCR testing with a 24-h delay and turnaround time of 24 h outperforms RA testing with a 24-h delay after the onset of symptoms. This arises from a higher rate of false negatives for the RA test, which allows infectious individuals to continue with their normal activities shortly (i.e., 15 min) after the test without isolation. In contrast, isolation of infectious individuals for a duration of 24 h before the RT-PCR test results are available would reduce the transmission (Fig. 5B). Furthermore, higher diagnostic sensitivity of the RT-PCR test compared with the RA test leads to fewer cases of false negatives. However, this comparative advantage diminishes with shorter delay in testing post symptom onset, and/or faster turnaround times for the RT-PCR test. For instance, reduction of incidence in RT-PCR testing with 2.4 h delay and turnaround time of 24 h is equivalent to that in RA testing with 2.4 h delay.
5.3 Effect of testing and isolation on the attack rate
The patterns observed in the control reproduction number, Rc, were also reflected in the attack rate (i.e., the proportion of the population infected throughout the epidemic). We computed the reduction in the attack rate achieved when isolation started at the time of testing as opposed to at the time of test result. This relative reduction (RR) was calculated byRR=Total infections (RT-PCR)TR−Total infections (RT-PCR)TTTotal infections (RT-PCR)TR,
where subscripts TR and TT represent the start of isolation at the time of result and testing, respectively. Fig. 6 shows that the RR reduces as delay in testing increases, but also increases as the turnaround time increases. The lowest RR (10%) occurs when tests are performed with a delay of at least 2 days post symptom onset and a short turnaround time of 4 h. The greatest RR (60%) is achieved when delay in testing is short (i.e., 2.4 h after the onset of symptoms), but the turnaround time is 2 days. This is expected as longer turnaround times extend the duration of isolation, which further reduces transmission in the scenario when isolation starts at the time of testing.Fig. 6 Relative reduction in the attack rate achieved with RT-PCR testing when isolation starts at the time of test result compared to at the time of testing.
Fig. 6
We then calculated the RR, comparing RT-PCR and RA testing. The RR was calculated byRR=Total infections (RA)-Total infections (RT-PCR)iTotal infections (RA),
where either i = TR or i = TT, with a fixed delay of 15 min for the RA test result. When isolation starts at the time of test result (Fig. 7 A), we found that the RT-PCR test outperforms the RA test (i.e., RR > 0) only if the test was performed within 20 h of symptom onset, and turnaround times were relatively short (i.e., within 4 h of testing). However, if isolation starts at the time of testing (Fig. 7B), the RT-PCR test (with a higher diagnostic sensitivity) outperforms the RA test regardless of delay in testing or turnaround times, but the RR reduces to under 10% as delay in testing increases to 2 days. Considering the practical timelines of RT-PCR testing post symptom onset and limited laboratory capacity that may lead to longer turnaround times during the peak of an epidemic, RA testing may lead to a similar or an even larger reduction in the attack rate.Fig. 7 Relative reduction in the attack rate achieved with RT-PCR testing compared with RA testing, as a function of delay in testing and turnaround time of the RT-PCR test result.
Fig. 7
6 Discussion
Testing and isolation of positive cases have been used as a key strategy to quell the spread of some infectious diseases, with COVID-19 and monkeypox presenting two recent examples of such diseases (MacIntyre, 2020; Titanji et al., 2022). There are several methods to test for the identification of infected individuals (CDC, 2022a). In the context of COVID-19 pandemic, the two commonly used methods are RT-PCR and rapid antigen tests, which have different temporal diagnostic sensitivities (Wells, Pandey, Moghadas, et al., 2022). The former has been shown to have a higher diagnostic sensitivity, but involves sample collection by trained individuals and laboratory assays, which would impose logistical considerations with implications for timely isolation of positive cases. The latter, on the other hand, can be self-administered with turnaround times within minutes, facilitating rapid case isolation despite having a comparatively lower diagnostic sensitivity. We considered these factors and developed a delay-differential model to investigate the trade-offs between higher sensitivity and rapid results of these tests with case isolation for mitigating disease spread.
Parameterizing the model with estimates associated with SARS-CoV-2 infection, and sensitivity of RT-PCR and Abbott-Panbio rapid antigen tests (Wells, Pandey, Moghadas, et al., 2022), we found that the reduction of disease transmission, reflected in Rc, using the RA test could be comparable to that achieved by the RT-PCR test. In practice, sample collection and laboratory results of the RT-PCR test are, in large measures, unlikely to occur within 1 day of symptom onset in the general population. If infected individuals are not isolated prior to the RT-PCR test result, our model indicates that using the RA test would lead to a significantly higher reduction in the attack rate. However, when isolation starts immediately after the test, then application of the RT-PCR test would confer a higher reduction in the attack rate, although this effect diminishes with delay in testing, thus arguing in favor of inexpensive, self-administered RA tests.
Our model provides a theoretical framework for evaluating the effect of testing and isolation. However, it is subject to several limitations. For example, we parameterized the model with a testing rate with some underlying assumptions that are plausible, but not based on specific quantifications in real-world settings. Furthermore, during an epidemic, it is likely that different methods of testing are used concomitantly with varying turnaround times. Our study considered testing of only symptomatic individuals, and after the onset of symptoms, while testing before the start of symptoms is also common during the epidemic, especially for close contacts and asymptomatic individuals (Day, 2020). We also assumed that those who are tested adhere to guidelines for isolation, for example in the scenario that isolation starts immediately after testing. However, with the perception that their symptoms may not indicate infection with the particular disease, they may defer isolation until confirmation of positive results. Under these assumptions, we highlight the qualitative aspects of our findings on the value of testing and timely isolation of infected cases.
Funding
SMM acknowledges the support from 10.13039/501100000038 Natural Sciences and Engineering Research Council of Canada (10.13039/501100000038 NSERC ) through Individual Discovery Grant, and EIDM Mathematics for Public Health (MfPH) grant.
Authors’ contributions
SMM conceived the study; AL and ZW developed the model, analyzed the results theoretically, and simulated the model; SMM analyzed simulation results and wrote the draft of the paper; AL and ZW contributed to providing contextual information, writing of the paper, and interpretation of the results.
Declaration of competing interest
None declared.
Appendix A Here, we show that if Rc<1, then the disease can be controlled. From (7), we see that S(t) ≤ S(0) for all t ≥ 0. Thus, we have(A.1) E′(t)≤βS(0)Λ(t)−σE(t).
The standard comparison argument implies that the solutions of (E, A, P, I, V, K, W) in the model are bounded by the solutions of the corresponding compartments of the following system:(A.2) E'(t)=βS(0)Λ(t)-σE(t),A'(t)=(1-p)σE(t)-γAA(t),P'(t)=pσE(t)-γPP(t),I'(t)=pσγP∫0∞b1(a)E(t-a)da-γPI(t),V'(t)=pσγP∫0∞e-γaE(t-a)∫0aeγuμ(u)b1(u)du-e-γτ∫0aeγuμ(u-τ)b1(u-τ)duda-γPV(t),K'(t)=pσγP∫0∞e-γ(a+τ)E(t-a)∫0aeγuq(u-τ)μ(u-τ)b1(u-τ)duda-γPK(t),W'(t)=pσγP∫0∞e-γ(a+τ)E(t-a)∫0aeγu(1-q(u-τ))μ(u-τ)b1(u-τ)duda-γPW(t).
We now substitute the ansatz ωe λt, where ω = (E 0, A 0, P 0, I 0, V 0, K 0, W 0) into the above equation. Without loss of generality, let us assume E 0 = 1 and eliminate e λt. It then follows that(A.3) λ=βS(0)δAA0+P0+δII0+δIV0+δkK0+δIW0-σ,A0λ=(1-p)σ-γAA0,P0λ=pσ-γPP0,I0λ=pσγP∫0∞b1(a)e-λada-γPI0,V0λ=pσγP∫0∞e-γa∫0aeγuμ(u)b1(u)du-e-γτ∫0aeγuμ(u-τ)b1(u-τ)due-λada-γPV0,K0λ=pσγP∫0∞e-γ(a+τ)∫0aeγuq(u-τ)μ(u-τ)b1(u-τ)due-λada-γPK0,W0λ=pσγP∫0∞e-γ(a+τ)∫0aeγu(1-q(u-τ))μ(u-τ)b1(u-τ)due-λada-γPW0.
Solving the equations for A 0, P 0, I 0, V 0, K 0 and W 0 givesA0=(1-p)σλ+γA,P0=pσλ+γP,I0=pσγP∫0∞b1(a)e-λadaλ+γP,V0=pσγP∫0∞e-γa∫0aeγuμ(u)b1(u)du-e-γτ∫0aeγuμ(u-τ)b1(u-τ)due-λadaλ+γP,K0=pσγP∫0∞e-γ(a+τ)∫0aeγuq(u-τ)μ(u-τ)b1(u-τ)due-λadaλ+γP,W0=pσγP∫0∞e-γ(a+τ)∫0aeγu(1-q(u-τ))μ(u-τ)b1(u-τ)due-λadaλ+γP
The characteristic equation can then be expressed by:(A.4) h(λ)=βS(0)σ[δA(1-p)λ+γA+pλ+γP+δIpγP(∫0∞b1(a)e-λadaλ+γP+∫0∞e-γa(∫0aeγuμ(u)b1(u)du-e-γτ∫0aeγuμ(u-τ)b1(u-τ)du)e-λadaλ+γP+∫0∞e-γ(a+τ)(∫0aeγu(1-q(u-τ))μ(u-τ)b1(u-τ)du)e-λadaλ+γP)+δkγP∫0∞e-γ(a+τ)(∫0aeγuq(u-τ)μ(u-τ)b1(u-τ)du)e-λadaλ+γP]-(λ+σ)
Let λ = x + iy be a complex root with x ≥ 0. Thus, |e −λa| ≤ 1 for all a ≥ 0. A simple calculation yields that1=|βS(0)σ(λ+σ)[δA(1-p)λ+γA+pλ+γP+δIpγP(∫0∞b1(a)e-λadaλ+γP+∫0∞e-γa(∫0aeγuμ(u)b1(u)du-e-γτ∫0aeγuμ(u-τ)b1(u-τ)du)e-λadaλ+γP+∫0∞e-γ(a+τ)(∫0aeγu(1-q(u-τ))μ(u-τ)b1(u-τ)du)e-λadaλ+γP)+δkpγP∫0∞e-γ(a+τ)(∫0aeγuq(u-τ)μ(u-τ)b1(u-τ)du)e-λadaλ+γP]|≤βS(0)σ|λ+σ|[δA(1-p)|λ+γA|+p|λ+γP|+δIpγP|λ+γP|(∫0∞b1(a)da+∫0∞e-γa(∫0aeγuμ(u)b1(u)du-e-γτ∫0aeγuμ(u-τ)b1(u-τ)du)da+∫0∞e-γ(a+τ)(∫0aeγu(1-q(u-τ))μ(u-τ)b1(u-τ)du)da)+δkpγP|λ+γP|∫0∞e-γ(a+τ)(∫0aeγuq(u-τ)μ(u-τ)b1(u-τ)du)da]<βS(0)[δA(1-p)γA+pγP+δIp(∫0∞b1(a)da+∫0∞e-γa(∫0aeγuμ(u)b1(u)du-e-γτ∫0aeγuμ(u-τ)b1(u-τ)du)da+∫0∞e-γ(a+τ)(∫0aeγu(1-q(u-τ))μ(u-τ)b1(u-τ)du)da)+δkp∫0∞e-γ(a+τ)(∫0aeγuq(u-τ)μ(u-τ)b1(u-τ)du)da]=Rc
Thus, if Rc<1, all complex roots of h(λ) have negative real parts. We now consider the real roots of h(λ). Let M=max−γP,−γA. It is easy to see that h′(λ) < 0 for all λ > M, indicating that h(λ) is monotonically decreasing on (M, ∞). Note that h(M) = ∞, h(∞) = −∞, and h(0)=σ(Rc−1). This implies that h(λ) has a unique negative real root if and only if Rc<1.
By considering the complex and real roots, we arrive at the conclusion that if Rc<1, all roots of the characteristic equation h(λ) have negative real parts. Hence, the zero solution of the model is asymptomatically stable.
Finally, we note that our system is a cooperative and irreducible, and thus a standard comparison argument shows that when Rc<1, the infection compartments are bounded by exponential functions with negative exponents for t ≥ a. Hence, the occurrence of an outbreak is prevented if Rc<1.
Peer review under responsibility of KeAi Communications Co., Ltd.
==== Refs
References
Alexander M.E. Moghadas S.M. Röst G. Wu J. A delay differential model for pandemic influenza with antiviral treatment Bulletin of Mathematical Biology 70 2008 382 397 17701376
CDC Centers for disease control and prevention, COVID-19 testing: What you need to know https://www.cdc.gov/coronavirus/2019-ncov/symptoms-testing/testing.html 2022
CDC Centers for disease control and prevention, isolation and precautions for people with COVID-19 https://www.cdc.gov/coronavirus/2019-ncov/your-health/isolation.html 2022
Day M. Covid-19: Identifying and isolating asymptomatic people helped eliminate virus in Italian village BMJ British Medical Journal 368 2020
Diekmann O. Heesterbeek J.A.P. Mathematical epidemiology of infectious diseases: Model building, analysis and interpretation Vol. 5 2000 John Wiley & Sons
Ferretti L. Wymant C. Kendall M. Zhao L. Nurtay A. Abeler-Dörner L. Parker M. Bonsall D. Fraser C. Quantifying sars-cov-2 transmission suggests epidemic control with digital contact tracing Science 368 2020 eabb6936
Gatto M. Bertuzzo E. Mari L. Miccoli S. Carraro L. Casagrandi R. Rinaldo A. Spread and dynamics of the covid-19 epidemic in Italy: Effects of emergency containment measures Proceedings of the National Academy of Sciences 117 2020 10484 10491
Grassly N.C. Pons-Salort M. Parker E.P. White P.J. Ferguson N.M. Ainslie K. Baguelin M. Bhatt S. Boonyasiri A. Brazeau N. Comparison of molecular testing strategies for covid-19 control: A mathematical modelling study The Lancet Infectious Diseases 20 2020 1381 1389 32822577
Kucharski A.J. Klepac P. Conlan A.J. Kissler S.M. Tang M.L. Fry H. Gog J.R. Edmunds W.J. Emery J.C. Medley G. Effectiveness of isolation, testing, contact tracing, and physical distancing on reducing transmission of sars-cov-2 in different settings: A mathematical modelling study The Lancet Infectious Diseases 20 2020 1151 1160 32559451
Li R. Pei S. Chen B. Song Y. Zhang T. Yang W. Shaman J. Substantial undocumented infection facilitates the rapid dissemination of novel coronavirus (sars-cov-2) Science 368 2020 489 493 32179701
MacIntyre C.R. Case isolation, contact tracing, and physical distancing are pillars of covid-19 pandemic control, not optional choices The Lancet Infectious Diseases 20 2020 1105 1106 32559453
Mercer T.R. Salit M. Testing at scale during the covid-19 pandemic Nature Reviews Genetics 22 2021 415 426
Moghadas S.M. Fitzpatrick M.C. Sah P. Pandey A. Shoukat A. Singer B.H. Galvani A.P. The implications of silent transmission for the control of covid-19 outbreaks Proceedings of the National Academy of Sciences 117 2020 17513 17515
Moghadas S.M. Fitzpatrick M.C. Shoukat A. Zhang K. Galvani A.P. Simulated identification of silent covid-19 infections among children and estimated future infection rates with vaccination JAMA Network Open 4 2021 e217097–e217097
Morens D.M. Daszak P. Markel H. Taubenberger J.K. Pandemic covid-19 joins history's pandemic legion mBio 11 2020 e00812 e00820 32471830
Rosenberg E.S. Holtgrave D.R. Widespread and frequent testing is essential to controlling coronavirus disease 2019 (covid-19) in the United States 2021
Sah P. Fitzpatrick M.C. Zimmer C.F. Abdollahi E. Juden-Kelly L. Moghadas S.M. Singer B.H. Galvani A.P. Asymptomatic sars-cov-2 infection: A systematic review and meta-analysis Proceedings of the national academy of Sciences Vol. 118 2021 e2109229118
Sayampanathan A.A. Heng C.S. Pin P.H. Pang J. Leong T.Y. Lee V.J. Infectivity of asymptomatic versus symptomatic covid-19 The Lancet 397 2021 93 94
Titanji B.K. Tegomoh B. Nematollahi S. Konomos M. Kulkarni P.A. Monkeypox: A contemporary review for healthcare professionals Open forum infectious diseases 2022 Oxford University Press ofac310
Wells C.R. Pandey A. Gokcebel S. Krieger G. Donoghue A.M. Singer B.H. Moghadas S.M. Galvani A.P. Townsend J.P. Quarantine and serial testing for variants of sars-cov-2 with benefits of vaccination and boosting on consequent control of covid-19 PNAS nexus 1 2022 pgac100 35909795
Wells C.R. Pandey A. Moghadas S.M. Singer B.H. Krieger G. Heron R.J. Turner D.E. Abshire J.P. Phillips K.M. Michael Donoghue A. Comparative analyses of eighteen rapid antigen tests and rt-pcr for covid-19 quarantine and surveillance-based isolation Communications medicine 2 2022 1 12 35603280
Wells C.R. Townsend J.P. Pandey A. Moghadas S.M. Krieger G. Singer B. McDonald R.H. Fitzpatrick M.C. Galvani A.P. Optimal covid-19 quarantine and testing strategies Nature Communications 12 2021 1 9
| 36467718 | PMC9704027 | NO-CC CODE | 2022-12-16 23:18:16 | no | Infect Dis Model. 2023 Mar 28; 8(1):58-71 | utf-8 | Infect Dis Model | 2,022 | 10.1016/j.idm.2022.11.008 | oa_other |
==== Front
J Emerg Nurs
J Emerg Nurs
Journal of Emergency Nursing
0099-1767
1527-2966
Emergency Nurses Association. Published by Elsevier Inc.
S0099-1767(22)00062-9
10.1016/j.jen.2022.03.006
Research
Examination of the Effects of 4-Hour Nonvalved Filtering Facepiece Respirator Use on Blood Gas Values of Health Care Professionals: A Before and After Study
Pasli Sinan MD ∗
Imamoglu Melih MD
Beser Muhammet Fatih MD
Sahin Abdul Samet MD
Ilhan Engin MD
Yadigaroglu Metin MD
∗ For correspondence, write: Sinan Pasli, Department of Emergency Medicine, Karadeniz Technical University, Trabzon 61080, Turkey.
9 5 2022
7 2022
9 5 2022
48 4 423429.e1
© 2022 Emergency Nurses Association. Published by Elsevier Inc. All rights reserved.
2022
Emergency Nurses Association
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Introduction
The use of personal protective equipment increased rapidly during the COVID-19 pandemic that began in 2019. The purpose of this study was to examine the effects of uninterrupted 4-hour use of internationally certified nonvalved filtering facepiece respirators on venous blood gas in health care workers during the COVID-19 pandemic.
Methods
A before-after design included venous blood gas analyses collected at the beginning of shifts before nonvalved filtering facepiece respirator had been put on and after 4-hour uninterrupted use of nonvalved filtering facepiece respirator.
Results
In this study, 33 volunteer health care workers took part. In terms of blood gas values, mean pCO2 values were 47.63 (SD = 5.16) before and 47.01 (SD = 5.07) after nonvalved filtering facepiece respirator use, mean HCO3 values were 23.68 (SD = 1.10) in first blood gas analysis and 24.06 (SD = 1.31) in second blood gas analysis, and no significant difference was observed between before and after the use of nonvalved filtering facepiece respirator (t = 0.67, P = .50, t = −2.0, P = .054, respectively). The only significant difference in parameters investigated between the groups was in pH levels, at pH = 7.35 (SD = 0.29) before and pH = 7.36 (SD = 0.20) after nonvalved filtering facepiece respirator use (t = −2.26, P = .03).
Conclusion
Continuous nonvalved filtering facepiece respirator use for 4 hours was not associated with clinician impairment in blood gas and peripheral SpO2 levels during nonexertional clinical ED work.
Graphical abstract
Key words
FFP2
N95 mask
Respirator
Venous blood analysis
==== Body
pmc Contribution to Emergency Nursing Practice
• Nonvalved filtering facepiece respirators are frequently used by nurses, paramedics, and other health care personnel who intervene in suspected and diagnosed patients with COVID-19, especially during the pandemic with high respiratory transmission.
• The main finding of this article is that continuous nonvalved filtering facepiece respirator use for 4 hours was not associated with impairment in blood gas and peripheral SpO2 levels during nonexertional clinical work in the emergency department. A statistically significant increase was observed in pH values, which remained within physiological limits and may not have clinical significance.
• Recommendations for translating the findings of this paper into emergency clinical practice include the following: continuous 4-hour use of nonvalved filtering facepiece respirators does not appear to have a negative impact on clinician respiratory physiology.
Introduction
Diseases transmitted by respiration can cause epidemics and pandemics. Health care workers adopted a series of measures to protect themselves during the severe acute respiratory syndrome (SARS) epidemic caused by the SARS coronavirus originating in Hong Kong in 2002 and during the influenza A H1N1 swine flu epidemic originating in Mexico in 2009. The most important of these measures was the use of surgical masks and nonvalved filtering facepiece (N95/FFP2) respirators.1 The use of personal protective equipment (PPE) increased rapidly during the COVID-19 pandemic that began spreading across the world from the Chinese city of Wuhan in 2019. Studies have also emphasized the importance of PPE use.2 Studies performed during the SARS outbreak have also revealed significant findings regarding the protective nature of PPE.3
Health care professionals serving patients with probable and confirmed COVID-19 were advised to use respirators (N95, FFP2, or equivalent standard), especially for aerosol-generating procedures.4 It may be reasonable to consider European FFP2 as “equivalent” to US NIOSH N95 respirators, for filtering at least 94% of non–oil-based particles such as virus bio-aerosols.5
We hypothesized that the pCO2 value in the blood gas might increase, and the pH and SpO2 values may decrease after the continuous use of N95/FFP2 respirators. The purpose of this study was to examine the effects of uninterrupted 4-hour N95/FFP2 respirator use on venous blood gas in health care professionals.
Methods
Study Design and Participants
We used a before-after design. The study population consisted of health care professionals (emergency nurses and physicians) working in the emergency department of a tertiary hospital between March 1, 2021, and April 1, 2021. Inclusion criteria were age 18 or over, voluntary participation with written consent, being clean-shaven, and receiving training in the use of PPE. Exclusion criteria were the presence of severe chronic pulmonary disease or mask use being medically contraindicated and participants who removed their N95/FFP2 respirator for any reason during the study period. No incentive was provided.
Ethical Considerations
This study was conducted under the approval of the local ethics committee (approval number: 2021/12).
Study Procedures
Venous blood gas measurements were performed at the beginning of shifts before the N95/FFP2 respirator had been put on and after 4-hour uninterrupted use of internationally certified N95/FFP2 (3M VFlex 9152E) respirators. Participants continued with their routine activities for 4 hours. They did not engage in high-effort interventions such as providing cardiopulmonary resuscitation and did not leave the emergency department for any reason. An average of 2 volunteers participated in the study each weekday, and the participants were observed by an author working in the same field.
Measures
Demographic Information
Age, sex, smoking, and medical history were recorded on a research form for each participant (see Supplementary Appendix).
Questionnaire for Side Effects
Major symptoms such as headache, nausea, palpitations, shortness of breath, or anxiety were questioned. At the end of 4-hour respirator use, participants were verbally asked whether they had any of these symptoms during the 4-hour period, and the results were recorded on a research form (see Supplementary Appendix).
Blood Gas Analysis
Blood gas analyses were performed on a bench analyzer (Rapidlab 1265, Bayer Health Care LLC, Pittsburgh, PA). The device was calibrated at regular intervals (1-point calibration every 4 hours, 2-point calibration every 8 hours). The pH (reference range: 7.35-7.45), pCO2 (reference range: 35-45 mmHg), and standard-HCO3 (reference range: 21.2-27 mmol/L) levels were measured. The results were printed on a paper and stored.
Statistical Methods
The study data were recorded onto Microsoft Excel software (Microsoft Corporation, Redmond, WA) and analyzed using Statistical Package of Social Sciences version 24.0 (IBM Corp, Armonk, NY) and MedCalc software (MedCalc Software Ltd, Oostende, Belgium). Compatibility with normal distribution was evaluated using the Shapiro Wilk or Kolmogorov Smirnov tests. Normally distributed numerical variables were expressed as mean and standard deviation and 95% confidence interval. Non-normally distributed numerical variables were expressed as median (minimum-maximum) and 95% confidence interval. Categorical variables were defined as n (number) and %. The matched pairs t test was used to compare before and after values of normally distributed numerical variables. Wilcoxon’s test was employed to compare non-normally distributed numerical variables. The independent samples t test was used to compare normally distributed numerical demographic variables between independent groups, and numerical data not exhibiting normal distribution were compared using the Mann-Whitney U test. P < .05 were regarded as statistically significant for all analyses. When the effect size was expected as d = 0.5, alpha error = 0.05, statistical power = 0.85 for the t test in 2 dependent groups (matched pairs) using the GPower 3.1.9.7. program (Heinrich-Heine-University, Düsseldorf, Germany), we calculated that the sample size of the study should be at least 31.
Results
Thirty-three volunteer health care professionals (emergency nurses and physicians) participated in the study. No participants were excluded from the study, and no data were missing. Men constituted n = 17 (51.5%) of the participants and women n = 16 (48.5%). Median ages were 28 years (24-47) among men and 27 years (24-29) among women. Nonsmokers (tobacco) were n = 26 (78.8%) while n = 7 (21.2%) smoked tobacco. Intermittent asthma was present in 2 participants (Table 1 ). Median fingertip oxygen saturation values in room air were SpO2 = 98 (95-99) before N95/FFP2 use and SpO2 = 98 (92-100) after N95/FFP2 use, and the difference was not statistically significant (z = −1.48, P = .13). Mean pCO2 values were 47.63 (SD = 5.16) before N95/FFP2 use and 47.01 (SD = 5.07) after N95/FFP2 use, and no significant difference was observed between these groups (t = 0.67, P = .50). The mean HCO3 levels were 23.68 (SD = 1.10) and 24.06 (SD = 1.3) before and after N95/FFP2 use, respectively, and no significant difference was observed (t = −2.0, P = .054). The only significant difference in parameters investigated between the groups was at pH levels. Mean pH levels were found at 7.35 (SD = 0.29) before N95/FFP2 use and pH = 7.36 (SD = 0.20) after N95/FFP2 use (t = −2.26, P = .03) (Table 2 ). No major side effects (headache, nausea, palpitations, dyspnea, or anxiety) were reported by the participants.Table 1 Participants’ demographic characteristics (N = 33)
Sex n %
Male 17 51.5
Female 16 48.5
Smoking
Yes 7 21.2
No 26 78.8
Past medical history
Asthma 2 6.1
Age (y) Median Min-Max t∗ P
Male 28 24-47 1.93 .06
Female 27 24-29
∗ Independent sample t test.
Table 2 A comparison of prerespirator and postrespirator use parameters
Variable Median Min-Max z∗ P
SpO2 before 98 95-99 −1.48 .13
SpO2 after 98 92-100
x¯ SD t† P
pCO2 before 47.63 5.16 0,67 .50
pCO2 after 47.01 5.07
pH before 7.35 0.29 −2,26 .03
pH after 7.36 0.20
HCO3 before 23.68 1.10 −2.0 .054
HCO3 after 24.06 1.31
∗ Wilcoxon test.
† Paired sample t test.
Discussion
The new global threat SARS-CoV-2 is transmitted by droplets and aerosols. Health care workers were advised to use PPE to prevent transmission of the disease during the COVID-19 pandemic.4 One recent meta-analysis showed that surgical masks provided a degree of protection comparable to that of N95/FFP2 respirators for aerosol-free procedures.6 However, in aerosol-generating procedures, the use of N95, FFP2, or FFP3 respirators may be appropriate.4
Although their findings vary, several studies have investigated the effects of N95/FFP2 and FFP3 respirators on metabolic and respiratory parameters. The present study examined the effect of 4-hour continuous N95/FFP2 respirator use on venous blood gas and peripheral oxygen saturation values. While a significant difference was observed in pH values, there was no significant difference between PCO2, HCO3, and peripheral SpO2 values. Despite the significant difference in pH values, both groups remained within physiological limits. Although the reason could not be determined through measurements, this might be related to increased respiration rates. A study showed that the use of N95/FFP2 for 1 hour was associated with increases in respiratory rate (range, 1.4-2.4 breaths per minute).7 Kao et al8 performed a blood gas study on 39 patients receiving dialysis during the SARS epidemic. They reported that 4-hour use of an N95/FFP2 respirator significantly reduced PaO2 in patients with end-stage kidney disease and increased respiratory side effects in those patients. These findings need to be confirmed in studies with a contemporaneous comparison or control group to address the unmeasured confounding of the work shift itself.
Ong et al9 and Lim et al10 both reported that N95/FFP2 caused a significant increase in headaches among health care workers. In a study performed during the COVID-19 pandemic, Ong et al9 reported headaches associated with PPE use in 82% of health care workers. In another study carried out during the SARS epidemic, Lim et al10 reported that 37.3% of participants had a headache after N95/FFP2 use. These authors interpreted that this headache finding might be related to increased inhaled CO2 levels, but blood gas measurements were not performed in their study. There was no difference between pCO2 levels before and after N95/FFP2 respirator use in our study, and the participants did not report any side effects. Coca et al11 recommended establishment of appropriate working and rest periods to avoid undesirable side effects of PPE. When using respirators for extended periods of time during the COVID-19 pandemic, health care professionals may have developed physiological and behavioral adaptations or ignored side effects with extended PPE wear. Changes in nutrition, hydration, and resting habits, and adjustment of work tempo may be some of these adaptation mechanisms. Further studies targeting pulmonary function tests or behavioral changes can be beneficial to clarify possible clinician adaptations.
Bharatendu et al12 examined the effect of N95/FFP2 use on end-tidal CO2 (ETCO2), and no significant increase in ETCO2 values was observed in participants using N95/FFP2. A recent study found that the use of N95/FFP2 for 1 hour in 10 healthy emergency residents did not cause a significant difference in the pH and pCO2 values measured at 20, 40, and 60 minutes.13 A study conducted with 57 nurses and 47 paramedics working in the SARS coronavirus-2 intensive care unit showed that the use of FFP2 and FFP3 respirators did not cause deterioration in blood gas values in a median time of 240 minutes.14 The results of these 2 studies are similar to our findings.
Yalciner et al15 reported that using an FFP3 respirator for at least 4 hours caused no significant change in blood gas parameters in 15 health care workers. None of the participants removed the respirator for any reason during the study period, and this is interpreted as it being well tolerated by participants.14 The methodology and results of our study are similar to those of Yalciner et al,15 although we used N95/FFP2 respirators in our research. In their study of 43 health care workers, Nafisah et al16 reported that continuous N95/FFP2 use significantly reduced pO2 levels and increased pCO2. The results of our study are not consistent with this research, suggesting that respiratory parameters can be affected differently by the use of different brands of respirators. In a study with the participation of 154 health care professionals, the effect of using only the N95 and combined use of the N95+Powered Air Purifying Respirator (PAPR) on cerebral hemodynamics and blood gas parameters was examined. They found a significant increase in ETCO2 values after using only the N95 respirator for 5 minutes. After the combination of N95+PAPR at 5 minutes, the ETCO2 values measured at 10 minutes returned to their basal values. These N95+PAPR results are not corroborated by our study. The reasons for this difference may be that clinician adaptations to extended PPE wear had not yet developed in the early pandemic period, or it may be related to the measurement method (capnometer). In addition, it is not clear whether the ETCO2 values returning to basal levels in that study is due to the combination of N95+PAPR or due to the adaptation that may occur over time. We did not perform a respirator fit test for the participants; therefore, air leaking from around the respirator may have affected the results and differences in our findings from those of other studies.12
Currently, the evidence in the published literature does not provide consistent results, especially regarding changes in pCO2 levels.13, 16 Most studies have shown that respirator use does not lead to significant changes in pCO2 levels.12, 13, 14 The fact that pCO2 levels were not impaired in our current study corroborates evidence in the published literature. However, our study contradicts other evidence in the published literature regarding side effects related to mask use.9, 10 Headache associated with respirator use has been reported frequently in the published literature, but in our study, the participants did not report any side effects. Respirator type, duration of use, age groups, sample size, or different working conditions may be the reason for different results. We recommend that determining the duration of PPE use, respirator change frequency, and respirator reuse be aligned with current World Health Organization recommendations, health care administrator recommendations, and manufacturer recommendations.4
Limitations
The relatively low number of participants may be regarded as a limitation of this study, and the average age of study participants was young, which may not be generalizable to the typical health care setting. In addition, the results were based on venous blood gas tests, and arterial blood gas samples might have yielded more accurate results. Another limitation may be that we were unable to show the effects of respirator use in participants with chronic disease or with exertion during the clinical shift. A history of intermittent asthma was present in only 2 members of the study population, and no statistical comparison was possible. As another limitation the pre-4-hour and post-4-hour measurements reflect values only at that immediate time and do not give an indication of short-term excursions that may have occurred during the 4-hour period. We did not perform a respirator fit test for the participants before the study; therefore, we cannot comment on whether there was air leakage around the mask that affected our results. The N95/FFP2 respirator was fixed to the dorsum of the nose with a tape to support a mask seal to the face.
Implications for Emergency Nurses
In light of evidence from the previously published literature with the findings of our study, using N95 respirators continuously for 4 hours by nurses and physicians working in nonexertional tasks in the emergency department who do not have pulmonary disease and have no contraindications for mask use is not associated with major side effects or blood gas deterioration. Our findings should be interpreted with caution and may not apply to older workers or those with characteristics markedly different from our study participants. We also think that it would be appropriate for other health care personnel working under similar conditions to use respirators for the same period of time.
Since the COVID 19 pandemic began, the use of PPE has been of the most significant importance for health care professionals’ self-protection. The importance of using proper PPE cannot be underestimated. It is important to consult with the respirator manufacturer regarding the maximum number of uses they recommend for the N95/FFP2 respirator. If no manufacturer guidance is available, data suggest limiting the number of reuses to no more than 5 total uses per device to ensure an adequate respirator performance.17 In a survey of 27 countries overall, 17 countries (63%) provide no information on their websites about the long-term use or reuse of N95/FFP2 respirators. Some countries have proposed specific methods for decontamination of N95/FFP2 respirators, and some countries have left the decision to health care administrators. The maximum extended use time ranged from 4 hours to 40 hours.18 World Health Organization recommends that wearing a respirator longer than 4 hours can cause discomfort and should be avoided.4
Conclusion
Contrary to our initial concerns, continuous N95/FFP2 respirator use for 4 hours was not associated with any impairment in blood gas and peripheral SpO2 levels. At the same time, a statistically significant increase was observed in pH values, although these remained within physiological limits and may not be clinically significant. In light of the evidence in the published literature and the results of our study, we conclude that N95/FFP2 respirators can be used safely for 4 hours without interruption for nonexertional clinical tasks in the emergency department. More research is needed on the impact of extended PPE wear during exertional activities, such as chest compressions in cardiopulmonary resuscitation.
Data, Code, and Research Materials Availability
Patient consent statement: written consent of the participants was obtained.
Author Disclosures
Conflicts of interest: none to report.
Sinan Pasli is an Assistant Professor, Faculty of Medicine, Department of Emergency Medicine, Karadeniz Technical University, Trabzon, Turkey. ORCID identifier:https://orcid.org/0000-0003-0052-2258.
Melih Imamoglu is an Assistant Professor, Faculty of Medicine, Department of Emergency Medicine, Karadeniz Technical University, Trabzon, Turkey. Twitter:@melihimam. ORCID identifier:https://orcid.org/0000-0003-4197-8999.
Muhammet Fatih Beser is a Resident, Faculty of Medicine, Department of Emergency Medicine, Karadeniz Technical University, Trabzon, Turkey. Twitter:@drfatih. ORCID identifier:https://orcid.org/0000-0003-2638-1297.
Abdul Samet Sahin is a Resident, Faculty of Medicine, Department of Emergency Medicine, Karadeniz Technical University, Trabzon, Turkey. ORCID identifier:https://orcid.org/0000-0001-9512-8741.
Engin Ilhan is a Resident, Faculty of Medicine, Department of Emergency Medicine, Karadeniz Technical University, Trabzon, Turkey. ORCID identifier:https://orcid.org/0000-0001-5476-9630.
Metin Yadigaroglu is an Assistant Professor, Faculty of Medicine, Department of Emergency Medicine, Samsun University, Samsun, Turkey. Twitter:@mtnydgrdr. ORCID identifier:https://orcid.org/0000-0003-1771-5523.
Supplementary Appendix
Form 1. Examination of the effects of 4-hour n95/FFP2 respirator use on blood gas values of health care professionals
DATA COLLECTION FORM
Age:
Gender:
Chronic Disease:
Smoking (tobacco):
Occupation:
Fingertip Sp0 2: Before N95/FFP2 use …………….. After N95/FFP use …………….
SIDE EFFECTS OCCURED DURING RESPIRATOR USE □ Headache
□ Nausea
□ Palpitations
□ Shortness of breath
□ Anxiety
□ Other (……………)
Supplementary data related to this article can be found at https://doi.org/10.1016/j.jen.2022.03.006. 1ST BLOOD GAS SAMPLE (Before N95/FFP2 use) 2ND BLOOD GAS SAMPLE (After N95/FFP2 use)
BARCODE NUMBER
==== Refs
References
1 Loeb M. Dafoe N. Mahony J. Surgical mask vs N95 respirator for preventing influenza among health care workers: a randomized trial JAMA 302 17 2009 1865 1871 10.1001/jama.2009.1466 19797474
2 Gamage B. Moore D. Copes R. Yassi A. Bryce E. Protecting health care workers from SARS and other respiratory pathogens: a review of the infection control literature Am J Infect Control 33 2 2005 114 121 10.1016/j.ajic.2004.12.002 15761412
3 Jefferson T. Foxlee R. Del Mar C. Physical interventions to interrupt or reduce the spread of respiratory viruses: systematic review BMJ 336 7635 2008 77 80 10.1136/bmj.39393.510347.be 18042961
4 World Health Organization. Rational use of personal protective equipment for coronavirus disease (COVID-19): interim Guidance, 27 February 2020 Accessed April 4, 2022 https://apps.who.int/iris/handle/10665/331215
5 Klimek L. Huppertz T. Alali A. A new form of irritant rhinitis to filtering facepiece particle (FFP) masks (FFP2/N95/KN95 respirators) during COVID-19 pandemic World Allergy Organ J 13 10 2020 100474 10.1016/j.waojou.2020.100474 33042359
6 Bartoszko J.J. Farooqi M.A.M. Alhazzani W. Loeb M. Medical masks vs N95 respirators for preventing COVID-19 in healthcare workers: a systematic review and meta-analysis of randomized trials Influenza Other Respir Viruses 14 4 2020 365 373 10.1111/irv.12745 32246890
7 Kim J.H. Benson S.M. Roberge R.J. Pulmonary and heart rate responses to wearing N95 filtering facepiece respirators Am J Infect Control 41 1 2013 24 27 10.1016/j.ajic.2012.02.037 22944510
8 Kao T.W. Huang K.C. Huang Y.L. Tsai T.J. Hsieh B.S. Wu M.S. The physiological impact of wearing an N95 mask during hemodialysis as a precaution against SARS in patients with end-stage renal disease J Formos Med Assoc 103 8 2004 624 628 15340662
9 Ong J.J.Y. Bharatendu C. Goh Y. Headaches associated with personal protective equipment - a cross-sectional study among frontline healthcare workers during COVID-19 Headache 60 5 2020 864 877 10.1111/head.13811 32232837
10 Lim E.C. Seet R.C. Lee K.H. Wilder-Smith E.P. Chuah B.Y. Ong B.K. Headaches and the N95 face-mask amongst healthcare providers Acta Neurol Scand 113 3 2006 199 202 10.1111/j.1600-0404.2005.00560.x 16441251
11 Coca A. Quinn T. Kim J.H. Physiological evaluation of personal protective ensembles recommended for use in West Africa Disaster Med Public Health Prep 11 5 2017 580 586 10.1017/dmp.2017.13 28303774
12 Bharatendu C. Ong J.J. Goh Y. Powered Air Purifying Respirator (PAPR) restores the N95 face mask induced cerebral hemodynamic alterations among healthcare workers during COVID-19 outbreak J Neurol Sci 417 2020 117078 10.1016/j.jns.2020.117078 32768718
13 Radio J. Willet K. Buyan L. 34 prolonged N-95 mask use did not result in carbon dioxide retention or clinically significant pH changes in one cohort of health care workers Ann Emerg Med 78 4 2021 S15 10.1016/j.annemergmed.2021.09.042
14 Mędrzycka-Dąbrowska W. Ślęzak D. Robakowska M. Evaluation of capillary blood gases in medical personnel caring for patients isolated due to SARS-CoV-2 in intensive care units before and after using enhanced filtration masks: a prospective cohort study Int J Environ Res Public Health 18 18 2021 9425 10.3390/ijerph18189425 34574350
15 Yalciner G. Babademez M.A. Gul F. Serifler S. Bulut K.S. Ozturk L. Consequences of FFP3 mask usage on venous blood gases Ir J Med Sci 190 4 2021 1565 1569 10.1007/s11845-020-02474-2 33459943
16 Nafisah S.B. Susi A. Alsaif E. Alqasmi M. Mzahim B. The effect of wearing an N95 mask on the blood gas values of healthcare providers Int J Med Sci Public Health 10 2 2021 31 34
17 Bergman M.S. Viscusi D.J. Zhuang Z. Palmiero A.J. Powell J.B. Shaffer R.E. Impact of multiple consecutive donnings on filtering facepiece respirator fit Am J Infect Control 40 4 2012 375 380 10.1016/j.ajic.2011.05.003 21864945
18 Kobayashi L.M. Marins B.R. dos Santos Costa P.C. Perazzo H. Castro R. Extended use or reuse of N95 respirators during COVID-19 pandemic: an overview of national regulatory authority recommendations Infect Control Hosp Epidemiol 41 11 2020 1364 1366 10.1017/ice.2020.173 32319884
| 35550305 | PMC9704115 | NO-CC CODE | 2022-12-10 23:20:28 | no | J Emerg Nurs. 2022 Jul 9; 48(4):423-429.e1 | utf-8 | J Emerg Nurs | 2,022 | 10.1016/j.jen.2022.03.006 | oa_other |
==== Front
Revista Médica Clínica Las Condes
0716-8640
0716-8640
S0716-8640(22)00120-1
10.1016/j.rmclc.2022.10.002
Article
Procesamiento de lenguaje natural para texto clínico en español: el caso de las listas de espera en Chile
Natural language processing for clinical text in Spanish: The case of waiting lists in ChileBáez Pablo a
Arancibia Antonia Paz b
Chaparro Matías Ignacio b
Bucarey Tomás b
Núñez Fredy cd
Dunstan Jocelyn def⁎
a Centro de Informática Médica y Telemedicina, Facultad de Medicina, Universidad de Chile. Santiago, Chile
b Escuela de Medicina, Facultad de Medicina, Universidad de Chile. Santiago, Chile
c Departamento de Ciencias del Lenguaje, Facultad de Letras, Pontificia Universidad Católica de Chile. Santiago, Chile
d Centro de Modelamiento Matemático, Facultad de Ciencias Físicas y Matemáticas, Universidad de Chile. Santiago, Chile
e Iniciativa de Datos & Inteligencia Artificial, Facultad de Ciencias Físicas y Matemáticas, Universidad de Chile. Santiago, Chile
f Instituto Milenio Engineering for Healthcare, ANID, Chile. Santiago, Chile
⁎ Autor para correspondencia.
28 11 2022
November-December 2022
28 11 2022
33 6 576582
30 7 2022
16 10 2022
18 10 2022
.
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Las listas de espera no cubiertas por el Plan de Garantías Explícitas en Salud para nueva consulta de especialidad en Chile se han visto incrementadas por los efectos de la pandemia del coronavirus SARS-CoV-2 (COVID-19). Esto representa un problema debido a la demora en la resolución y priorización de cada caso de derivación al nivel secundario de atención en salud. El objetivo de este artículo es exponer el problema de la lista de espera en el sistema de salud de Chile, y abordarlo como ejemplo de la aplicación de técnicas de Procesamiento del Lenguaje Natural (PLN). Específicamente, se describe una metodología para el reconocimiento de información clave en narrativas médicas. Actualmente, contamos con un conjunto de interconsultas médicas manualmente anotadas en el desarrollo del Corpus de Lista de Espera Chilena, y con una fracción de 2.000 interconsultas en las que las entidades médicas anotadas fueron normalizadas de forma automatizada a los conceptos del Sistema de Lenguaje Médico Unificado empleando el léxico MedLexSp. Este y otros recursos lingüísticos y herramientas de PLN están siendo desarrollados por el grupo de PLN en Medicina del Centro de Modelamiento Matemático de la Universidad de Chile y otros grupos a nivel nacional, los cuales constituyen aportes relevantes que pueden ser transferidos al sistema de salud chileno, con el objetivo de apoyar la gestión del texto clínico en español.
The waiting lists not covered by the Explicit Health Guarantee Plan for new specialty consultation in Chile increased due to the effects of the SARS-CoV-2 coronavirus (COVID-19) pandemic. This represents a problem derived from the delay in the resolution and prioritization of each case. This paper aims to describe the issue of the waiting lists in the Chilean health system and present an example of the application of Natural Language Processing (NLP). Specifically, a methodology for recognizing key information in medical narratives is described. Currently, we have a set of manually annotated medical referrals in the development of the Chilean Waiting List Corpus, with a fraction of 2,000 referrals in which the annotated medical entities were automatically normalized to the Unified Medical Language System concepts using the lexicon MedLexSp. The clinical NLP Group of the Center for Mathematical Modeling of the University of Chile, and other national NLP groups, are developing several tools and resources in medicine that can be transferred to the Chilean health system to support managing clinical text in Spanish.
Palabras clave
Procesamiento del Lenguaje Natural
Informática Médica
Inteligencia Artificial
Listas de Espera
Keywords
Natural Language Processing
Medical Informatics
Artificial Intelligence
Waiting Lists
==== Body
pmcIntroducción
Tal como ocurre en muchos países del continente y el mundo, el sistema de salud en Chile enfrenta problemáticas derivadas del envejecimiento poblacional, la mayor expectativa de vida, la multimorbilidad y el incremento en los servicios de salud, entre otras. Actualmente, diversos aspectos se han agravado debido a la pandemia del coronavirus SARS-CoV-2 (COVID-19), cuyas consecuencias no sólo han impactado negativamente en el ámbito económico, sino también en la salud pública del país. A la par, estamos experimentando una transformación digital sin precedentes, cuyo núcleo es la aplicación de la inteligencia artificial (en adelante IA) para generar herramientas que contribuyan a desarrollar un sistema de salud más eficiente, que cubra las necesidades de los pacientes con soluciones costo-efectivas.
Un problema de salud pública que se ha acrecentado significativamente por la pandemia y que puede ser abordado con el apoyo de la IA es el de las listas de espera. En particular, y tal como se explicará en detalle en las siguientes secciones, la razón de interconsulta en la lista de espera para nueva consulta de especialidad en hospitales públicos chilenos está escrita en texto no estructurado, un problema que ha motivado por varios años nuestro trabajo como grupo de Procesamiento de Lenguaje Natural en Medicina del Centro de Modelamiento Matemático de la Universidad de Chile (http://pln.cmm.uchile.cl).
En este artículo describimos el problema de las listas de espera en Chile y lo presentamos como ejemplo de la aplicación de procesamiento de lenguaje natural (en adelante PLN). Nos centraremos en cómo esta rama de la inteligencia artificial puede proporcionar soluciones a problemas de salud pública, y mostraremos otras tareas de PLN en el contexto clínico chileno.
El sistema de salud chileno y las listas de espera
El sistema de salud en Chile se compone de un sistema mixto de atención integrado por el seguro público, el Fondo Nacional de Salud (en adelante FONASA), y uno privado formado por Instituciones de Salud Previsional (ISAPRES)1. Es de interés el sistema público, ya que al año 2019 concentraba el 78% de la población2.
En el sistema público no se puede acceder directamente a una atención con un especialista. Cuando, a juicio del equipo de salud, se requiere la opinión de éste, se emite una solicitud de interconsulta junto con una derivación al nivel secundario de atención en salud. Es en estos centros donde se solicitan exámenes, se definen tratamientos y seguimiento. Ante la imposibilidad de atender esta consulta inmediatamente, surge el concepto de lista de espera3. En términos generales, las listas de espera pueden ser definidas desde dos perspectivas:• Genéricamente, se definen como un conjunto de personas que, en un momento dado, se encuentran en espera de ser atendidas para una consulta de especialidad médica u odontológica, para un procedimiento o prueba diagnóstica o para una intervención quirúrgica programada, solicitada por un profesional médico u odontológico autorizado en la red de salud y teniendo documentada tal petición.
• Operacionalmente, se entienden como el universo de registros que no poseen una causal de salida y aquellos identificados por la causal de salida N°3, es decir, aquellos pacientes que se encuentran en re-evaluación para el diagnóstico y por ende, no salen de las listas de espera4.
Las Garantías Explícitas en Salud (en adelante, GES) corresponden a los beneficios, garantizados por ley, para la población afiliada tanto a las ISAPRES como a FONASA5, correspondiente a un conjunto de 87 patologías. Algunos ejemplos de las patologías cubiertas por el GES son: enfermedad renal crónica etapa 4 y 5, diabetes mellitus tipo I y II, e hipertensión arterial6.
Las garantías exigibles para las patologías GES son:
Acceso: es el derecho a recibir las atenciones correspondientes, según la Ley, para cada patología listada.
Oportunidad: tiempos máximos de espera para el otorgamiento de las prestaciones.
Protección financiera: la persona beneficiaria cancelará un porcentaje de la afiliación.
Calidad: otorgamiento de las prestaciones por un prestador acreditado o certificado5.
El problema que presenta la lista de espera radica en la demora de resolución y priorización de cada caso, más que en el número de personas que conforman dicha lista7. Específicamente, durante el año 2016 fallecieron 16.625 personas esperando ser atendidas por un médico especialista para patologías no cubiertas por el GES, mientras que, en el mismo año, 993 personas fallecieron esperando en una lista de espera GES. Lo anterior se estableció combinando tanto las listas de espera para consultas de especialidad como la lista quirúrgica8. Así, al 31 de marzo del 2021, el número de personas en lista de espera para patologías no GES ascendía a 1.932.4229. En consecuencia, tanto la alta mortalidad de la lista de espera no GES, como el tiempo que los pacientes esperan para ser atendidos, nos ha motivado como grupo a estudiar esta lista de espera en particular.
Esta situación es preocupante para la población chilena y las autoridades, por lo que deben buscarse medidas para revertir este tiempo prolongado de espera, junto con sus desfavorables consecuencias. Para abordar este problema, es importante identificar la dinámica en la que se registra la información. El médico tratante primero debe ingresar la información clínica del paciente a una plataforma de ficha clínica electrónica, la que contiene campos tanto de información estructurada como no estructurada, en donde se justifica la solicitud. La información estructurada puede ser categórica o numérica, mientras que la no estructurada corresponde a información de tipo imágenes, señales o texto10. Del total de la información, un 40% corresponde a las denominadas ‘narrativas clínicas’ o texto libre11.
La información estructurada es de fácil análisis y comprensión, lo que posibilita generar documentos equivalentes y acotados, pero puede omitir información relevante respecto al proceso diagnóstico. Por otro lado, el texto libre podría contener faltas de ortografía, uso de abreviaciones, jerga local, entre otros, lo que dificulta su análisis. Su contenido, sin embargo, es de gran relevancia, ya que las narrativas contienen información detallada sobre enfermedades, su sintomatología y el fundamento clínico escrito por profesionales de la salud.
Al analizar las interconsultas que forman parte de la lista de espera no GES, una contribución a la resolución del problema podría ser la implementación de criterios de prioridad, la realización de estadísticas de la derivación, la determinación de los casos que pueden ser resueltos por medio de telemedicina, la identificación de factores de riesgo y, finalmente, la búsqueda de relaciones con los antecedentes familiares, entre otros. Todas estas tareas pueden ser agilizadas al utilizar técnicas de PLN12, 13.
Procesamiento de lenguaje natural (PLN)
El PLN es una rama de la inteligencia artificial que puede ser utilizada para el análisis del texto y discurso producido por humanos14. Específicamente, el PLN tiene como objetivos estudiar, diseñar y aplicar sistemas informáticos que faciliten la comunicación entre personas y, entre personas y máquinas15, 16. En otras palabras, se persigue imitar artificialmente algunos de los aspectos de la capacidad humana para el lenguaje, lo que se traduce en procesos de producción y comprensión.
En definitiva, el PLN busca entregar soluciones a los problemas concretos derivados del intento por reproducir artificialmente los patrones con los que funcionan la mente y el lenguaje humanos, para transferirlos a la relación entre humanos y máquinas. Algunas de las tareas tradicionales que aborda el PLN son las siguientes:1. Desarrollo de sistemas de diálogo: sistemas conversacionales que tienen por objetivo optimizar la precisión del proceso comunicativo, ya sea por medio de un chat o del reconocimiento acústico-fonético por parte de la máquina17.
2. Extracción y recuperación de información: ámbito de investigación dedicado al desarrollo de procedimientos informáticos para encontrar documentos que contienen datos textuales no estructurados (es decir, que carecen de una estructura semánticamente abierta y fácil de usar por una máquina)18.
3. Traducción automática o semiautomática: conjunto de técnicas para el desarrollo de programas especializados que buscan traducir datos textuales desde una lengua origen hacia una lengua meta, a partir de implementaciones estadísticas19 o neuronales20.
En cuanto a la aplicación del PLN en el ámbito de la medicina, algunas de las herramientas de mayor utilidad sirven para la extracción automática de información en documentos electrónicos, como lo son las fichas clínicas21. Una gran proporción de la información contenida en las fichas clínicas es no estructurada, lo que hace que su análisis sea de mayor dificultad pues no puede ser fácilmente resumida22. El procesamiento de las fichas clínicas permite asistir a los profesionales de la salud en estudios retrospectivos, en la toma de decisiones clínicas, obtención y comparación de datos sobre enfermedades poco frecuentes, creación de estadísticas e identificación de factores de riesgo23, así como también en la evaluación de eficiencia y costos del cuidado de la salud24.
Otras aplicaciones de PLN en medicina son: detección de reacciones adversas a medicamentos25, 26, detección de eventos y síntomas que preceden al diagnóstico de cáncer27, 28, evaluación del riesgo de suicidio en personas con depresión29, 30, asociación de comorbilidades y otras aplicaciones que pueden ayudar a elaborar estudios epidemiológicos11. Asimismo, cabe nombrar la utilidad en la asignación automática de códigos CIE-10 (u otras nomenclaturas)31, 32 al texto libre, normalizando diversas expresiones empleadas para describir la misma patología.
Uno de los enfoques más comunes en PLN en el ámbito médico es emplear el texto clínico (fichas clínicas o parte de ellas como reportes de imagenología, anatomía patológica o interconsultas, entre otros) para construir un corpus. Un corpus es un conjunto ordenado de datos o textos que sirve de base a una investigación, y que es representativo de un ámbito o variedad lingüística particular33. Estos corpus pueden enriquecerse con información interpretativa mediante un proceso denominado anotación34. La información contenida en los textos es analizada por una persona con experiencia (anotador), quien identifica y señala la información clave (por ejemplo, las enfermedades que presentan los pacientes). Estos textos anotados pueden ser utilizados para entrenar sistemas de reconocimiento automatizado de dicha información clave11.
Precisamente, la detección de entidades nombradas (NER por su sigla en inglés), consiste en el reconocimiento de clases semánticas en texto libre no estructurado35. La detección de entidades como enfermedades, signos y síntomas o procedimientos, son ejemplos de NER en el contexto médico. Sin embargo, gran parte del trabajo previo está desarrollado en la lengua inglesa, lo que junto a la carencia de recursos lingüísticos para la lengua española en el dominio clínico, supone una dificultad para el análisis de textos en esta lengua22.
Aportes al problema de la lista de espera con PLN
Nuestro grupo construyó el primer corpus de texto clínico chileno de libre acceso, a partir de datos de la lista de espera no cubierta por el Plan GES para nueva consulta de especialidad. Esta muestra está compuesta por 2.592.925 interconsultas integradas por las 50 especialidades médicas y dentales22. Adicionalmente, empleamos 10.000 interconsultas para construir un corpus manualmente anotado con hallazgos clínicos, resultados de test o de laboratorio, signos y síntomas, enfermedades, partes del cuerpo, medicamentos, abreviaciones, miembros de la familia y procedimientos de laboratorio, diagnósticos y de tratamiento12.
El corpus anotado se empleó para entrenar un modelo de NER, el cual permite identificar de forma automática enfermedades, medicamentos y partes del cuerpo, con un F1-score de 82,9%, 84,3% y 86,5%, respectivamente. El F1-score es una métrica ampliamente utilizada en la estimación del rendimiento de sistemas de PLN o de recuperación de la información, la cual combina la precisión y el recall o exhaustividad36 y permite hacer comparaciones entre diversas aproximaciones para resolver una tarea en cuestión. Nuestro objetivo es que dicho corpus, que actualmente se encuentra disponible gratuitamente para uso no comercial (https://zenodo.org/record/5591011#.Yt3-K-zMI-R), permita reducir la lista de espera mediante la identificación de casos que, a su vez, pueden ser solucionados mediante otras estrategias tales como telemedicina y la priorización de pacientes mediante análisis de comorbilidades. Otro aspecto importante que queremos abordar en futuras investigaciones y colaboraciones con el Ministerio de Salud es la identificación de procedimientos pendientes.
Codificación automática
Una de las dificultades en el análisis de texto no estructurado radica en la variabilidad del lenguaje que puede ser empleado para referirse a una misma entidad; específicamente, casos de ambigüedad en la asignación de diferentes etiquetas lingüísticas para el mismo sentido. Por ejemplo, las expresiones «cáncer de mama», «tumor mamario», «ca. mama», «tu. mamario», «carcinoma mamario», pueden ser empleadas, indistintamente, para hacer alusión a la misma enfermedad. Una de las soluciones para reducir esta variabilidad es la normalización de los términos médicos, mediante la asignación de, por ejemplo, su código correspondiente desde los conceptos del Sistema de Lenguaje Médico Unificado (UMLS por su sigla en inglés).
El UMLS Metathesaurus, desarrollado por la Biblioteca Nacional de Medicina de los Estados Unidos37, es una herramienta creada principalmente para resolver dos barreras importantes frente a la capacidad de las máquinas para extraer información: la variedad de nombres para referirse al mismo concepto, como ya lo hemos mencionado, y la ausencia de un formato establecido para distribuir terminologías. Esta herramienta contiene un compilado de nombres, relaciones e información asociada de una variedad de sistemas biomédicos que integra más de dos millones de nombres de aproximadamente 900.000 conceptos de vocabulario médico, y no sólo eso, sino que también posee más de 12 millones de relaciones entre todos estos conceptos.
Actualmente, contamos con una fracción de 2.000 interconsultas médicas anotadas en el Corpus de Lista de Espera Chilena, cuyas entidades médicas fueron normalizadas de forma automatizada empleando el léxico MedLexSp38, 39 asignándole uno o múltiples códigos únicos de identificación a cada entidad. Este recurso estará disponible para su libre uso próximamente.
Otros sistemas de PLN clínico desarrollados en Chile
El Plan Nacional del Cáncer 2018-2028 en Chile propone una línea estratégica para fortalecer los Sistemas de Registro, Información y Vigilancia del Cáncer. A partir de lo anterior, el PLN puede apoyar la extracción automática de información y la sistematización de bases de datos para este fin. Recientemente, hemos desarrollado un sistema de apoyo que facilita la codificación CIE-O de la morfología y topografía de los tumores en los informes de patología, una tarea esencial para los registros de cáncer40. Este sistema puede probarse en el siguiente enlace: https://topomorfo.oncodata.org.
Otro de nuestros desarrollos en esta área permite detectar automáticamente menciones de metástasis a distancia en reportes de imagenología. Específicamente, la clasificación TNM de tumores aporta información relevante en la definición de un estadio, definiendo las características del tumor primario (T), la posible propagación a ganglios linfáticos cercanos (N) y la presencia o ausencia de metástasis a otras partes del cuerpo (M). Como antecedente, es relevante considerar que la detección manual de los parámetros TNM consume mucho tiempo y horas de personal, pues implica el análisis individual de cada reporte. Tanto este desarrollo como el anterior se enmarcan en una colaboración con la Fundación Arturo López Pérez, considerado como el mayor centro oncológico de Chile, que atiende a más de 50 mil pacientes al año, la mayoría pertenecientes al sistema público de salud41.
Un algoritmo de clasificación que emplea técnicas de PLN e historias clínicas anonimizadas de un hospital de Chile fue desarrollado por Ramos et al.42 para clasificar los diagnósticos de los pacientes discriminando entre las clases ‘cáncer’ frente a ‘no cáncer’ y ‘cáncer de mama’ frente a ‘otro cáncer’. Este algoritmo podría utilizarse como herramienta de apoyo y recomendación del diagnóstico de los pacientes, principalmente para médicos que inician sus labores en sectores alejados con poco personal especializado como, por ejemplo, hospitales rurales.
Existen varios trabajos enfocados en el uso secundario de datos médicos y clasificación de textos desarrollados por investigadores de la Universidad de Concepción43, 44, 45. En esta propuesta se emplearon textos clínicos en español, provenientes del Hospital Clínico Regional Dr. Guillermo Grant Benavente de Concepción, para identificar y extraer información sobre el estado de tabaquismo de los pacientes, mediante técnicas de PLN y minería textual46, junto con información sobre medidas de peso corporal y comorbilidades47. En ambos casos, los datos extraídos fueron posteriormente utilizados como corpus para algoritmos de clasificación de textos. A partir de esta implementación, por ejemplo, fue posible determinar si un paciente era fumador, no fumador, fumador actual o fumador pasado.
Lecaros et al.48, por su parte, examinaron derivaciones contenidas en el corpus de la lista de espera para detectar casos de pacientes con psoriasis y, así, determinar la incidencia de dicha patología en Chile. Recientemente, Figueroa-Barra et al.49 emplearon técnicas de PLN dentro de un sistema de análisis automático del lenguaje para identificar y predecir esquizofrenia en el primer episodio de psicosis. Este sistema se empleó en entrevistas clínicas en español y se basa en el análisis de 30 rasgos lingüísticos que permiten distinguir los controles sanos de los pacientes con esquizofrenia crónica, y predecir el diagnóstico de esquizofrenia en pacientes con primer episodio de psicosis.
Conclusiones
El texto clínico representa una proporción importante de la información registrada de los y las pacientes. Su procesamiento masivo y utilización en la toma de decisiones requiere herramientas del estado del arte en PLN; correspondiente a una de las tres ramas de la inteligencia artificial, junto con la visión por computador y la robótica.
Un área que hemos comenzado a desarrollar es el estudio de las propiedades lingüísticas de distintos textos clínicos, lo que es relevante para desarrollar sistemas en diversas especialidades médicas. En efecto, el reconocimiento de patrones lingüísticos presenta implicaciones significativas en términos de la generalización, ya que las herramientas de PLN suelen adaptarse de manera eficiente a dominios estrechos, con lenguajes bien definidos y comprendidos50. Además, nuestro equipo colabora con grupos en España y Argentina, con el fin de aunar esfuerzos para la construcción de recursos lingüísticos que consideren la variedad de estructuras presentes en la lengua española, y de esta manera evaluar sus similitudes o diferencias.
Para avanzar en el uso del PLN en español es necesario obtener narrativas clínicas, frecuentemente anotadas por humanos, así como modelos computacionales que permitan realizar la tarea en cuestión. Ejemplos de tareas relevantes de PLN clínico son: la detección de información clave, la clasificación de textos o la asignación de códigos internacionales. En el presente artículo, mostramos ejemplos de los avances en estas tareas, lideradas por nuestro grupo de investigación y otros grupos nacionales.
Según lo anterior, recibir apoyo político y gubernamental es esencial para que trabajos como estos puedan ser transferidos al sistema de salud chileno. En particular, desde octubre de 2022 hemos comenzado a colaborar con el Departamento de Estadísticas e Información de Salud (DEIS) del Ministerio de Salud. El objetivo de esta colaboración es transferir nuestro conocimiento adquirido en el trabajo con texto libre de la lista de espera no GES para apoyar la gestión de ésta. En particular, nos hemos propuesto codificar todas las enfermedades mencionadas en las razones de interconsultas, y con ello apoyar estrategias tales como la detección de pacientes GES en la lista no GES, el Registro Nacional de Cáncer o la articulación con el Hospital Digital para el uso de la telemedicina.
Un desafío actual en la investigación en PLN clínico es lograr acceso a narrativas médicas preservando la privacidad de los pacientes y las regulaciones cada vez más estrictas en algunos países. Una solución es la creación de un corpus sintético que preserve las propiedades lingüísticas del corpus original, pero cuya información no guarde relación con los pacientes reales51. Este enfoque es muy promisorio, dado que podría corregir las graves brechas de representación de grupos identitarios52.
Varias condiciones deben darse para que nuevas tecnologías como el PLN sean utilizadas tanto en el sistema público como el privado. Obviamente, es necesaria la voluntad política, por un lado, y el acceso a poder de cómputo, por otro. Además, se debe promover una estrategia que permita asegurar un uso ético de la IA. En efecto, la Organización Mundial de la Salud (https://www.who.int/publications/i/item/9789240029200) recomienda seis principios que deben tenerse en cuenta, los que incluyen: no perder la autonomía humana en la toma de decisiones, incluir medidas de mejoramiento continuo, explicabilidad de los modelos, responsabilidad sobre las tareas que está realizando la máquina, garantizar un uso no-discriminatorio y sustentabilidad. En definitiva, el cumplimiento de todas estas condiciones no es una tarea sencilla. No obstante, contar con ellas como meta constituye una ayuda relevante para crear consciencia y promover el trabajo interdisciplinario y el mejoramiento continuo de los modelos.
PLN en medicina es un área transdisciplinaria por definición, en la que confluyen las ciencias de la computación, la medicina y la lingüística. Para lograr un mayor impacto, se requiere del trabajo conjunto de profesionales de la salud y tomadores de decisión que conozcan y valoren los alcances de la inteligencia artificial aplicada al ámbito de la salud.
Declaración de conflicto de interés
Los autores declaran no tener conflictos de intereses.
Financiamiento
Este trabajo ha sido financiado por la ANID a través de los Fondos Basales para Centros de Excelencia FB210005 (Centro de Modelamiento Matemático), Fondecyt de Iniciación 11201250 (J. Dunstan) y Fondecyt de Postdoctorado 3210395 (P. Báez). Además, la investigación conducida por J. Dunstan es apoyada por los Institutos Milenio ICN2021_004 (iHealth) e ICN17_002 (IMFD).
==== Refs
Bibliografía
1 Cómo funciona el Sistema de Salud en Chile Cómo funciona el Sistema de Salud en Chile [Internet]. Orientación en Salud. Superintendencia de Salud, Gobierno de Chile 2022 [citado el 23 de julio de 2022]. Disponible en: http://www., supersalud., gob., cl/difusion/665/w3-article-17328., html.
2 Informe CDD: Caracterización sociodemográfica y socioeconómica en la población asegurada inscrita. Gobierno de Chile. [Internet]. División de Desarrollo Institucional, Departamento de Estudios y Estadísticas; 2020 [citado el 23 de julio de 2022]. Disponible en: https://www.fonasa.cl/sites/fonasa/adjuntos/Informe_caracterizacion_poblacion_asegurada.
3 Estay R. Cuadrado C. Crispi F. González F. Alvarado F. Cabrera N. Desde el conflicto de listas de espera, hacia el fortalecimiento de los prestadores públicos de salud: Una propuesta para Chile Cuad Med Soc. 57 1 2017 49 63
4 Cordero A, Valderrama S, Duran L, Santelices P, Olave M, Jiron P, et al., Norma técnica para el registro de las listas de espera. 2016; Disponible en: https://www.minsal.cl/sites/default/files/files/Nueva%20Norma%20de%20Listas%20de%20Espera%202011.pdf.
5 ¿Qué es el AUGE o GES? - ¿Qué es el AUGE o GES? [Internet]. Preguntas Frecuentes. Superintendencia de Salud, Gobierno de Chile. [citado el 18 de julio de 2022]. Disponible en: http://www.supersalud.gob.cl/consultas/667/w3-article-4605.html.
6 Garantías Explícitas en Salud (GES) [Internet]. Orientación en Salud. Superintendencia de Salud, Gobierno de Chile. [citado el 18 de julio de 2022]. Disponible en: http://www.supersalud.gob.cl/difusion/665/w3-propertyvalue-1962.html.
7 Martinez D.A. Zhang H. Bastias M. Feijoo F. Hinson J. Martinez R. Prolonged wait time is associated with increased mortality for Chilean waiting list patients with non-prioritized conditions BMC Public Health. 19 1 2019 233 doi: 10.1186/s12889-019-6526.-6. 30808318
8 MINSAL. Estado de Situacion Personas Fallecidas en Listas de Espera No-GES y Garantias Retrasadas GES [Internet]. 2017 [citado el 25 de julio de 2022]. Disponible en: https://www.minsal.cl/wp-content/uploads/2018/01/Informe-Final-Comision-Asesora-LE-y-Garantias-Retrasadas-GES-17082017_.pdf.
9 MINSAL. Glosa 06 - Lista de espera no GES y garantías de oportunidad GES retrasadas [Internet]. 2021 [citado el 23 de julio de 2022]. Disponible en: https://www.minsal.cl/wp-content/uploads/2021/05/ORD-1204-DIGERA-Glosa-06.pdf.
10 Villena F. Dunstan J. Obtención automática de palabras clave en textos clínicos: una aplicación de procesamiento del lenguaje natural a datos masivos de sospecha diagnóstica en Chile. [Automatic keyword retrieval from clinical texts: an application of natural language processing to massive data of Chilean suspected diagnosis] Rev Med Chile. 147 10 2019 1229 1238 32186630
11 Dalianis H. Clinical text mining: Secondary use of electronic patient records 2018 Springer Nature
12 Báez P, Villena F, Rojas M, Durán M, Dunstan J. The Chilean Waiting List Corpus: a new resource for clinical named entity recognition in Spanish. In: Proceedings of the 3rd Clinical Natural Language Processing Workshop. Association for Computational Linguistics. 2020. p.291-300. doi 10.18653/v1/2020.clinicalnlp-1.32.
13 Báez P. Bravo-Marquez F. Dunstan J. Rojas M. Villena F. Automatic Extraction of Nested Entities in Clinical Referrals in Spanish ACM Trans Comput Healthc Health. 3 3 2022 1 22 doi: 10.1145/3498.324.
14 Jurafsky D, Martin JH. Speech and Language Processing [Internet]. Third Ed. draft. 2022 [citado el 18 de julio de 2022]. Disponible en: https://web.stanford.edu/∼jurafsky/slp3/ed3book_jan122022.pdf.
15 Periñán Pascual J.C. Mairal Usón R. Bringing Role and Reference Grammar to natural language understanding Proces Leng Nat. 43 2009 265 273
16 Choi R.Y. Coyner A.S. Kalpathy-Cramer J. Chiang M.F. Campbell J.P. Introduction to Machine Learning, Neural Networks, and Deep Learning Transl Vis Sci Technol. 9 2 2020 14 doi: 10.1167/tvst.9.2.14.
17 Gerbino E, Baggia P, Giachin E, Rullent C. Analysis and Evaluation of Spontaneous Speech Utterances in Focused Dialogue Contexts. Proc. of ESCA Workshop. 1995.
18 Savoy J. Gaussier É. Information Retrieval Indurkhya N. Damerau F.J. Handbook of Natural Language Processing 2nd Ed. 2010 Chapman and Hall/CRC
19 Sinhal R.A. Gupta K.O. Machine translation approaches and design aspects IOSR J Comput Eng. 16 1 2014 22 25
20 Dabre R. Chu C. Kunchukuttan A. A Survey of Multilingual Neural Machine Translation ACM Comput Surv. 53 5 2020 99 1-99:38. doi: 10.1145/3406095.
21 Oronoz M. Gojenola K. Pérez A. de Ilarraza A.D. Casillas A. On the creation of a clinical gold standard corpus in Spanish: Mining adverse drug reactions J Biomed Inform. 56 2015 318 332 doi: 10.1016/j.jbi.2015.06.016. 26141794
22 Báez P. Villena F. Zúñiga K. Jones N. Fernández G. Durán M. Construcción de recursos de texto para la identificación automática de información clínica en narrativas no estructuradas. [Construction of text resources for automatic identification of clinical information in unstructured narratives] Rev Med Chil. 149 7 2021 1014 1022 Spanish. doi: 10.4067/s0034-98872021000701014. 34751303
23 Névéol A. Dalianis H. Velupillai S. Savova G. Zweigenbaum P. Clinical Natural Language Processing in languages other than English: opportunities and challenges J Biomed Semantics 9 1 2018 12 doi: 10.1186/s13326-018-0179-8. 29602312
24 Koleck T.A. Dreisbach C. Bourne P.E. Bakken S. Natural language processing of symptoms documented in free-text narratives of electronic health records: a systematic review J Am Med Inform Assoc. 26 4 2019 364 379 doi: 10.1093/jamia/ocy173. 30726935
25 Jagannatha A. Liu F. Liu W. Yu H. Overview of the First Natural Language Processing Challenge for Extracting Medication. Indication, and Adverse Drug Events from Electronic Health Record Notes (MADE 1.0) Drug Saf. 42 1 2019 99 111 doi: 10.1007/s40264-018-0762-z. 30649735
26 Chen L. Gu Y. Ji X. Sun Z. Li H. Gao Y. Huang Y. Extracting medications and associated adverse drug events using a natural language processing system combining knowledge base and deep learning J Am Med Inform Assoc. 27 1 2020 56 64 doi: 10.1093/jamia/ocz141. 31591641
27 Weegar R. Kvist M. Sundström K. Brunak S. Dalianis H. Finding Cervical Cancer Symptoms in Swedish Clinical Text using a Machine Learning Approach and NegEx AMIA Annu Symp Proc 2015 2015 1296 1305 26958270
28 Weegar R. Mining events preceding a cancer diagnosis En IEEE Computer Society 2018 295 296 doi: 10.1109/eScience.2018.00059.
29 Orooji A. Langarizadeh M. Using of natural language processing techniques in suicide research Emerg Sci J. 1 2 2017 89 96 doi: 10.28991/esj-2017-01120.
30 Levis M. Leonard Westgate C. Gui J. Watts B.V. Shiner B. Natural language processing of clinical mental health notes may add predictive value to existing suicide risk models Psychol Med. 2020 1 10 doi: 10.1017/S0033291720000173
31 Henry S. Wang Y. Shen F. Uzuner O. The 2019 National Natural language processing (NLP) Clinical Challenges (n2c2)/Open Health NLP (OHNLP) shared task on clinical concept normalization for clinical records J Am Med Inform Assoc. 27 10 2020 1529 1537 doi: 10.1093/jamia/ocaa106. Erratumin: J Am Med Inform Assoc. 20***21;28(11):2546 32968800
32 Blanco A. Perez-de-Viñaspre O. Pérez A. Casillas A. Boosting ICD multi-label classification of health records with contextual embeddings and label-granularity Comput Methods Programs Biomed. 188 2020 105264 doi: 10.1016/j.cmpb.2019.1052.64. 31851906
33 Dunstan J, Maass A, Tobar F. Una mirada a la era de los datos. 1 ed. Ed. Universitaria SA; 2022. 134 p.
34 Fort K. Collaborative Annotation for Reliable Natural Language Processing: Technical and Sociological Aspects 2016 John Wiley & Sons
35 Pérez A. Weegar R. Casillas A. Gojenola K. Oronoz M. Dalianis H. Semi-supervised medical entity recognition: A study on Spanish and Swedish clinical corpora J Biomed Inform. 71 2017 16 30 doi: 10.1016/j.jbi.2017.05.009. 28526460
36 Goutte C. Gaussier E. A Probabilistic Interpretation of Precision, Recall and F-Score with Implication for Evaluation Losada D.E. Fernández-Luna J.M. Advances in Information Retrieval. ECIR. Lecture Notes in Computer Science, vol 3408 2005 Springer Berlin, Heidelberg doi: 10.1007/978-3-540-31865-1_25.
37 Schuyler P.L. Hole W.T. Tuttle M.S. Sherertz D.D. The UMLS Metathesaurus: representing different views of biomedical concepts Bull Med Libr Assoc. 81 2 1993 217 222 8472007
38 Campillos-Llanos L. Medical Lexicon for Spanish (MedLexSp) 2022 doi: 10.20350/digitalCSIC/1465.6.
39 Campillos-Llanos L. First steps towards building a medical Lexicon for Spanish with linguistic and semantic information In Proceedings of the 18th BioNLP Workshop and Shared Task. doi:10.18653/v1/W19-5017 2019 152 164
40 Villena F, Báez P, Peñafiel S, Rojas M, Paredes I, Dunstan J. Automatic Support System for Tumor Coding in Pathology Reports in Spanish. SSRN Electron J [Internet]. 2021. doi:10.2139/ssrn.3982259. Disponible en: https://www.ssrn.com/abstract=3982259.
41 Memoria de Gestión FALP [Internet]. Fundación Arturo López Pérez. 2017 [citado el 18 de julio de 2022]. Disponible en: https://www.institutoncologicofalp.cl/fundacion/memoria-falp/.
42 Ramos A.A. Allende-Cid H. Taramasco C. Becerra C. Figueroa R.L. Application of Machine Learning and Word Embeddings in the Classification of Cancer Diagnosis Using Patient Anamnesis IEEE Access. 8 2020 106198 106213
43 Flores C.A. Figueroa R.L. Pezoa J.E. Active Learning for Biomedical Text Classification Based on Automatically Generated Regular Expressions IEEE Access. 9 2021 38767 38777 doi: 10.1109/ACCESS. 2021.3064.000.
44 Flores C.A. Figueroa R.L. Pezoa J.E. Zeng-Treitler Q. CREGEX: A Biomedical Text Classifier Based on Automatically Generated Regular Expressions IEEE Access. 8 2020 29270 29280 doi: 10.1109/ACCESS. 2020.2972.205.
45 Flores C.A. Figueroa R.L. Pezoa J.E. FREGEX: A Feature Extraction Method for Biomedical Text Classification using Regular Expressions Annu Int Conf IEEE Eng Med Biol Soc. 2019 2019 6085 6088 doi: 10.1109/EMBC. 2019.8857.471. 31947233
46 Figueroa R.L. Soto D.A. Pino E.J. Identifying and extracting patient smoking status information from clinical narrative texts in Spanish Annu Int Conf IEEE Eng Med Biol Soc. 2014 2014 2710 2713 doi: 10.1109/EMBC. 2014.6944.182. 25570550
47 Figueroa R.L. Flores C.A. Extracting Information from Electronic Medical Records to Identify the Obesity Status of a Patient Based on Comorbidities and Bodyweight Measures J Med Syst. 40 8 2016 191 doi: 10.1007/s10916-016-0548-8. 27402260
48 Lecaros C. Dunstan J. Villena F. Ashcroft D.M. Parisi R. Griffiths C.E.M. The incidence of psoriasis in Chile: an analysis of the National Waiting List Repository Clin Exp Dermatol. 46 7 2021 1262 1269 doi: 10.1111/ced.14713. 33914930
49 Figueroa-Barra A. Del Aguila D. Cerda M. Gaspar P.A. Terissi L.D. Durán M. Automatic language analysis identifies and predicts schizophrenia in first-episode of psychosis Schizophrenia (Heidelb). 8 1 2022 53 doi: 10.1038/s41537-022-0025.9-3. 35853943
50 Zeng Q.T. Redd D. Divita G. Jarad S. Brandt C. Nebeker J.R. Characterizing clinical text and sublanguage: A case study of the VA clinical notes J Health Med Informat S. 2011 S3 doi:10.4172/2157-7420.S3-001.
51 Bannour N. Wajsbürt P. Rance B. Tannier X. Névéol A. Privacy-preserving mimic models for clinical named entity recognition in French J Biomed Inform. 130 2022 104073 doi: 10.1016/j.jbi.2022.1040.73. 35427797
52 Criado-Perez C. La mujer invisible: descubre cómo los datos configuran un mundo hecho por y para los hombres 2020 Seix Barral España
| 0 | PMC9704358 | NO-CC CODE | 2022-12-01 23:19:04 | no | 2022 Nov 28 November-December; 33(6):576-582 | utf-8 | null | null | null | oa_other |
==== Front
Revista Médica Clínica Las Condes
0716-8640
0716-8640
S0716-8640(22)00120-1
10.1016/j.rmclc.2022.10.002
Article
Procesamiento de lenguaje natural para texto clínico en español: el caso de las listas de espera en Chile
Natural language processing for clinical text in Spanish: The case of waiting lists in ChileBáez Pablo a
Arancibia Antonia Paz b
Chaparro Matías Ignacio b
Bucarey Tomás b
Núñez Fredy cd
Dunstan Jocelyn def⁎
a Centro de Informática Médica y Telemedicina, Facultad de Medicina, Universidad de Chile. Santiago, Chile
b Escuela de Medicina, Facultad de Medicina, Universidad de Chile. Santiago, Chile
c Departamento de Ciencias del Lenguaje, Facultad de Letras, Pontificia Universidad Católica de Chile. Santiago, Chile
d Centro de Modelamiento Matemático, Facultad de Ciencias Físicas y Matemáticas, Universidad de Chile. Santiago, Chile
e Iniciativa de Datos & Inteligencia Artificial, Facultad de Ciencias Físicas y Matemáticas, Universidad de Chile. Santiago, Chile
f Instituto Milenio Engineering for Healthcare, ANID, Chile. Santiago, Chile
⁎ Autor para correspondencia.
28 11 2022
November-December 2022
28 11 2022
33 6 576582
30 7 2022
16 10 2022
18 10 2022
.
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Las listas de espera no cubiertas por el Plan de Garantías Explícitas en Salud para nueva consulta de especialidad en Chile se han visto incrementadas por los efectos de la pandemia del coronavirus SARS-CoV-2 (COVID-19). Esto representa un problema debido a la demora en la resolución y priorización de cada caso de derivación al nivel secundario de atención en salud. El objetivo de este artículo es exponer el problema de la lista de espera en el sistema de salud de Chile, y abordarlo como ejemplo de la aplicación de técnicas de Procesamiento del Lenguaje Natural (PLN). Específicamente, se describe una metodología para el reconocimiento de información clave en narrativas médicas. Actualmente, contamos con un conjunto de interconsultas médicas manualmente anotadas en el desarrollo del Corpus de Lista de Espera Chilena, y con una fracción de 2.000 interconsultas en las que las entidades médicas anotadas fueron normalizadas de forma automatizada a los conceptos del Sistema de Lenguaje Médico Unificado empleando el léxico MedLexSp. Este y otros recursos lingüísticos y herramientas de PLN están siendo desarrollados por el grupo de PLN en Medicina del Centro de Modelamiento Matemático de la Universidad de Chile y otros grupos a nivel nacional, los cuales constituyen aportes relevantes que pueden ser transferidos al sistema de salud chileno, con el objetivo de apoyar la gestión del texto clínico en español.
The waiting lists not covered by the Explicit Health Guarantee Plan for new specialty consultation in Chile increased due to the effects of the SARS-CoV-2 coronavirus (COVID-19) pandemic. This represents a problem derived from the delay in the resolution and prioritization of each case. This paper aims to describe the issue of the waiting lists in the Chilean health system and present an example of the application of Natural Language Processing (NLP). Specifically, a methodology for recognizing key information in medical narratives is described. Currently, we have a set of manually annotated medical referrals in the development of the Chilean Waiting List Corpus, with a fraction of 2,000 referrals in which the annotated medical entities were automatically normalized to the Unified Medical Language System concepts using the lexicon MedLexSp. The clinical NLP Group of the Center for Mathematical Modeling of the University of Chile, and other national NLP groups, are developing several tools and resources in medicine that can be transferred to the Chilean health system to support managing clinical text in Spanish.
Palabras clave
Procesamiento del Lenguaje Natural
Informática Médica
Inteligencia Artificial
Listas de Espera
Keywords
Natural Language Processing
Medical Informatics
Artificial Intelligence
Waiting Lists
==== Body
pmcIntroducción
Tal como ocurre en muchos países del continente y el mundo, el sistema de salud en Chile enfrenta problemáticas derivadas del envejecimiento poblacional, la mayor expectativa de vida, la multimorbilidad y el incremento en los servicios de salud, entre otras. Actualmente, diversos aspectos se han agravado debido a la pandemia del coronavirus SARS-CoV-2 (COVID-19), cuyas consecuencias no sólo han impactado negativamente en el ámbito económico, sino también en la salud pública del país. A la par, estamos experimentando una transformación digital sin precedentes, cuyo núcleo es la aplicación de la inteligencia artificial (en adelante IA) para generar herramientas que contribuyan a desarrollar un sistema de salud más eficiente, que cubra las necesidades de los pacientes con soluciones costo-efectivas.
Un problema de salud pública que se ha acrecentado significativamente por la pandemia y que puede ser abordado con el apoyo de la IA es el de las listas de espera. En particular, y tal como se explicará en detalle en las siguientes secciones, la razón de interconsulta en la lista de espera para nueva consulta de especialidad en hospitales públicos chilenos está escrita en texto no estructurado, un problema que ha motivado por varios años nuestro trabajo como grupo de Procesamiento de Lenguaje Natural en Medicina del Centro de Modelamiento Matemático de la Universidad de Chile (http://pln.cmm.uchile.cl).
En este artículo describimos el problema de las listas de espera en Chile y lo presentamos como ejemplo de la aplicación de procesamiento de lenguaje natural (en adelante PLN). Nos centraremos en cómo esta rama de la inteligencia artificial puede proporcionar soluciones a problemas de salud pública, y mostraremos otras tareas de PLN en el contexto clínico chileno.
El sistema de salud chileno y las listas de espera
El sistema de salud en Chile se compone de un sistema mixto de atención integrado por el seguro público, el Fondo Nacional de Salud (en adelante FONASA), y uno privado formado por Instituciones de Salud Previsional (ISAPRES)1. Es de interés el sistema público, ya que al año 2019 concentraba el 78% de la población2.
En el sistema público no se puede acceder directamente a una atención con un especialista. Cuando, a juicio del equipo de salud, se requiere la opinión de éste, se emite una solicitud de interconsulta junto con una derivación al nivel secundario de atención en salud. Es en estos centros donde se solicitan exámenes, se definen tratamientos y seguimiento. Ante la imposibilidad de atender esta consulta inmediatamente, surge el concepto de lista de espera3. En términos generales, las listas de espera pueden ser definidas desde dos perspectivas:• Genéricamente, se definen como un conjunto de personas que, en un momento dado, se encuentran en espera de ser atendidas para una consulta de especialidad médica u odontológica, para un procedimiento o prueba diagnóstica o para una intervención quirúrgica programada, solicitada por un profesional médico u odontológico autorizado en la red de salud y teniendo documentada tal petición.
• Operacionalmente, se entienden como el universo de registros que no poseen una causal de salida y aquellos identificados por la causal de salida N°3, es decir, aquellos pacientes que se encuentran en re-evaluación para el diagnóstico y por ende, no salen de las listas de espera4.
Las Garantías Explícitas en Salud (en adelante, GES) corresponden a los beneficios, garantizados por ley, para la población afiliada tanto a las ISAPRES como a FONASA5, correspondiente a un conjunto de 87 patologías. Algunos ejemplos de las patologías cubiertas por el GES son: enfermedad renal crónica etapa 4 y 5, diabetes mellitus tipo I y II, e hipertensión arterial6.
Las garantías exigibles para las patologías GES son:
Acceso: es el derecho a recibir las atenciones correspondientes, según la Ley, para cada patología listada.
Oportunidad: tiempos máximos de espera para el otorgamiento de las prestaciones.
Protección financiera: la persona beneficiaria cancelará un porcentaje de la afiliación.
Calidad: otorgamiento de las prestaciones por un prestador acreditado o certificado5.
El problema que presenta la lista de espera radica en la demora de resolución y priorización de cada caso, más que en el número de personas que conforman dicha lista7. Específicamente, durante el año 2016 fallecieron 16.625 personas esperando ser atendidas por un médico especialista para patologías no cubiertas por el GES, mientras que, en el mismo año, 993 personas fallecieron esperando en una lista de espera GES. Lo anterior se estableció combinando tanto las listas de espera para consultas de especialidad como la lista quirúrgica8. Así, al 31 de marzo del 2021, el número de personas en lista de espera para patologías no GES ascendía a 1.932.4229. En consecuencia, tanto la alta mortalidad de la lista de espera no GES, como el tiempo que los pacientes esperan para ser atendidos, nos ha motivado como grupo a estudiar esta lista de espera en particular.
Esta situación es preocupante para la población chilena y las autoridades, por lo que deben buscarse medidas para revertir este tiempo prolongado de espera, junto con sus desfavorables consecuencias. Para abordar este problema, es importante identificar la dinámica en la que se registra la información. El médico tratante primero debe ingresar la información clínica del paciente a una plataforma de ficha clínica electrónica, la que contiene campos tanto de información estructurada como no estructurada, en donde se justifica la solicitud. La información estructurada puede ser categórica o numérica, mientras que la no estructurada corresponde a información de tipo imágenes, señales o texto10. Del total de la información, un 40% corresponde a las denominadas ‘narrativas clínicas’ o texto libre11.
La información estructurada es de fácil análisis y comprensión, lo que posibilita generar documentos equivalentes y acotados, pero puede omitir información relevante respecto al proceso diagnóstico. Por otro lado, el texto libre podría contener faltas de ortografía, uso de abreviaciones, jerga local, entre otros, lo que dificulta su análisis. Su contenido, sin embargo, es de gran relevancia, ya que las narrativas contienen información detallada sobre enfermedades, su sintomatología y el fundamento clínico escrito por profesionales de la salud.
Al analizar las interconsultas que forman parte de la lista de espera no GES, una contribución a la resolución del problema podría ser la implementación de criterios de prioridad, la realización de estadísticas de la derivación, la determinación de los casos que pueden ser resueltos por medio de telemedicina, la identificación de factores de riesgo y, finalmente, la búsqueda de relaciones con los antecedentes familiares, entre otros. Todas estas tareas pueden ser agilizadas al utilizar técnicas de PLN12, 13.
Procesamiento de lenguaje natural (PLN)
El PLN es una rama de la inteligencia artificial que puede ser utilizada para el análisis del texto y discurso producido por humanos14. Específicamente, el PLN tiene como objetivos estudiar, diseñar y aplicar sistemas informáticos que faciliten la comunicación entre personas y, entre personas y máquinas15, 16. En otras palabras, se persigue imitar artificialmente algunos de los aspectos de la capacidad humana para el lenguaje, lo que se traduce en procesos de producción y comprensión.
En definitiva, el PLN busca entregar soluciones a los problemas concretos derivados del intento por reproducir artificialmente los patrones con los que funcionan la mente y el lenguaje humanos, para transferirlos a la relación entre humanos y máquinas. Algunas de las tareas tradicionales que aborda el PLN son las siguientes:1. Desarrollo de sistemas de diálogo: sistemas conversacionales que tienen por objetivo optimizar la precisión del proceso comunicativo, ya sea por medio de un chat o del reconocimiento acústico-fonético por parte de la máquina17.
2. Extracción y recuperación de información: ámbito de investigación dedicado al desarrollo de procedimientos informáticos para encontrar documentos que contienen datos textuales no estructurados (es decir, que carecen de una estructura semánticamente abierta y fácil de usar por una máquina)18.
3. Traducción automática o semiautomática: conjunto de técnicas para el desarrollo de programas especializados que buscan traducir datos textuales desde una lengua origen hacia una lengua meta, a partir de implementaciones estadísticas19 o neuronales20.
En cuanto a la aplicación del PLN en el ámbito de la medicina, algunas de las herramientas de mayor utilidad sirven para la extracción automática de información en documentos electrónicos, como lo son las fichas clínicas21. Una gran proporción de la información contenida en las fichas clínicas es no estructurada, lo que hace que su análisis sea de mayor dificultad pues no puede ser fácilmente resumida22. El procesamiento de las fichas clínicas permite asistir a los profesionales de la salud en estudios retrospectivos, en la toma de decisiones clínicas, obtención y comparación de datos sobre enfermedades poco frecuentes, creación de estadísticas e identificación de factores de riesgo23, así como también en la evaluación de eficiencia y costos del cuidado de la salud24.
Otras aplicaciones de PLN en medicina son: detección de reacciones adversas a medicamentos25, 26, detección de eventos y síntomas que preceden al diagnóstico de cáncer27, 28, evaluación del riesgo de suicidio en personas con depresión29, 30, asociación de comorbilidades y otras aplicaciones que pueden ayudar a elaborar estudios epidemiológicos11. Asimismo, cabe nombrar la utilidad en la asignación automática de códigos CIE-10 (u otras nomenclaturas)31, 32 al texto libre, normalizando diversas expresiones empleadas para describir la misma patología.
Uno de los enfoques más comunes en PLN en el ámbito médico es emplear el texto clínico (fichas clínicas o parte de ellas como reportes de imagenología, anatomía patológica o interconsultas, entre otros) para construir un corpus. Un corpus es un conjunto ordenado de datos o textos que sirve de base a una investigación, y que es representativo de un ámbito o variedad lingüística particular33. Estos corpus pueden enriquecerse con información interpretativa mediante un proceso denominado anotación34. La información contenida en los textos es analizada por una persona con experiencia (anotador), quien identifica y señala la información clave (por ejemplo, las enfermedades que presentan los pacientes). Estos textos anotados pueden ser utilizados para entrenar sistemas de reconocimiento automatizado de dicha información clave11.
Precisamente, la detección de entidades nombradas (NER por su sigla en inglés), consiste en el reconocimiento de clases semánticas en texto libre no estructurado35. La detección de entidades como enfermedades, signos y síntomas o procedimientos, son ejemplos de NER en el contexto médico. Sin embargo, gran parte del trabajo previo está desarrollado en la lengua inglesa, lo que junto a la carencia de recursos lingüísticos para la lengua española en el dominio clínico, supone una dificultad para el análisis de textos en esta lengua22.
Aportes al problema de la lista de espera con PLN
Nuestro grupo construyó el primer corpus de texto clínico chileno de libre acceso, a partir de datos de la lista de espera no cubierta por el Plan GES para nueva consulta de especialidad. Esta muestra está compuesta por 2.592.925 interconsultas integradas por las 50 especialidades médicas y dentales22. Adicionalmente, empleamos 10.000 interconsultas para construir un corpus manualmente anotado con hallazgos clínicos, resultados de test o de laboratorio, signos y síntomas, enfermedades, partes del cuerpo, medicamentos, abreviaciones, miembros de la familia y procedimientos de laboratorio, diagnósticos y de tratamiento12.
El corpus anotado se empleó para entrenar un modelo de NER, el cual permite identificar de forma automática enfermedades, medicamentos y partes del cuerpo, con un F1-score de 82,9%, 84,3% y 86,5%, respectivamente. El F1-score es una métrica ampliamente utilizada en la estimación del rendimiento de sistemas de PLN o de recuperación de la información, la cual combina la precisión y el recall o exhaustividad36 y permite hacer comparaciones entre diversas aproximaciones para resolver una tarea en cuestión. Nuestro objetivo es que dicho corpus, que actualmente se encuentra disponible gratuitamente para uso no comercial (https://zenodo.org/record/5591011#.Yt3-K-zMI-R), permita reducir la lista de espera mediante la identificación de casos que, a su vez, pueden ser solucionados mediante otras estrategias tales como telemedicina y la priorización de pacientes mediante análisis de comorbilidades. Otro aspecto importante que queremos abordar en futuras investigaciones y colaboraciones con el Ministerio de Salud es la identificación de procedimientos pendientes.
Codificación automática
Una de las dificultades en el análisis de texto no estructurado radica en la variabilidad del lenguaje que puede ser empleado para referirse a una misma entidad; específicamente, casos de ambigüedad en la asignación de diferentes etiquetas lingüísticas para el mismo sentido. Por ejemplo, las expresiones «cáncer de mama», «tumor mamario», «ca. mama», «tu. mamario», «carcinoma mamario», pueden ser empleadas, indistintamente, para hacer alusión a la misma enfermedad. Una de las soluciones para reducir esta variabilidad es la normalización de los términos médicos, mediante la asignación de, por ejemplo, su código correspondiente desde los conceptos del Sistema de Lenguaje Médico Unificado (UMLS por su sigla en inglés).
El UMLS Metathesaurus, desarrollado por la Biblioteca Nacional de Medicina de los Estados Unidos37, es una herramienta creada principalmente para resolver dos barreras importantes frente a la capacidad de las máquinas para extraer información: la variedad de nombres para referirse al mismo concepto, como ya lo hemos mencionado, y la ausencia de un formato establecido para distribuir terminologías. Esta herramienta contiene un compilado de nombres, relaciones e información asociada de una variedad de sistemas biomédicos que integra más de dos millones de nombres de aproximadamente 900.000 conceptos de vocabulario médico, y no sólo eso, sino que también posee más de 12 millones de relaciones entre todos estos conceptos.
Actualmente, contamos con una fracción de 2.000 interconsultas médicas anotadas en el Corpus de Lista de Espera Chilena, cuyas entidades médicas fueron normalizadas de forma automatizada empleando el léxico MedLexSp38, 39 asignándole uno o múltiples códigos únicos de identificación a cada entidad. Este recurso estará disponible para su libre uso próximamente.
Otros sistemas de PLN clínico desarrollados en Chile
El Plan Nacional del Cáncer 2018-2028 en Chile propone una línea estratégica para fortalecer los Sistemas de Registro, Información y Vigilancia del Cáncer. A partir de lo anterior, el PLN puede apoyar la extracción automática de información y la sistematización de bases de datos para este fin. Recientemente, hemos desarrollado un sistema de apoyo que facilita la codificación CIE-O de la morfología y topografía de los tumores en los informes de patología, una tarea esencial para los registros de cáncer40. Este sistema puede probarse en el siguiente enlace: https://topomorfo.oncodata.org.
Otro de nuestros desarrollos en esta área permite detectar automáticamente menciones de metástasis a distancia en reportes de imagenología. Específicamente, la clasificación TNM de tumores aporta información relevante en la definición de un estadio, definiendo las características del tumor primario (T), la posible propagación a ganglios linfáticos cercanos (N) y la presencia o ausencia de metástasis a otras partes del cuerpo (M). Como antecedente, es relevante considerar que la detección manual de los parámetros TNM consume mucho tiempo y horas de personal, pues implica el análisis individual de cada reporte. Tanto este desarrollo como el anterior se enmarcan en una colaboración con la Fundación Arturo López Pérez, considerado como el mayor centro oncológico de Chile, que atiende a más de 50 mil pacientes al año, la mayoría pertenecientes al sistema público de salud41.
Un algoritmo de clasificación que emplea técnicas de PLN e historias clínicas anonimizadas de un hospital de Chile fue desarrollado por Ramos et al.42 para clasificar los diagnósticos de los pacientes discriminando entre las clases ‘cáncer’ frente a ‘no cáncer’ y ‘cáncer de mama’ frente a ‘otro cáncer’. Este algoritmo podría utilizarse como herramienta de apoyo y recomendación del diagnóstico de los pacientes, principalmente para médicos que inician sus labores en sectores alejados con poco personal especializado como, por ejemplo, hospitales rurales.
Existen varios trabajos enfocados en el uso secundario de datos médicos y clasificación de textos desarrollados por investigadores de la Universidad de Concepción43, 44, 45. En esta propuesta se emplearon textos clínicos en español, provenientes del Hospital Clínico Regional Dr. Guillermo Grant Benavente de Concepción, para identificar y extraer información sobre el estado de tabaquismo de los pacientes, mediante técnicas de PLN y minería textual46, junto con información sobre medidas de peso corporal y comorbilidades47. En ambos casos, los datos extraídos fueron posteriormente utilizados como corpus para algoritmos de clasificación de textos. A partir de esta implementación, por ejemplo, fue posible determinar si un paciente era fumador, no fumador, fumador actual o fumador pasado.
Lecaros et al.48, por su parte, examinaron derivaciones contenidas en el corpus de la lista de espera para detectar casos de pacientes con psoriasis y, así, determinar la incidencia de dicha patología en Chile. Recientemente, Figueroa-Barra et al.49 emplearon técnicas de PLN dentro de un sistema de análisis automático del lenguaje para identificar y predecir esquizofrenia en el primer episodio de psicosis. Este sistema se empleó en entrevistas clínicas en español y se basa en el análisis de 30 rasgos lingüísticos que permiten distinguir los controles sanos de los pacientes con esquizofrenia crónica, y predecir el diagnóstico de esquizofrenia en pacientes con primer episodio de psicosis.
Conclusiones
El texto clínico representa una proporción importante de la información registrada de los y las pacientes. Su procesamiento masivo y utilización en la toma de decisiones requiere herramientas del estado del arte en PLN; correspondiente a una de las tres ramas de la inteligencia artificial, junto con la visión por computador y la robótica.
Un área que hemos comenzado a desarrollar es el estudio de las propiedades lingüísticas de distintos textos clínicos, lo que es relevante para desarrollar sistemas en diversas especialidades médicas. En efecto, el reconocimiento de patrones lingüísticos presenta implicaciones significativas en términos de la generalización, ya que las herramientas de PLN suelen adaptarse de manera eficiente a dominios estrechos, con lenguajes bien definidos y comprendidos50. Además, nuestro equipo colabora con grupos en España y Argentina, con el fin de aunar esfuerzos para la construcción de recursos lingüísticos que consideren la variedad de estructuras presentes en la lengua española, y de esta manera evaluar sus similitudes o diferencias.
Para avanzar en el uso del PLN en español es necesario obtener narrativas clínicas, frecuentemente anotadas por humanos, así como modelos computacionales que permitan realizar la tarea en cuestión. Ejemplos de tareas relevantes de PLN clínico son: la detección de información clave, la clasificación de textos o la asignación de códigos internacionales. En el presente artículo, mostramos ejemplos de los avances en estas tareas, lideradas por nuestro grupo de investigación y otros grupos nacionales.
Según lo anterior, recibir apoyo político y gubernamental es esencial para que trabajos como estos puedan ser transferidos al sistema de salud chileno. En particular, desde octubre de 2022 hemos comenzado a colaborar con el Departamento de Estadísticas e Información de Salud (DEIS) del Ministerio de Salud. El objetivo de esta colaboración es transferir nuestro conocimiento adquirido en el trabajo con texto libre de la lista de espera no GES para apoyar la gestión de ésta. En particular, nos hemos propuesto codificar todas las enfermedades mencionadas en las razones de interconsultas, y con ello apoyar estrategias tales como la detección de pacientes GES en la lista no GES, el Registro Nacional de Cáncer o la articulación con el Hospital Digital para el uso de la telemedicina.
Un desafío actual en la investigación en PLN clínico es lograr acceso a narrativas médicas preservando la privacidad de los pacientes y las regulaciones cada vez más estrictas en algunos países. Una solución es la creación de un corpus sintético que preserve las propiedades lingüísticas del corpus original, pero cuya información no guarde relación con los pacientes reales51. Este enfoque es muy promisorio, dado que podría corregir las graves brechas de representación de grupos identitarios52.
Varias condiciones deben darse para que nuevas tecnologías como el PLN sean utilizadas tanto en el sistema público como el privado. Obviamente, es necesaria la voluntad política, por un lado, y el acceso a poder de cómputo, por otro. Además, se debe promover una estrategia que permita asegurar un uso ético de la IA. En efecto, la Organización Mundial de la Salud (https://www.who.int/publications/i/item/9789240029200) recomienda seis principios que deben tenerse en cuenta, los que incluyen: no perder la autonomía humana en la toma de decisiones, incluir medidas de mejoramiento continuo, explicabilidad de los modelos, responsabilidad sobre las tareas que está realizando la máquina, garantizar un uso no-discriminatorio y sustentabilidad. En definitiva, el cumplimiento de todas estas condiciones no es una tarea sencilla. No obstante, contar con ellas como meta constituye una ayuda relevante para crear consciencia y promover el trabajo interdisciplinario y el mejoramiento continuo de los modelos.
PLN en medicina es un área transdisciplinaria por definición, en la que confluyen las ciencias de la computación, la medicina y la lingüística. Para lograr un mayor impacto, se requiere del trabajo conjunto de profesionales de la salud y tomadores de decisión que conozcan y valoren los alcances de la inteligencia artificial aplicada al ámbito de la salud.
Declaración de conflicto de interés
Los autores declaran no tener conflictos de intereses.
Financiamiento
Este trabajo ha sido financiado por la ANID a través de los Fondos Basales para Centros de Excelencia FB210005 (Centro de Modelamiento Matemático), Fondecyt de Iniciación 11201250 (J. Dunstan) y Fondecyt de Postdoctorado 3210395 (P. Báez). Además, la investigación conducida por J. Dunstan es apoyada por los Institutos Milenio ICN2021_004 (iHealth) e ICN17_002 (IMFD).
==== Refs
Bibliografía
1 Cómo funciona el Sistema de Salud en Chile Cómo funciona el Sistema de Salud en Chile [Internet]. Orientación en Salud. Superintendencia de Salud, Gobierno de Chile 2022 [citado el 23 de julio de 2022]. Disponible en: http://www., supersalud., gob., cl/difusion/665/w3-article-17328., html.
2 Informe CDD: Caracterización sociodemográfica y socioeconómica en la población asegurada inscrita. Gobierno de Chile. [Internet]. División de Desarrollo Institucional, Departamento de Estudios y Estadísticas; 2020 [citado el 23 de julio de 2022]. Disponible en: https://www.fonasa.cl/sites/fonasa/adjuntos/Informe_caracterizacion_poblacion_asegurada.
3 Estay R. Cuadrado C. Crispi F. González F. Alvarado F. Cabrera N. Desde el conflicto de listas de espera, hacia el fortalecimiento de los prestadores públicos de salud: Una propuesta para Chile Cuad Med Soc. 57 1 2017 49 63
4 Cordero A, Valderrama S, Duran L, Santelices P, Olave M, Jiron P, et al., Norma técnica para el registro de las listas de espera. 2016; Disponible en: https://www.minsal.cl/sites/default/files/files/Nueva%20Norma%20de%20Listas%20de%20Espera%202011.pdf.
5 ¿Qué es el AUGE o GES? - ¿Qué es el AUGE o GES? [Internet]. Preguntas Frecuentes. Superintendencia de Salud, Gobierno de Chile. [citado el 18 de julio de 2022]. Disponible en: http://www.supersalud.gob.cl/consultas/667/w3-article-4605.html.
6 Garantías Explícitas en Salud (GES) [Internet]. Orientación en Salud. Superintendencia de Salud, Gobierno de Chile. [citado el 18 de julio de 2022]. Disponible en: http://www.supersalud.gob.cl/difusion/665/w3-propertyvalue-1962.html.
7 Martinez D.A. Zhang H. Bastias M. Feijoo F. Hinson J. Martinez R. Prolonged wait time is associated with increased mortality for Chilean waiting list patients with non-prioritized conditions BMC Public Health. 19 1 2019 233 doi: 10.1186/s12889-019-6526.-6. 30808318
8 MINSAL. Estado de Situacion Personas Fallecidas en Listas de Espera No-GES y Garantias Retrasadas GES [Internet]. 2017 [citado el 25 de julio de 2022]. Disponible en: https://www.minsal.cl/wp-content/uploads/2018/01/Informe-Final-Comision-Asesora-LE-y-Garantias-Retrasadas-GES-17082017_.pdf.
9 MINSAL. Glosa 06 - Lista de espera no GES y garantías de oportunidad GES retrasadas [Internet]. 2021 [citado el 23 de julio de 2022]. Disponible en: https://www.minsal.cl/wp-content/uploads/2021/05/ORD-1204-DIGERA-Glosa-06.pdf.
10 Villena F. Dunstan J. Obtención automática de palabras clave en textos clínicos: una aplicación de procesamiento del lenguaje natural a datos masivos de sospecha diagnóstica en Chile. [Automatic keyword retrieval from clinical texts: an application of natural language processing to massive data of Chilean suspected diagnosis] Rev Med Chile. 147 10 2019 1229 1238 32186630
11 Dalianis H. Clinical text mining: Secondary use of electronic patient records 2018 Springer Nature
12 Báez P, Villena F, Rojas M, Durán M, Dunstan J. The Chilean Waiting List Corpus: a new resource for clinical named entity recognition in Spanish. In: Proceedings of the 3rd Clinical Natural Language Processing Workshop. Association for Computational Linguistics. 2020. p.291-300. doi 10.18653/v1/2020.clinicalnlp-1.32.
13 Báez P. Bravo-Marquez F. Dunstan J. Rojas M. Villena F. Automatic Extraction of Nested Entities in Clinical Referrals in Spanish ACM Trans Comput Healthc Health. 3 3 2022 1 22 doi: 10.1145/3498.324.
14 Jurafsky D, Martin JH. Speech and Language Processing [Internet]. Third Ed. draft. 2022 [citado el 18 de julio de 2022]. Disponible en: https://web.stanford.edu/∼jurafsky/slp3/ed3book_jan122022.pdf.
15 Periñán Pascual J.C. Mairal Usón R. Bringing Role and Reference Grammar to natural language understanding Proces Leng Nat. 43 2009 265 273
16 Choi R.Y. Coyner A.S. Kalpathy-Cramer J. Chiang M.F. Campbell J.P. Introduction to Machine Learning, Neural Networks, and Deep Learning Transl Vis Sci Technol. 9 2 2020 14 doi: 10.1167/tvst.9.2.14.
17 Gerbino E, Baggia P, Giachin E, Rullent C. Analysis and Evaluation of Spontaneous Speech Utterances in Focused Dialogue Contexts. Proc. of ESCA Workshop. 1995.
18 Savoy J. Gaussier É. Information Retrieval Indurkhya N. Damerau F.J. Handbook of Natural Language Processing 2nd Ed. 2010 Chapman and Hall/CRC
19 Sinhal R.A. Gupta K.O. Machine translation approaches and design aspects IOSR J Comput Eng. 16 1 2014 22 25
20 Dabre R. Chu C. Kunchukuttan A. A Survey of Multilingual Neural Machine Translation ACM Comput Surv. 53 5 2020 99 1-99:38. doi: 10.1145/3406095.
21 Oronoz M. Gojenola K. Pérez A. de Ilarraza A.D. Casillas A. On the creation of a clinical gold standard corpus in Spanish: Mining adverse drug reactions J Biomed Inform. 56 2015 318 332 doi: 10.1016/j.jbi.2015.06.016. 26141794
22 Báez P. Villena F. Zúñiga K. Jones N. Fernández G. Durán M. Construcción de recursos de texto para la identificación automática de información clínica en narrativas no estructuradas. [Construction of text resources for automatic identification of clinical information in unstructured narratives] Rev Med Chil. 149 7 2021 1014 1022 Spanish. doi: 10.4067/s0034-98872021000701014. 34751303
23 Névéol A. Dalianis H. Velupillai S. Savova G. Zweigenbaum P. Clinical Natural Language Processing in languages other than English: opportunities and challenges J Biomed Semantics 9 1 2018 12 doi: 10.1186/s13326-018-0179-8. 29602312
24 Koleck T.A. Dreisbach C. Bourne P.E. Bakken S. Natural language processing of symptoms documented in free-text narratives of electronic health records: a systematic review J Am Med Inform Assoc. 26 4 2019 364 379 doi: 10.1093/jamia/ocy173. 30726935
25 Jagannatha A. Liu F. Liu W. Yu H. Overview of the First Natural Language Processing Challenge for Extracting Medication. Indication, and Adverse Drug Events from Electronic Health Record Notes (MADE 1.0) Drug Saf. 42 1 2019 99 111 doi: 10.1007/s40264-018-0762-z. 30649735
26 Chen L. Gu Y. Ji X. Sun Z. Li H. Gao Y. Huang Y. Extracting medications and associated adverse drug events using a natural language processing system combining knowledge base and deep learning J Am Med Inform Assoc. 27 1 2020 56 64 doi: 10.1093/jamia/ocz141. 31591641
27 Weegar R. Kvist M. Sundström K. Brunak S. Dalianis H. Finding Cervical Cancer Symptoms in Swedish Clinical Text using a Machine Learning Approach and NegEx AMIA Annu Symp Proc 2015 2015 1296 1305 26958270
28 Weegar R. Mining events preceding a cancer diagnosis En IEEE Computer Society 2018 295 296 doi: 10.1109/eScience.2018.00059.
29 Orooji A. Langarizadeh M. Using of natural language processing techniques in suicide research Emerg Sci J. 1 2 2017 89 96 doi: 10.28991/esj-2017-01120.
30 Levis M. Leonard Westgate C. Gui J. Watts B.V. Shiner B. Natural language processing of clinical mental health notes may add predictive value to existing suicide risk models Psychol Med. 2020 1 10 doi: 10.1017/S0033291720000173
31 Henry S. Wang Y. Shen F. Uzuner O. The 2019 National Natural language processing (NLP) Clinical Challenges (n2c2)/Open Health NLP (OHNLP) shared task on clinical concept normalization for clinical records J Am Med Inform Assoc. 27 10 2020 1529 1537 doi: 10.1093/jamia/ocaa106. Erratumin: J Am Med Inform Assoc. 20***21;28(11):2546 32968800
32 Blanco A. Perez-de-Viñaspre O. Pérez A. Casillas A. Boosting ICD multi-label classification of health records with contextual embeddings and label-granularity Comput Methods Programs Biomed. 188 2020 105264 doi: 10.1016/j.cmpb.2019.1052.64. 31851906
33 Dunstan J, Maass A, Tobar F. Una mirada a la era de los datos. 1 ed. Ed. Universitaria SA; 2022. 134 p.
34 Fort K. Collaborative Annotation for Reliable Natural Language Processing: Technical and Sociological Aspects 2016 John Wiley & Sons
35 Pérez A. Weegar R. Casillas A. Gojenola K. Oronoz M. Dalianis H. Semi-supervised medical entity recognition: A study on Spanish and Swedish clinical corpora J Biomed Inform. 71 2017 16 30 doi: 10.1016/j.jbi.2017.05.009. 28526460
36 Goutte C. Gaussier E. A Probabilistic Interpretation of Precision, Recall and F-Score with Implication for Evaluation Losada D.E. Fernández-Luna J.M. Advances in Information Retrieval. ECIR. Lecture Notes in Computer Science, vol 3408 2005 Springer Berlin, Heidelberg doi: 10.1007/978-3-540-31865-1_25.
37 Schuyler P.L. Hole W.T. Tuttle M.S. Sherertz D.D. The UMLS Metathesaurus: representing different views of biomedical concepts Bull Med Libr Assoc. 81 2 1993 217 222 8472007
38 Campillos-Llanos L. Medical Lexicon for Spanish (MedLexSp) 2022 doi: 10.20350/digitalCSIC/1465.6.
39 Campillos-Llanos L. First steps towards building a medical Lexicon for Spanish with linguistic and semantic information In Proceedings of the 18th BioNLP Workshop and Shared Task. doi:10.18653/v1/W19-5017 2019 152 164
40 Villena F, Báez P, Peñafiel S, Rojas M, Paredes I, Dunstan J. Automatic Support System for Tumor Coding in Pathology Reports in Spanish. SSRN Electron J [Internet]. 2021. doi:10.2139/ssrn.3982259. Disponible en: https://www.ssrn.com/abstract=3982259.
41 Memoria de Gestión FALP [Internet]. Fundación Arturo López Pérez. 2017 [citado el 18 de julio de 2022]. Disponible en: https://www.institutoncologicofalp.cl/fundacion/memoria-falp/.
42 Ramos A.A. Allende-Cid H. Taramasco C. Becerra C. Figueroa R.L. Application of Machine Learning and Word Embeddings in the Classification of Cancer Diagnosis Using Patient Anamnesis IEEE Access. 8 2020 106198 106213
43 Flores C.A. Figueroa R.L. Pezoa J.E. Active Learning for Biomedical Text Classification Based on Automatically Generated Regular Expressions IEEE Access. 9 2021 38767 38777 doi: 10.1109/ACCESS. 2021.3064.000.
44 Flores C.A. Figueroa R.L. Pezoa J.E. Zeng-Treitler Q. CREGEX: A Biomedical Text Classifier Based on Automatically Generated Regular Expressions IEEE Access. 8 2020 29270 29280 doi: 10.1109/ACCESS. 2020.2972.205.
45 Flores C.A. Figueroa R.L. Pezoa J.E. FREGEX: A Feature Extraction Method for Biomedical Text Classification using Regular Expressions Annu Int Conf IEEE Eng Med Biol Soc. 2019 2019 6085 6088 doi: 10.1109/EMBC. 2019.8857.471. 31947233
46 Figueroa R.L. Soto D.A. Pino E.J. Identifying and extracting patient smoking status information from clinical narrative texts in Spanish Annu Int Conf IEEE Eng Med Biol Soc. 2014 2014 2710 2713 doi: 10.1109/EMBC. 2014.6944.182. 25570550
47 Figueroa R.L. Flores C.A. Extracting Information from Electronic Medical Records to Identify the Obesity Status of a Patient Based on Comorbidities and Bodyweight Measures J Med Syst. 40 8 2016 191 doi: 10.1007/s10916-016-0548-8. 27402260
48 Lecaros C. Dunstan J. Villena F. Ashcroft D.M. Parisi R. Griffiths C.E.M. The incidence of psoriasis in Chile: an analysis of the National Waiting List Repository Clin Exp Dermatol. 46 7 2021 1262 1269 doi: 10.1111/ced.14713. 33914930
49 Figueroa-Barra A. Del Aguila D. Cerda M. Gaspar P.A. Terissi L.D. Durán M. Automatic language analysis identifies and predicts schizophrenia in first-episode of psychosis Schizophrenia (Heidelb). 8 1 2022 53 doi: 10.1038/s41537-022-0025.9-3. 35853943
50 Zeng Q.T. Redd D. Divita G. Jarad S. Brandt C. Nebeker J.R. Characterizing clinical text and sublanguage: A case study of the VA clinical notes J Health Med Informat S. 2011 S3 doi:10.4172/2157-7420.S3-001.
51 Bannour N. Wajsbürt P. Rance B. Tannier X. Névéol A. Privacy-preserving mimic models for clinical named entity recognition in French J Biomed Inform. 130 2022 104073 doi: 10.1016/j.jbi.2022.1040.73. 35427797
52 Criado-Perez C. La mujer invisible: descubre cómo los datos configuran un mundo hecho por y para los hombres 2020 Seix Barral España
| 0 | PMC9704488 | NO-CC CODE | 2022-12-01 23:19:05 | no | Hematol Transfus Cell Ther. 2022 Oct 15; 44:S369-S370 | latin-1 | Hematol Transfus Cell Ther | 2,022 | 10.1016/j.htct.2022.09.625 | oa_other |
==== Front
Revista Médica Clínica Las Condes
0716-8640
0716-8640
S0716-8640(22)00120-1
10.1016/j.rmclc.2022.10.002
Article
Procesamiento de lenguaje natural para texto clínico en español: el caso de las listas de espera en Chile
Natural language processing for clinical text in Spanish: The case of waiting lists in ChileBáez Pablo a
Arancibia Antonia Paz b
Chaparro Matías Ignacio b
Bucarey Tomás b
Núñez Fredy cd
Dunstan Jocelyn def⁎
a Centro de Informática Médica y Telemedicina, Facultad de Medicina, Universidad de Chile. Santiago, Chile
b Escuela de Medicina, Facultad de Medicina, Universidad de Chile. Santiago, Chile
c Departamento de Ciencias del Lenguaje, Facultad de Letras, Pontificia Universidad Católica de Chile. Santiago, Chile
d Centro de Modelamiento Matemático, Facultad de Ciencias Físicas y Matemáticas, Universidad de Chile. Santiago, Chile
e Iniciativa de Datos & Inteligencia Artificial, Facultad de Ciencias Físicas y Matemáticas, Universidad de Chile. Santiago, Chile
f Instituto Milenio Engineering for Healthcare, ANID, Chile. Santiago, Chile
⁎ Autor para correspondencia.
28 11 2022
November-December 2022
28 11 2022
33 6 576582
30 7 2022
16 10 2022
18 10 2022
.
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Las listas de espera no cubiertas por el Plan de Garantías Explícitas en Salud para nueva consulta de especialidad en Chile se han visto incrementadas por los efectos de la pandemia del coronavirus SARS-CoV-2 (COVID-19). Esto representa un problema debido a la demora en la resolución y priorización de cada caso de derivación al nivel secundario de atención en salud. El objetivo de este artículo es exponer el problema de la lista de espera en el sistema de salud de Chile, y abordarlo como ejemplo de la aplicación de técnicas de Procesamiento del Lenguaje Natural (PLN). Específicamente, se describe una metodología para el reconocimiento de información clave en narrativas médicas. Actualmente, contamos con un conjunto de interconsultas médicas manualmente anotadas en el desarrollo del Corpus de Lista de Espera Chilena, y con una fracción de 2.000 interconsultas en las que las entidades médicas anotadas fueron normalizadas de forma automatizada a los conceptos del Sistema de Lenguaje Médico Unificado empleando el léxico MedLexSp. Este y otros recursos lingüísticos y herramientas de PLN están siendo desarrollados por el grupo de PLN en Medicina del Centro de Modelamiento Matemático de la Universidad de Chile y otros grupos a nivel nacional, los cuales constituyen aportes relevantes que pueden ser transferidos al sistema de salud chileno, con el objetivo de apoyar la gestión del texto clínico en español.
The waiting lists not covered by the Explicit Health Guarantee Plan for new specialty consultation in Chile increased due to the effects of the SARS-CoV-2 coronavirus (COVID-19) pandemic. This represents a problem derived from the delay in the resolution and prioritization of each case. This paper aims to describe the issue of the waiting lists in the Chilean health system and present an example of the application of Natural Language Processing (NLP). Specifically, a methodology for recognizing key information in medical narratives is described. Currently, we have a set of manually annotated medical referrals in the development of the Chilean Waiting List Corpus, with a fraction of 2,000 referrals in which the annotated medical entities were automatically normalized to the Unified Medical Language System concepts using the lexicon MedLexSp. The clinical NLP Group of the Center for Mathematical Modeling of the University of Chile, and other national NLP groups, are developing several tools and resources in medicine that can be transferred to the Chilean health system to support managing clinical text in Spanish.
Palabras clave
Procesamiento del Lenguaje Natural
Informática Médica
Inteligencia Artificial
Listas de Espera
Keywords
Natural Language Processing
Medical Informatics
Artificial Intelligence
Waiting Lists
==== Body
pmcIntroducción
Tal como ocurre en muchos países del continente y el mundo, el sistema de salud en Chile enfrenta problemáticas derivadas del envejecimiento poblacional, la mayor expectativa de vida, la multimorbilidad y el incremento en los servicios de salud, entre otras. Actualmente, diversos aspectos se han agravado debido a la pandemia del coronavirus SARS-CoV-2 (COVID-19), cuyas consecuencias no sólo han impactado negativamente en el ámbito económico, sino también en la salud pública del país. A la par, estamos experimentando una transformación digital sin precedentes, cuyo núcleo es la aplicación de la inteligencia artificial (en adelante IA) para generar herramientas que contribuyan a desarrollar un sistema de salud más eficiente, que cubra las necesidades de los pacientes con soluciones costo-efectivas.
Un problema de salud pública que se ha acrecentado significativamente por la pandemia y que puede ser abordado con el apoyo de la IA es el de las listas de espera. En particular, y tal como se explicará en detalle en las siguientes secciones, la razón de interconsulta en la lista de espera para nueva consulta de especialidad en hospitales públicos chilenos está escrita en texto no estructurado, un problema que ha motivado por varios años nuestro trabajo como grupo de Procesamiento de Lenguaje Natural en Medicina del Centro de Modelamiento Matemático de la Universidad de Chile (http://pln.cmm.uchile.cl).
En este artículo describimos el problema de las listas de espera en Chile y lo presentamos como ejemplo de la aplicación de procesamiento de lenguaje natural (en adelante PLN). Nos centraremos en cómo esta rama de la inteligencia artificial puede proporcionar soluciones a problemas de salud pública, y mostraremos otras tareas de PLN en el contexto clínico chileno.
El sistema de salud chileno y las listas de espera
El sistema de salud en Chile se compone de un sistema mixto de atención integrado por el seguro público, el Fondo Nacional de Salud (en adelante FONASA), y uno privado formado por Instituciones de Salud Previsional (ISAPRES)1. Es de interés el sistema público, ya que al año 2019 concentraba el 78% de la población2.
En el sistema público no se puede acceder directamente a una atención con un especialista. Cuando, a juicio del equipo de salud, se requiere la opinión de éste, se emite una solicitud de interconsulta junto con una derivación al nivel secundario de atención en salud. Es en estos centros donde se solicitan exámenes, se definen tratamientos y seguimiento. Ante la imposibilidad de atender esta consulta inmediatamente, surge el concepto de lista de espera3. En términos generales, las listas de espera pueden ser definidas desde dos perspectivas:• Genéricamente, se definen como un conjunto de personas que, en un momento dado, se encuentran en espera de ser atendidas para una consulta de especialidad médica u odontológica, para un procedimiento o prueba diagnóstica o para una intervención quirúrgica programada, solicitada por un profesional médico u odontológico autorizado en la red de salud y teniendo documentada tal petición.
• Operacionalmente, se entienden como el universo de registros que no poseen una causal de salida y aquellos identificados por la causal de salida N°3, es decir, aquellos pacientes que se encuentran en re-evaluación para el diagnóstico y por ende, no salen de las listas de espera4.
Las Garantías Explícitas en Salud (en adelante, GES) corresponden a los beneficios, garantizados por ley, para la población afiliada tanto a las ISAPRES como a FONASA5, correspondiente a un conjunto de 87 patologías. Algunos ejemplos de las patologías cubiertas por el GES son: enfermedad renal crónica etapa 4 y 5, diabetes mellitus tipo I y II, e hipertensión arterial6.
Las garantías exigibles para las patologías GES son:
Acceso: es el derecho a recibir las atenciones correspondientes, según la Ley, para cada patología listada.
Oportunidad: tiempos máximos de espera para el otorgamiento de las prestaciones.
Protección financiera: la persona beneficiaria cancelará un porcentaje de la afiliación.
Calidad: otorgamiento de las prestaciones por un prestador acreditado o certificado5.
El problema que presenta la lista de espera radica en la demora de resolución y priorización de cada caso, más que en el número de personas que conforman dicha lista7. Específicamente, durante el año 2016 fallecieron 16.625 personas esperando ser atendidas por un médico especialista para patologías no cubiertas por el GES, mientras que, en el mismo año, 993 personas fallecieron esperando en una lista de espera GES. Lo anterior se estableció combinando tanto las listas de espera para consultas de especialidad como la lista quirúrgica8. Así, al 31 de marzo del 2021, el número de personas en lista de espera para patologías no GES ascendía a 1.932.4229. En consecuencia, tanto la alta mortalidad de la lista de espera no GES, como el tiempo que los pacientes esperan para ser atendidos, nos ha motivado como grupo a estudiar esta lista de espera en particular.
Esta situación es preocupante para la población chilena y las autoridades, por lo que deben buscarse medidas para revertir este tiempo prolongado de espera, junto con sus desfavorables consecuencias. Para abordar este problema, es importante identificar la dinámica en la que se registra la información. El médico tratante primero debe ingresar la información clínica del paciente a una plataforma de ficha clínica electrónica, la que contiene campos tanto de información estructurada como no estructurada, en donde se justifica la solicitud. La información estructurada puede ser categórica o numérica, mientras que la no estructurada corresponde a información de tipo imágenes, señales o texto10. Del total de la información, un 40% corresponde a las denominadas ‘narrativas clínicas’ o texto libre11.
La información estructurada es de fácil análisis y comprensión, lo que posibilita generar documentos equivalentes y acotados, pero puede omitir información relevante respecto al proceso diagnóstico. Por otro lado, el texto libre podría contener faltas de ortografía, uso de abreviaciones, jerga local, entre otros, lo que dificulta su análisis. Su contenido, sin embargo, es de gran relevancia, ya que las narrativas contienen información detallada sobre enfermedades, su sintomatología y el fundamento clínico escrito por profesionales de la salud.
Al analizar las interconsultas que forman parte de la lista de espera no GES, una contribución a la resolución del problema podría ser la implementación de criterios de prioridad, la realización de estadísticas de la derivación, la determinación de los casos que pueden ser resueltos por medio de telemedicina, la identificación de factores de riesgo y, finalmente, la búsqueda de relaciones con los antecedentes familiares, entre otros. Todas estas tareas pueden ser agilizadas al utilizar técnicas de PLN12, 13.
Procesamiento de lenguaje natural (PLN)
El PLN es una rama de la inteligencia artificial que puede ser utilizada para el análisis del texto y discurso producido por humanos14. Específicamente, el PLN tiene como objetivos estudiar, diseñar y aplicar sistemas informáticos que faciliten la comunicación entre personas y, entre personas y máquinas15, 16. En otras palabras, se persigue imitar artificialmente algunos de los aspectos de la capacidad humana para el lenguaje, lo que se traduce en procesos de producción y comprensión.
En definitiva, el PLN busca entregar soluciones a los problemas concretos derivados del intento por reproducir artificialmente los patrones con los que funcionan la mente y el lenguaje humanos, para transferirlos a la relación entre humanos y máquinas. Algunas de las tareas tradicionales que aborda el PLN son las siguientes:1. Desarrollo de sistemas de diálogo: sistemas conversacionales que tienen por objetivo optimizar la precisión del proceso comunicativo, ya sea por medio de un chat o del reconocimiento acústico-fonético por parte de la máquina17.
2. Extracción y recuperación de información: ámbito de investigación dedicado al desarrollo de procedimientos informáticos para encontrar documentos que contienen datos textuales no estructurados (es decir, que carecen de una estructura semánticamente abierta y fácil de usar por una máquina)18.
3. Traducción automática o semiautomática: conjunto de técnicas para el desarrollo de programas especializados que buscan traducir datos textuales desde una lengua origen hacia una lengua meta, a partir de implementaciones estadísticas19 o neuronales20.
En cuanto a la aplicación del PLN en el ámbito de la medicina, algunas de las herramientas de mayor utilidad sirven para la extracción automática de información en documentos electrónicos, como lo son las fichas clínicas21. Una gran proporción de la información contenida en las fichas clínicas es no estructurada, lo que hace que su análisis sea de mayor dificultad pues no puede ser fácilmente resumida22. El procesamiento de las fichas clínicas permite asistir a los profesionales de la salud en estudios retrospectivos, en la toma de decisiones clínicas, obtención y comparación de datos sobre enfermedades poco frecuentes, creación de estadísticas e identificación de factores de riesgo23, así como también en la evaluación de eficiencia y costos del cuidado de la salud24.
Otras aplicaciones de PLN en medicina son: detección de reacciones adversas a medicamentos25, 26, detección de eventos y síntomas que preceden al diagnóstico de cáncer27, 28, evaluación del riesgo de suicidio en personas con depresión29, 30, asociación de comorbilidades y otras aplicaciones que pueden ayudar a elaborar estudios epidemiológicos11. Asimismo, cabe nombrar la utilidad en la asignación automática de códigos CIE-10 (u otras nomenclaturas)31, 32 al texto libre, normalizando diversas expresiones empleadas para describir la misma patología.
Uno de los enfoques más comunes en PLN en el ámbito médico es emplear el texto clínico (fichas clínicas o parte de ellas como reportes de imagenología, anatomía patológica o interconsultas, entre otros) para construir un corpus. Un corpus es un conjunto ordenado de datos o textos que sirve de base a una investigación, y que es representativo de un ámbito o variedad lingüística particular33. Estos corpus pueden enriquecerse con información interpretativa mediante un proceso denominado anotación34. La información contenida en los textos es analizada por una persona con experiencia (anotador), quien identifica y señala la información clave (por ejemplo, las enfermedades que presentan los pacientes). Estos textos anotados pueden ser utilizados para entrenar sistemas de reconocimiento automatizado de dicha información clave11.
Precisamente, la detección de entidades nombradas (NER por su sigla en inglés), consiste en el reconocimiento de clases semánticas en texto libre no estructurado35. La detección de entidades como enfermedades, signos y síntomas o procedimientos, son ejemplos de NER en el contexto médico. Sin embargo, gran parte del trabajo previo está desarrollado en la lengua inglesa, lo que junto a la carencia de recursos lingüísticos para la lengua española en el dominio clínico, supone una dificultad para el análisis de textos en esta lengua22.
Aportes al problema de la lista de espera con PLN
Nuestro grupo construyó el primer corpus de texto clínico chileno de libre acceso, a partir de datos de la lista de espera no cubierta por el Plan GES para nueva consulta de especialidad. Esta muestra está compuesta por 2.592.925 interconsultas integradas por las 50 especialidades médicas y dentales22. Adicionalmente, empleamos 10.000 interconsultas para construir un corpus manualmente anotado con hallazgos clínicos, resultados de test o de laboratorio, signos y síntomas, enfermedades, partes del cuerpo, medicamentos, abreviaciones, miembros de la familia y procedimientos de laboratorio, diagnósticos y de tratamiento12.
El corpus anotado se empleó para entrenar un modelo de NER, el cual permite identificar de forma automática enfermedades, medicamentos y partes del cuerpo, con un F1-score de 82,9%, 84,3% y 86,5%, respectivamente. El F1-score es una métrica ampliamente utilizada en la estimación del rendimiento de sistemas de PLN o de recuperación de la información, la cual combina la precisión y el recall o exhaustividad36 y permite hacer comparaciones entre diversas aproximaciones para resolver una tarea en cuestión. Nuestro objetivo es que dicho corpus, que actualmente se encuentra disponible gratuitamente para uso no comercial (https://zenodo.org/record/5591011#.Yt3-K-zMI-R), permita reducir la lista de espera mediante la identificación de casos que, a su vez, pueden ser solucionados mediante otras estrategias tales como telemedicina y la priorización de pacientes mediante análisis de comorbilidades. Otro aspecto importante que queremos abordar en futuras investigaciones y colaboraciones con el Ministerio de Salud es la identificación de procedimientos pendientes.
Codificación automática
Una de las dificultades en el análisis de texto no estructurado radica en la variabilidad del lenguaje que puede ser empleado para referirse a una misma entidad; específicamente, casos de ambigüedad en la asignación de diferentes etiquetas lingüísticas para el mismo sentido. Por ejemplo, las expresiones «cáncer de mama», «tumor mamario», «ca. mama», «tu. mamario», «carcinoma mamario», pueden ser empleadas, indistintamente, para hacer alusión a la misma enfermedad. Una de las soluciones para reducir esta variabilidad es la normalización de los términos médicos, mediante la asignación de, por ejemplo, su código correspondiente desde los conceptos del Sistema de Lenguaje Médico Unificado (UMLS por su sigla en inglés).
El UMLS Metathesaurus, desarrollado por la Biblioteca Nacional de Medicina de los Estados Unidos37, es una herramienta creada principalmente para resolver dos barreras importantes frente a la capacidad de las máquinas para extraer información: la variedad de nombres para referirse al mismo concepto, como ya lo hemos mencionado, y la ausencia de un formato establecido para distribuir terminologías. Esta herramienta contiene un compilado de nombres, relaciones e información asociada de una variedad de sistemas biomédicos que integra más de dos millones de nombres de aproximadamente 900.000 conceptos de vocabulario médico, y no sólo eso, sino que también posee más de 12 millones de relaciones entre todos estos conceptos.
Actualmente, contamos con una fracción de 2.000 interconsultas médicas anotadas en el Corpus de Lista de Espera Chilena, cuyas entidades médicas fueron normalizadas de forma automatizada empleando el léxico MedLexSp38, 39 asignándole uno o múltiples códigos únicos de identificación a cada entidad. Este recurso estará disponible para su libre uso próximamente.
Otros sistemas de PLN clínico desarrollados en Chile
El Plan Nacional del Cáncer 2018-2028 en Chile propone una línea estratégica para fortalecer los Sistemas de Registro, Información y Vigilancia del Cáncer. A partir de lo anterior, el PLN puede apoyar la extracción automática de información y la sistematización de bases de datos para este fin. Recientemente, hemos desarrollado un sistema de apoyo que facilita la codificación CIE-O de la morfología y topografía de los tumores en los informes de patología, una tarea esencial para los registros de cáncer40. Este sistema puede probarse en el siguiente enlace: https://topomorfo.oncodata.org.
Otro de nuestros desarrollos en esta área permite detectar automáticamente menciones de metástasis a distancia en reportes de imagenología. Específicamente, la clasificación TNM de tumores aporta información relevante en la definición de un estadio, definiendo las características del tumor primario (T), la posible propagación a ganglios linfáticos cercanos (N) y la presencia o ausencia de metástasis a otras partes del cuerpo (M). Como antecedente, es relevante considerar que la detección manual de los parámetros TNM consume mucho tiempo y horas de personal, pues implica el análisis individual de cada reporte. Tanto este desarrollo como el anterior se enmarcan en una colaboración con la Fundación Arturo López Pérez, considerado como el mayor centro oncológico de Chile, que atiende a más de 50 mil pacientes al año, la mayoría pertenecientes al sistema público de salud41.
Un algoritmo de clasificación que emplea técnicas de PLN e historias clínicas anonimizadas de un hospital de Chile fue desarrollado por Ramos et al.42 para clasificar los diagnósticos de los pacientes discriminando entre las clases ‘cáncer’ frente a ‘no cáncer’ y ‘cáncer de mama’ frente a ‘otro cáncer’. Este algoritmo podría utilizarse como herramienta de apoyo y recomendación del diagnóstico de los pacientes, principalmente para médicos que inician sus labores en sectores alejados con poco personal especializado como, por ejemplo, hospitales rurales.
Existen varios trabajos enfocados en el uso secundario de datos médicos y clasificación de textos desarrollados por investigadores de la Universidad de Concepción43, 44, 45. En esta propuesta se emplearon textos clínicos en español, provenientes del Hospital Clínico Regional Dr. Guillermo Grant Benavente de Concepción, para identificar y extraer información sobre el estado de tabaquismo de los pacientes, mediante técnicas de PLN y minería textual46, junto con información sobre medidas de peso corporal y comorbilidades47. En ambos casos, los datos extraídos fueron posteriormente utilizados como corpus para algoritmos de clasificación de textos. A partir de esta implementación, por ejemplo, fue posible determinar si un paciente era fumador, no fumador, fumador actual o fumador pasado.
Lecaros et al.48, por su parte, examinaron derivaciones contenidas en el corpus de la lista de espera para detectar casos de pacientes con psoriasis y, así, determinar la incidencia de dicha patología en Chile. Recientemente, Figueroa-Barra et al.49 emplearon técnicas de PLN dentro de un sistema de análisis automático del lenguaje para identificar y predecir esquizofrenia en el primer episodio de psicosis. Este sistema se empleó en entrevistas clínicas en español y se basa en el análisis de 30 rasgos lingüísticos que permiten distinguir los controles sanos de los pacientes con esquizofrenia crónica, y predecir el diagnóstico de esquizofrenia en pacientes con primer episodio de psicosis.
Conclusiones
El texto clínico representa una proporción importante de la información registrada de los y las pacientes. Su procesamiento masivo y utilización en la toma de decisiones requiere herramientas del estado del arte en PLN; correspondiente a una de las tres ramas de la inteligencia artificial, junto con la visión por computador y la robótica.
Un área que hemos comenzado a desarrollar es el estudio de las propiedades lingüísticas de distintos textos clínicos, lo que es relevante para desarrollar sistemas en diversas especialidades médicas. En efecto, el reconocimiento de patrones lingüísticos presenta implicaciones significativas en términos de la generalización, ya que las herramientas de PLN suelen adaptarse de manera eficiente a dominios estrechos, con lenguajes bien definidos y comprendidos50. Además, nuestro equipo colabora con grupos en España y Argentina, con el fin de aunar esfuerzos para la construcción de recursos lingüísticos que consideren la variedad de estructuras presentes en la lengua española, y de esta manera evaluar sus similitudes o diferencias.
Para avanzar en el uso del PLN en español es necesario obtener narrativas clínicas, frecuentemente anotadas por humanos, así como modelos computacionales que permitan realizar la tarea en cuestión. Ejemplos de tareas relevantes de PLN clínico son: la detección de información clave, la clasificación de textos o la asignación de códigos internacionales. En el presente artículo, mostramos ejemplos de los avances en estas tareas, lideradas por nuestro grupo de investigación y otros grupos nacionales.
Según lo anterior, recibir apoyo político y gubernamental es esencial para que trabajos como estos puedan ser transferidos al sistema de salud chileno. En particular, desde octubre de 2022 hemos comenzado a colaborar con el Departamento de Estadísticas e Información de Salud (DEIS) del Ministerio de Salud. El objetivo de esta colaboración es transferir nuestro conocimiento adquirido en el trabajo con texto libre de la lista de espera no GES para apoyar la gestión de ésta. En particular, nos hemos propuesto codificar todas las enfermedades mencionadas en las razones de interconsultas, y con ello apoyar estrategias tales como la detección de pacientes GES en la lista no GES, el Registro Nacional de Cáncer o la articulación con el Hospital Digital para el uso de la telemedicina.
Un desafío actual en la investigación en PLN clínico es lograr acceso a narrativas médicas preservando la privacidad de los pacientes y las regulaciones cada vez más estrictas en algunos países. Una solución es la creación de un corpus sintético que preserve las propiedades lingüísticas del corpus original, pero cuya información no guarde relación con los pacientes reales51. Este enfoque es muy promisorio, dado que podría corregir las graves brechas de representación de grupos identitarios52.
Varias condiciones deben darse para que nuevas tecnologías como el PLN sean utilizadas tanto en el sistema público como el privado. Obviamente, es necesaria la voluntad política, por un lado, y el acceso a poder de cómputo, por otro. Además, se debe promover una estrategia que permita asegurar un uso ético de la IA. En efecto, la Organización Mundial de la Salud (https://www.who.int/publications/i/item/9789240029200) recomienda seis principios que deben tenerse en cuenta, los que incluyen: no perder la autonomía humana en la toma de decisiones, incluir medidas de mejoramiento continuo, explicabilidad de los modelos, responsabilidad sobre las tareas que está realizando la máquina, garantizar un uso no-discriminatorio y sustentabilidad. En definitiva, el cumplimiento de todas estas condiciones no es una tarea sencilla. No obstante, contar con ellas como meta constituye una ayuda relevante para crear consciencia y promover el trabajo interdisciplinario y el mejoramiento continuo de los modelos.
PLN en medicina es un área transdisciplinaria por definición, en la que confluyen las ciencias de la computación, la medicina y la lingüística. Para lograr un mayor impacto, se requiere del trabajo conjunto de profesionales de la salud y tomadores de decisión que conozcan y valoren los alcances de la inteligencia artificial aplicada al ámbito de la salud.
Declaración de conflicto de interés
Los autores declaran no tener conflictos de intereses.
Financiamiento
Este trabajo ha sido financiado por la ANID a través de los Fondos Basales para Centros de Excelencia FB210005 (Centro de Modelamiento Matemático), Fondecyt de Iniciación 11201250 (J. Dunstan) y Fondecyt de Postdoctorado 3210395 (P. Báez). Además, la investigación conducida por J. Dunstan es apoyada por los Institutos Milenio ICN2021_004 (iHealth) e ICN17_002 (IMFD).
==== Refs
Bibliografía
1 Cómo funciona el Sistema de Salud en Chile Cómo funciona el Sistema de Salud en Chile [Internet]. Orientación en Salud. Superintendencia de Salud, Gobierno de Chile 2022 [citado el 23 de julio de 2022]. Disponible en: http://www., supersalud., gob., cl/difusion/665/w3-article-17328., html.
2 Informe CDD: Caracterización sociodemográfica y socioeconómica en la población asegurada inscrita. Gobierno de Chile. [Internet]. División de Desarrollo Institucional, Departamento de Estudios y Estadísticas; 2020 [citado el 23 de julio de 2022]. Disponible en: https://www.fonasa.cl/sites/fonasa/adjuntos/Informe_caracterizacion_poblacion_asegurada.
3 Estay R. Cuadrado C. Crispi F. González F. Alvarado F. Cabrera N. Desde el conflicto de listas de espera, hacia el fortalecimiento de los prestadores públicos de salud: Una propuesta para Chile Cuad Med Soc. 57 1 2017 49 63
4 Cordero A, Valderrama S, Duran L, Santelices P, Olave M, Jiron P, et al., Norma técnica para el registro de las listas de espera. 2016; Disponible en: https://www.minsal.cl/sites/default/files/files/Nueva%20Norma%20de%20Listas%20de%20Espera%202011.pdf.
5 ¿Qué es el AUGE o GES? - ¿Qué es el AUGE o GES? [Internet]. Preguntas Frecuentes. Superintendencia de Salud, Gobierno de Chile. [citado el 18 de julio de 2022]. Disponible en: http://www.supersalud.gob.cl/consultas/667/w3-article-4605.html.
6 Garantías Explícitas en Salud (GES) [Internet]. Orientación en Salud. Superintendencia de Salud, Gobierno de Chile. [citado el 18 de julio de 2022]. Disponible en: http://www.supersalud.gob.cl/difusion/665/w3-propertyvalue-1962.html.
7 Martinez D.A. Zhang H. Bastias M. Feijoo F. Hinson J. Martinez R. Prolonged wait time is associated with increased mortality for Chilean waiting list patients with non-prioritized conditions BMC Public Health. 19 1 2019 233 doi: 10.1186/s12889-019-6526.-6. 30808318
8 MINSAL. Estado de Situacion Personas Fallecidas en Listas de Espera No-GES y Garantias Retrasadas GES [Internet]. 2017 [citado el 25 de julio de 2022]. Disponible en: https://www.minsal.cl/wp-content/uploads/2018/01/Informe-Final-Comision-Asesora-LE-y-Garantias-Retrasadas-GES-17082017_.pdf.
9 MINSAL. Glosa 06 - Lista de espera no GES y garantías de oportunidad GES retrasadas [Internet]. 2021 [citado el 23 de julio de 2022]. Disponible en: https://www.minsal.cl/wp-content/uploads/2021/05/ORD-1204-DIGERA-Glosa-06.pdf.
10 Villena F. Dunstan J. Obtención automática de palabras clave en textos clínicos: una aplicación de procesamiento del lenguaje natural a datos masivos de sospecha diagnóstica en Chile. [Automatic keyword retrieval from clinical texts: an application of natural language processing to massive data of Chilean suspected diagnosis] Rev Med Chile. 147 10 2019 1229 1238 32186630
11 Dalianis H. Clinical text mining: Secondary use of electronic patient records 2018 Springer Nature
12 Báez P, Villena F, Rojas M, Durán M, Dunstan J. The Chilean Waiting List Corpus: a new resource for clinical named entity recognition in Spanish. In: Proceedings of the 3rd Clinical Natural Language Processing Workshop. Association for Computational Linguistics. 2020. p.291-300. doi 10.18653/v1/2020.clinicalnlp-1.32.
13 Báez P. Bravo-Marquez F. Dunstan J. Rojas M. Villena F. Automatic Extraction of Nested Entities in Clinical Referrals in Spanish ACM Trans Comput Healthc Health. 3 3 2022 1 22 doi: 10.1145/3498.324.
14 Jurafsky D, Martin JH. Speech and Language Processing [Internet]. Third Ed. draft. 2022 [citado el 18 de julio de 2022]. Disponible en: https://web.stanford.edu/∼jurafsky/slp3/ed3book_jan122022.pdf.
15 Periñán Pascual J.C. Mairal Usón R. Bringing Role and Reference Grammar to natural language understanding Proces Leng Nat. 43 2009 265 273
16 Choi R.Y. Coyner A.S. Kalpathy-Cramer J. Chiang M.F. Campbell J.P. Introduction to Machine Learning, Neural Networks, and Deep Learning Transl Vis Sci Technol. 9 2 2020 14 doi: 10.1167/tvst.9.2.14.
17 Gerbino E, Baggia P, Giachin E, Rullent C. Analysis and Evaluation of Spontaneous Speech Utterances in Focused Dialogue Contexts. Proc. of ESCA Workshop. 1995.
18 Savoy J. Gaussier É. Information Retrieval Indurkhya N. Damerau F.J. Handbook of Natural Language Processing 2nd Ed. 2010 Chapman and Hall/CRC
19 Sinhal R.A. Gupta K.O. Machine translation approaches and design aspects IOSR J Comput Eng. 16 1 2014 22 25
20 Dabre R. Chu C. Kunchukuttan A. A Survey of Multilingual Neural Machine Translation ACM Comput Surv. 53 5 2020 99 1-99:38. doi: 10.1145/3406095.
21 Oronoz M. Gojenola K. Pérez A. de Ilarraza A.D. Casillas A. On the creation of a clinical gold standard corpus in Spanish: Mining adverse drug reactions J Biomed Inform. 56 2015 318 332 doi: 10.1016/j.jbi.2015.06.016. 26141794
22 Báez P. Villena F. Zúñiga K. Jones N. Fernández G. Durán M. Construcción de recursos de texto para la identificación automática de información clínica en narrativas no estructuradas. [Construction of text resources for automatic identification of clinical information in unstructured narratives] Rev Med Chil. 149 7 2021 1014 1022 Spanish. doi: 10.4067/s0034-98872021000701014. 34751303
23 Névéol A. Dalianis H. Velupillai S. Savova G. Zweigenbaum P. Clinical Natural Language Processing in languages other than English: opportunities and challenges J Biomed Semantics 9 1 2018 12 doi: 10.1186/s13326-018-0179-8. 29602312
24 Koleck T.A. Dreisbach C. Bourne P.E. Bakken S. Natural language processing of symptoms documented in free-text narratives of electronic health records: a systematic review J Am Med Inform Assoc. 26 4 2019 364 379 doi: 10.1093/jamia/ocy173. 30726935
25 Jagannatha A. Liu F. Liu W. Yu H. Overview of the First Natural Language Processing Challenge for Extracting Medication. Indication, and Adverse Drug Events from Electronic Health Record Notes (MADE 1.0) Drug Saf. 42 1 2019 99 111 doi: 10.1007/s40264-018-0762-z. 30649735
26 Chen L. Gu Y. Ji X. Sun Z. Li H. Gao Y. Huang Y. Extracting medications and associated adverse drug events using a natural language processing system combining knowledge base and deep learning J Am Med Inform Assoc. 27 1 2020 56 64 doi: 10.1093/jamia/ocz141. 31591641
27 Weegar R. Kvist M. Sundström K. Brunak S. Dalianis H. Finding Cervical Cancer Symptoms in Swedish Clinical Text using a Machine Learning Approach and NegEx AMIA Annu Symp Proc 2015 2015 1296 1305 26958270
28 Weegar R. Mining events preceding a cancer diagnosis En IEEE Computer Society 2018 295 296 doi: 10.1109/eScience.2018.00059.
29 Orooji A. Langarizadeh M. Using of natural language processing techniques in suicide research Emerg Sci J. 1 2 2017 89 96 doi: 10.28991/esj-2017-01120.
30 Levis M. Leonard Westgate C. Gui J. Watts B.V. Shiner B. Natural language processing of clinical mental health notes may add predictive value to existing suicide risk models Psychol Med. 2020 1 10 doi: 10.1017/S0033291720000173
31 Henry S. Wang Y. Shen F. Uzuner O. The 2019 National Natural language processing (NLP) Clinical Challenges (n2c2)/Open Health NLP (OHNLP) shared task on clinical concept normalization for clinical records J Am Med Inform Assoc. 27 10 2020 1529 1537 doi: 10.1093/jamia/ocaa106. Erratumin: J Am Med Inform Assoc. 20***21;28(11):2546 32968800
32 Blanco A. Perez-de-Viñaspre O. Pérez A. Casillas A. Boosting ICD multi-label classification of health records with contextual embeddings and label-granularity Comput Methods Programs Biomed. 188 2020 105264 doi: 10.1016/j.cmpb.2019.1052.64. 31851906
33 Dunstan J, Maass A, Tobar F. Una mirada a la era de los datos. 1 ed. Ed. Universitaria SA; 2022. 134 p.
34 Fort K. Collaborative Annotation for Reliable Natural Language Processing: Technical and Sociological Aspects 2016 John Wiley & Sons
35 Pérez A. Weegar R. Casillas A. Gojenola K. Oronoz M. Dalianis H. Semi-supervised medical entity recognition: A study on Spanish and Swedish clinical corpora J Biomed Inform. 71 2017 16 30 doi: 10.1016/j.jbi.2017.05.009. 28526460
36 Goutte C. Gaussier E. A Probabilistic Interpretation of Precision, Recall and F-Score with Implication for Evaluation Losada D.E. Fernández-Luna J.M. Advances in Information Retrieval. ECIR. Lecture Notes in Computer Science, vol 3408 2005 Springer Berlin, Heidelberg doi: 10.1007/978-3-540-31865-1_25.
37 Schuyler P.L. Hole W.T. Tuttle M.S. Sherertz D.D. The UMLS Metathesaurus: representing different views of biomedical concepts Bull Med Libr Assoc. 81 2 1993 217 222 8472007
38 Campillos-Llanos L. Medical Lexicon for Spanish (MedLexSp) 2022 doi: 10.20350/digitalCSIC/1465.6.
39 Campillos-Llanos L. First steps towards building a medical Lexicon for Spanish with linguistic and semantic information In Proceedings of the 18th BioNLP Workshop and Shared Task. doi:10.18653/v1/W19-5017 2019 152 164
40 Villena F, Báez P, Peñafiel S, Rojas M, Paredes I, Dunstan J. Automatic Support System for Tumor Coding in Pathology Reports in Spanish. SSRN Electron J [Internet]. 2021. doi:10.2139/ssrn.3982259. Disponible en: https://www.ssrn.com/abstract=3982259.
41 Memoria de Gestión FALP [Internet]. Fundación Arturo López Pérez. 2017 [citado el 18 de julio de 2022]. Disponible en: https://www.institutoncologicofalp.cl/fundacion/memoria-falp/.
42 Ramos A.A. Allende-Cid H. Taramasco C. Becerra C. Figueroa R.L. Application of Machine Learning and Word Embeddings in the Classification of Cancer Diagnosis Using Patient Anamnesis IEEE Access. 8 2020 106198 106213
43 Flores C.A. Figueroa R.L. Pezoa J.E. Active Learning for Biomedical Text Classification Based on Automatically Generated Regular Expressions IEEE Access. 9 2021 38767 38777 doi: 10.1109/ACCESS. 2021.3064.000.
44 Flores C.A. Figueroa R.L. Pezoa J.E. Zeng-Treitler Q. CREGEX: A Biomedical Text Classifier Based on Automatically Generated Regular Expressions IEEE Access. 8 2020 29270 29280 doi: 10.1109/ACCESS. 2020.2972.205.
45 Flores C.A. Figueroa R.L. Pezoa J.E. FREGEX: A Feature Extraction Method for Biomedical Text Classification using Regular Expressions Annu Int Conf IEEE Eng Med Biol Soc. 2019 2019 6085 6088 doi: 10.1109/EMBC. 2019.8857.471. 31947233
46 Figueroa R.L. Soto D.A. Pino E.J. Identifying and extracting patient smoking status information from clinical narrative texts in Spanish Annu Int Conf IEEE Eng Med Biol Soc. 2014 2014 2710 2713 doi: 10.1109/EMBC. 2014.6944.182. 25570550
47 Figueroa R.L. Flores C.A. Extracting Information from Electronic Medical Records to Identify the Obesity Status of a Patient Based on Comorbidities and Bodyweight Measures J Med Syst. 40 8 2016 191 doi: 10.1007/s10916-016-0548-8. 27402260
48 Lecaros C. Dunstan J. Villena F. Ashcroft D.M. Parisi R. Griffiths C.E.M. The incidence of psoriasis in Chile: an analysis of the National Waiting List Repository Clin Exp Dermatol. 46 7 2021 1262 1269 doi: 10.1111/ced.14713. 33914930
49 Figueroa-Barra A. Del Aguila D. Cerda M. Gaspar P.A. Terissi L.D. Durán M. Automatic language analysis identifies and predicts schizophrenia in first-episode of psychosis Schizophrenia (Heidelb). 8 1 2022 53 doi: 10.1038/s41537-022-0025.9-3. 35853943
50 Zeng Q.T. Redd D. Divita G. Jarad S. Brandt C. Nebeker J.R. Characterizing clinical text and sublanguage: A case study of the VA clinical notes J Health Med Informat S. 2011 S3 doi:10.4172/2157-7420.S3-001.
51 Bannour N. Wajsbürt P. Rance B. Tannier X. Névéol A. Privacy-preserving mimic models for clinical named entity recognition in French J Biomed Inform. 130 2022 104073 doi: 10.1016/j.jbi.2022.1040.73. 35427797
52 Criado-Perez C. La mujer invisible: descubre cómo los datos configuran un mundo hecho por y para los hombres 2020 Seix Barral España
| 0 | PMC9704489 | NO-CC CODE | 2022-12-01 23:19:05 | no | Hematol Transfus Cell Ther. 2022 Oct 15; 44:S675 | latin-1 | Hematol Transfus Cell Ther | 2,022 | 10.1016/j.htct.2022.09.1159 | oa_other |
==== Front
Hematol Transfus Cell Ther
Hematol Transfus Cell Ther
Hematology, Transfusion and Cell Therapy
2531-1379
2531-1387
Published by Elsevier Editora Ltda.
S2531-1379(22)00784-2
10.1016/j.htct.2022.09.669
Article
PERFIL DOS CANDIDATOS À DOAÇÃO DE SANGUE APÓS PICO A PANDEMIA DA COVID-19
Bento RA
Rodrigues APC
Sessin APC
Pitol ACA
Giacomo JED
Santos JAD
Rossetto DE
Grupo GSH, São Paulo, SP, Brasil
15 10 2022
10 2022
15 10 2022
44 S395S396
Copyright © 2022 Published by Elsevier Editora Ltda.
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Objetivos
Comparar o perfil dos candidatos à doação de sangue do Banco de Sangue São Paulo e Banco de Sangue Santos (Grupo GSH) na pandemia (2021) e após (2022) o pico da pandemia da Covid-19.
Material e método
Os dados para o levantamento do histórico dos candidatos à doação de sangue da unidade do Banco de Sangue de São Paulo, foram extraídos dos períodos de 01/04/2020 a 30/09/2020 e 01/01/2022 à 30/06/2022, a partir do banco de dados do Grupo GSH por meio do Sistema Real Blood (TDSA Sistemas) a fim de comparação dos resultados. A análise do perfil epidemiológico corresponde ao: tipo de doação (espontanea, reposição, autologa), tipo de doador (primeira vez, repetição, esporádico), gênero (masculino e feminino), faixa etária (> 18 anos, 18-29 anos, 30-39 anos, 40-49 anos, 50-59 anos e maiores de 60 anos).
Resultados
Conforme os dois períodos analisados (pandemia e pós pandemia) observou-se: aumento de 89,6% no comparecimento as unidades de doação citadas (12.422 para 23.586); de 98,3% do sexo masculino (6.364 para 12.625); 81% do sexo feminino (6.057 para 10.961); 61,6% quanto ao tipo de doação espontânea (4.501 para 7.278) e de 171% de doação de reposição (5.253 para 14.240); 153,7% quanto ao tipo de doador primeira vez (4.716 para 11.967) e 133,8% de doador esporádico (2.567 para 6.002). Por outro lado, houve uma queda de 9,3% quanto tipo de doador de repetição (5.138 para 5.617); 26,9% (165 para 130) quanto a faixa etária da procura pela doação em candidatos menores de 18 anos e de 17,7% (1.904 para 1.617) em candidatos de 50 a 59 anos.
Discussão
O Brasil passou a viver uma nova realidade exigida pela Covid 19. É dificil determinar quais serão os comportamentos do “novo normal”. Estudiosos e pesquisadores, estimam que grande parte das mudanças provocadas pela surto do Covid-19 podem ser duradouras. No auge da pandemia no ano de 2020 e ainda em 2021 presenciamos um cenário gerado com determinações de flexibilização, afetando diretamente a diminuição do transporte público e com isso a restrição de circulação de pessoas nas ruas, que acabaram atingindo o perfil dos candidatos que buscaram os bancos de sangue para realizarem sua doação. Medidas como doações com horário agendado, distribuição de senhas para evitar aglomerações, espaçamento entre cadeiras para garantir a segurança do doador, foram ações implantadas pelo banco de sangue para garantir a presença de candidatos a doação para manutenção do estoque. Mesmo com a perspectiva de um cenário mais estável, precauções como a continuidade da vacinação e o uso de máscaras ainda permanecem fundamentais para o controle do vírus.
Conclusão
Os bancos de sangue foram atingidos severamente pela proliferação da Covid 19. Entretanto após o pico da pandemia no período analisado (2022), observa-se o aumento das categorias do perfil dos candidatos, esse aumento se deve a criação de novas estratégias para o recrutamento de doadores, que pudessem garantir o abastecimento dos estoques de sangue. Além disso após o nova onda da pandemia (2022), foi possível investirmos mais no aprimoramento das estrategias de mobilização e conscientização da população em relação a importância em se doar sangue.
==== Body
pmc
| 0 | PMC9704490 | NO-CC CODE | 2022-12-01 23:19:05 | no | Hematol Transfus Cell Ther. 2022 Oct 15; 44:S395-S396 | utf-8 | Hematol Transfus Cell Ther | 2,022 | 10.1016/j.htct.2022.09.669 | oa_other |
==== Front
Hematol Transfus Cell Ther
Hematol Transfus Cell Ther
Hematology, Transfusion and Cell Therapy
2531-1379
2531-1387
Published by Elsevier Editora Ltda.
S2531-1379(22)01241-X
10.1016/j.htct.2022.09.1126
Article
ALTERAÇÕES HEMATOLÓGICAS EM PACIENTES COM COVID-19: REVISÃO SISTEMÁTICA E METANÁLISE
Costa LG a
Souza PIM a
Pereira M b
Neto MMS a
a Universidade do Estado da Bahia (UNEB), Salvador, BA, Brasil
b Universidade Federal da Bahia (UFBA), Salvador, BA, Brasil
15 10 2022
10 2022
15 10 2022
44 S655S656
Copyright © 2022 Published by Elsevier Editora Ltda.
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Objetivo
Realizar uma revisão sistemática com metanálise sobre as manifestações hematológicas da COVID-19, comparando as alterações entre os grupos de gravidade clínica: doença leve ou moderada versus doença grave ou crítica.
Material e Métodos
Foi realizada uma revisão sistemática baseada no protocolo PRISMA 2020, nas bases de dados Pubmed, Embase, LILACS e SciElo, usando os seguintes descritores do MeSH: COVID-19 ou SARS-CoV-2; hematological tests; erythrocyte count, red blood cell count, leukocyte count, platelet count, ferritin, coagulopathy, prothrombin time, partial thromboplastin time, c-reactive protein e fibrinogen. Foram excluídos artigos não disponíveis na íntegra, revisões de literatura, revisões sistemáticas, opiniões de experts e artigos que não faziam a comparação dos parâmetros entre os grupos de gravidade. O software Stata versão 14.0 foi utilizado para a análise estatística e foi feito cálculo de risco relativo com Intervalo de Confiança de 95% para avaliar as diferenças entres os grupos. A heterogeneidade foi calculada com o teste do ꭓ2 e o teste do I2. A heterogeneidade significativa foi definida com p<0,10 ou I2 >50%.
Resultados
A pesquisa sistemática identificou um total de 2.682 artigos, sendo que ao final da triagem, 55 foram selecionados para a revisão e 18 para metanálise. Os artigos selecionados arrolaram um total de 13.289 participantes, sendo 10.312 com quadro clínico leve a moderado e 3.977 com quadro clínico grave a crítico. As médias de idade foram de 49,8 anos para participantes com doença leve a moderada e de 61,3 anos para o grupo de doença grave a crítica. As mulheres representaram 44,5% dos indivíduos do grupo com doença leve e 32,94% dos indivíduos do grupo grave. Identificou-se que os valores médios de leucócitos (SMD=0,47; 95% IC 0,24‒0,70; I2=86,2%), neutrófilos (SMD=1.44; 95% IC 0,92‒1,96; I2=0,0%), PCR (SMD=3,98; 95% IC 2,6‒5,80; I2=98,2%), ferritina (SMD=1,13; 95% IC 0,57‒1,69; I2=72,6%), fibrinogênio (SMD=0,55; 95% IC 0,16‒0,93; I2=40,5%) e TP (SMD=0,53; 95% IC 0,24‒0,82; I2=55,0%) foram significativamente mais elevados e a contagem de linfócitos (SMD = -1,25; 95% IC -1,67 ‒ -0,83; I2=95,7%) foi significativamente reduzida no grupo de doença grave.
Discussão
A infecção por SARS-CoV-2 induz uma série de mudanças nos exames laboratoriais e algumas delas podem ser usadas para monitorar a gravidade e prever o prognóstico. Diversos trabalhos demonstraram que casos graves são mais propensos a níveis elevados de leucócitos, neutrófilos, aumento da ferritina e alargamento do TP. Apesar de o presente trabalho não demonstrar associação entre níveis de plaquetas e gravidade da doença, vários estudos sugerem que a trombocitopenia também está significativamente associada à doença grave. Neste estudo também não foram observadas diferenças significativas entre os níveis de D-dímero dos grupos avaliados, mas outros trabalhos apontam que a elevação do D-dímero é um fator de risco independente para morte.
Conclusão
Os indivíduos com COVID-19 grave apresentaram redução da contagem de linfócitos e elevação de leucócitos, neutrófilos, PCR, ferritina, fibrinogênio e TP.
==== Body
pmc
| 0 | PMC9704491 | NO-CC CODE | 2022-12-01 23:19:05 | no | Hematol Transfus Cell Ther. 2022 Oct 15; 44:S655-S656 | utf-8 | Hematol Transfus Cell Ther | 2,022 | 10.1016/j.htct.2022.09.1126 | oa_other |
==== Front
Child Youth Serv Rev
Child Youth Serv Rev
Children and Youth Services Review
0190-7409
0190-7409
Elsevier Ltd.
S0190-7409(22)00381-4
10.1016/j.childyouth.2022.106745
106745
Discussion
The moment is now: Strengthening communities and families for the future of our nation
Rostad Whitney L. a
Ports Katie A. b⁎
Merrick Melissa c
Hughes Laura d
a Independent Scholar, Seattle, WA, USA
b Health Equity Research Applied, Atlanta, GA, USA
c Prevent Child Abuse America, Chicago, IL, USA
d Gusto Partners, LLC, Detroit, MI, USA
⁎ Corresponding author.
28 11 2022
1 2023
28 11 2022
144 106745106745
17 12 2020
28 10 2021
24 11 2022
© 2022 Elsevier Ltd. All rights reserved.
2022
Elsevier Ltd
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
COVID-19 has highlighted the historical lack of investment in the conditions that children need to thrive, and demonstrates how a crisis can exacerbate children’s vulnerability to disease and violence. Exposure to early adversity already affects millions of children across the country and puts them at risk for poor outcomes. With the uncertainty of the pandemic, many more families are struggling and subsequently, more children are at risk for exposure to adversity. Preventing early adversity and promoting the prosperity of our nation requires assuring that all children, regardless of sociodemographic characteristics, have what they need to reach their full health and life potential. Now is the time to address the social and structural conditions that contribute to the inequitable distribution of risk for some families and which contribute to their unequal burden and impacts of adversity, COVID-19, racial injustice, and other health crises. While many look forward to “a return to normal,” returning to normal would be a missed opportunity to learn from our mistakes and ensure a bright future for our nation. We must invest in children and families for the future health of Americans.
Keywords
Adverse childhood experiences
Prevention
Equity
Child well-being
==== Body
pmc1 Introduction
The world is still in the midst of a pandemic. COVID-19 is wreaking havoc across the globe and devastating communities as it spreads. To slow the virus’s spread, countries and states across the U.S. ordered residents to stay home and practice social-distancing while also encouraging and even incentivizing vaccinations. In this sense, we are all experiencing public health in action. Integral to public health, however, is health equity and what a society does collectively to assure the conditions in which all people can be healthy (Institute of Medicine, 1993). As we all experience this public health crisis, we see that disproportionate impacts abound, and children and families are even more vulnerable to the negative conditions that will impact their health now and into the future.
Given the significant stressors parents and caregivers are experiencing in this global pandemic, including social isolation, job loss, and reduced access to concrete supports, many children are at higher risk of experiencing child abuse and neglect (CAN), and other adversities. Unfortunately, as a country, we have never prioritized CAN prevention as a public health imperative. The science of brain development, childhood adversity, and toxic stress demonstrates strong associations between CAN and longer-term health consequences, including changes in physiological development; physical and mental health problems; engagement in health risk behaviors such as smoking and substance use; and premature death (Anda et al., 2006, Bellis et al., 2019, Merrick et al., 2019). Many at increased risk right now are the same children at high risk before the pandemic—children who, because of racism, historical trauma, and other structural or systemic factors, consistently comprise lower income households, lack stable housing, and live in divested communities, and consequently remain the most vulnerable to COVID-19 and the resultant economic fallout. COVID-19 highlights the historical lack of investment in the conditions that children need to thrive, and demonstrates how a crisis can exacerbate children’s vulnerability. While many look forward to “a return to normal,” simply returning to normal would be a missed opportunity to learn from our cross-generational mistakes and ensure a bright future for our nation. Every member of our society, and particularly those who have been privileged by our policies and practices over time, has a role to play in developing or supporting safe, stable, and nurturing environments for all children and families (e.g., Merrick et al., 2019). Heretofore the authors use ‘we’ to refer to the collective action needed to ensure healthy conditions. We must invest in children and families for the future health of our nation, and now is the time!
We must assure the environments, physical and sociopolitical, in which families live support relationships that promote safety, stability, and nurturance. This is particularly important in the current crisis, but the innovations employed to get families through crises today need to be sustained in what will be a long recovery. We must address the social and structural conditions that contribute to the inequitable distribution of risk that some families experience, and which contribute to their unequal burden and disproportionate impacts of CAN, COVID-19, and other health crises.
2 Social determinants of child well-being
The conditions in which individuals live have profound implications for their lifelong health and well-being. Research has long highlighted the connection between structural and societal conditions and exposure to childhood adversity and violence (Klevens & Metzler, 2019). The COVID-19 pandemic has underscored the inequitable distribution of conditions that promote health. Indeed, the pivot of many employees to remote work and home confinement was certainly smoother for some families than others. While all families are experiencing additional stressors in this uncertain time, many blue-collar and service industry workers cannot work remotely and certainly cannot afford not to work during a pandemic regardless of symptoms, exposure, and risk level. In fact, among those who make up the bottom 10 % in earnings, only 31 % have access to paid sick leave (Desilver, 2020), which means that many families have had to choose between disease prevention or forgoing basic concrete needs.
Many households report they struggle with an unexpected $400 expense (Brady, Fullerton, & Cross, 2009). As such, those families finding themselves unexpectedly unemployed do not have the means to support their families during months of economic shutdowns. As we highlight further in the paper, government subsidies rolled out in 2021, but many families had already experienced economic instability for months before payments had been received. The workers at highest risk for job instability before and during the crisis are comprised of women and racial/ethnic minorities who earn less wages than their white, male counterparts, and thus, greater work instability from COVID-19 will further widen the income inequality we see in this country. In addition, daycare closures, schools moving to remote learning, and mandatory quarantines for kids exhibiting systems regardless of testing results have created new child care burdens for mothers and fathers, making it a challenge to hold down a job that does not provide remote options and additional flexibilities. This will contribute to the unequal burden of childhood adversity, as economic instability and parental stress are key risk factors for CAN. Families in deep poverty are also likely to be living in marginalized communities, and measures to mitigate the virus’s spread likely exacerbated their social isolation and lack of access to social supports.
Social determinants of health, such as access to education, nutritious food, and healthcare are crucial to children’s health and well-being. As schools moved to online platforms, the digital divide and inequitable access to education has become even more apparent, as students from low-income and minority families are more likely to report that they cannot complete assignments because they lack reliable access to the internet or a computer (Anderson & Perrin, 2018). Additionally, over 30 million children rely on free or reduced meals at school and many low-income families lack access to nutritious food as they live in neighborhoods without affordable, nutritious foods (i.e., “food deserts”). Pandemic-related labor and distribution shortages have made it even more challenging to get nutritious foods to families who rely on free and reduced lunches and other food assistance programs.
In the midst of a pandemic, the need for universal healthcare could not be more apparent, but unfortunately, over 27 million people in the US remain uninsured (Berchick, Barnett, & Upton, 2019). Healthcare access may be particularly consequential for children of color, as Medicaid expansion has been linked to reduced mortality among Black adolescents (Wherry & Meyer, 2015). Expanded health insurance coverage for low-income children has also been associated with higher levels of education completion (Cohodes, Grossman, Kleiner, & Lovenheim, 2014) and, in adulthood, they pay more in taxes and claim less in tax credits (Brown, Kowalski, & Lurie, 2015). Those without access to internet who cannot utilize telehealth options, and who lack health care coverage are particularly vulnerable during a public health crisis.
Recent events have highlighted the ramifications of racism and continued oppression of communities of color in public health and social crises. Criminal justice policies, particularly regarding drug possession, have led to the mass incarceration of men of color and exacerbated the instability of their children and families. One in every 28 children has a parent behind bars, but for Black children the number is 1 in 9, a rate that has quadrupled in the past 25 years (The Pew Charitable Trusts: Pew Center on the States, 2010). Incarceration can devastate the health and well-being of individuals, their families, their communities, and society at large (Jubitana, 2019). In the current pandemic, outbreaks in prisons have been common and decisions to reduce crowding have resulted in the early release of non-violent offenders. Still, people who have incarceration histories have a harder time acquiring stable housing, employment, and education (London and Myers, 2006, Pager et al., 2009), which will be even more difficult with business closures and surging unemployment.
COVID-19 has hit communities of color and lower SES the hardest, both in terms of illness severity and deaths (Rossen et al., 2021). Similarly, risk for exposure to CAN is unequal. While all children are at risk, conditions like poverty, racism, and historical trauma keep some children and families at greater risk of experiencing CAN and other public health crises. However, these same conditions also put some children and families at greater risk of unnecessary involvement with punitive systems like child protection because poverty is often confused with neglect (Milner & Kelly, 2020). Indeed, while child protection systems should not remove children because of poverty-related concerns alone, the unfortunate reality is that it happens all too often (Dettlaff et al., 2011, Zlotnick, 2009). Given the pandemic has been accompanied by significant economic shocks, the pandemic may not only lead to increased maltreatment by exacerbating parenting stress, more children may be removed simply because more families are living in poverty as a result of the pandemic’s economic fallout. As such, traumatic intervention will not be necessary if we truly transform how we as a society provide supports and services for families before they find themselves in crisis.
The disproportionate burden of violence experienced by some children and families, primarily those of color, without citizenship, and living in low-income households, is increasingly understood as a major contributor to health inequities in the U.S.; health inequities that are contributing to increased death rates in these same families amidst the current public health crisis. Indeed, conditions like poverty and racism are pre-existing conditions that increase risk of various health problems, and that has been painfully clear with COVID-19. As such, addressing the social and structural determinants that influence the conditions in which families are raising children can point to powerful levers of change that can sustain long-term prevention and health promotion. Prevention strategies that address risk and protective factors that contribute to structural conditions (e.g., federal, state, local, and organizational policies) not only have the potential for large impacts on population health, but also require less individual effort. Policies that promote family well-being (e.g., livable wages, universal healthcare, affordable housing, childcare subsidies) are needed, and to do this effectively, we need to change our national narrative around why people, especially children, struggle.
3 Developing new narratives that are supportive of children and families
There remains a gap between what we know about the consequences of social and economic inequality and what evidence-based solutions we choose to implement. This may be the result of dominant narratives regarding poverty and adversity, which tend to focus on blaming individuals for their hardship, and therefore, believing they are undeserving of assistance. Subsequently, policies shown to be effective at lifting people out of poverty are often unpopular and evoke pejoratives such as ‘free handouts’, ‘welfare queens’, and ‘nanny states’. These negative tropes reinforce dominant and misleading narratives that can be used to strategically suppress the political will needed to create supportive environments for children and families.
Narratives are stories rooted in shared values and common themes that influence how individuals and groups process information about circumstances and experiences and subsequently guide decisions and actions (or support for actions) (Jenkins, 2018). Narratives surrounding public health issues like CAN provide information on who and what is responsible for the problem and how to solve it (Niederdeppe, Bu, Borah, Kindig, & Robert, 2008). Ultimately, narratives help people make sense of their world, and they are powerful tools for garnering support for policies. To move toward adopting the policies proposed below, there needs to be a societal shift in our understanding of why people experience adversity, including poverty. This understanding could lead to a shift in our narratives and ultimately, in greater support for equitable, family-friendly policies.
A dominant narrative in American culture is that individual choices determine an individual’s outcomes (i.e., fallacy of pulling yourself up by the bootstraps). In addition to implicit biases, this narrative leads to blaming individuals for their problems and, thus the default solution is one that is focused on fixing “bad” children and parents. The perception that people in poverty are not trying hard enough and therefore unworthy is deeply engrained in the ethos of the US. These narratives perpetuate ideologies that some lives matter more than others in that there are ‘deserving’ and ‘undeserving’ people in poverty, that sharing resources is not possible, or that we are all on our own. These fallacies are, at their root, the logic of white supremacy whereby hyper-individualism, greed, violence, and fear are used to shape harmful narratives that drive ineffective action in unjust systems, and perpetuate stigma on those seeking help, which reduces the generosity of the public and limits effective implementation of policies that address the conditions children and families need for maximal health and prosperity (Wilson, 1987). For example, in the US, the narrative that individuals are entirely responsible for their outcomes results in people viewing social and economic supports as a welfare system whereby individuals are undeserving of assistance if they are not working for pay. As such, “acceptable” policies almost always come in the form of a tax credit where stipulations are placed on the poor in exchange for aid, requiring that those receiving public assistance demonstrate they are using the support to change their behaviors and choices in a way that lifts them out of poverty (e.g., work requirements, drug testing) (Lindsay, 2004). Working against lifting people out of poverty are the too common policy restrictions on accruing assets while receiving support. For example, many economic support policies have a benefits cliff, such that small increases in earnings can push people over the eligibility threshold for benefits even though they still need economic support (National Conference on State Legislatures, 2019). In addition, people must pass a savings or resource test where they demonstrate a lack of most assets before they can receive and continue to receive benefits. Lifting one’s self out of poverty requires the accrual of assets so that the next time they are in need they can use their savings versus social supports. Temporary income supports that penalize people for accruing assets is counterproductive to their intent and renders our current policies largely ineffective at actually lifting people out of poverty for the long term (Wilson, 1987). This is vastly different than other developed nations whereby services and economic supports are seen as a necessity to family and child well-being, and are provided more universally (Lindsay, 2004).
In the 1990 s, several wealthy nations adopted child tax credits (CTC) to address concerns of increasing child poverty (McCabe & Popp Berman, 2016). In Canada, the discussion regarding CTCs centered on children, whereas in the U.S., the distinction between the taxpayer and welfare recipient overshadowed children. As a result, the “logic of tax relief for taxpayers” was adopted in the U.S., in which “families should be exempt from paying taxes until they have enough income to avoid poverty” and the extent of relief was determined by how many dependents a family was supporting (McCabe & Popp Berman, 2016). Within this logic, providing tax relief to the poorest families is perceived as inappropriate because these families do not have tax liability; alternatively, supplementing the income for working families is perceived as inappropriate because they do not need it. Conversely, in Canada and other countries, the “logic of income supplementation for families with children” dominated, wherein “neither wages nor unemployment benefits take into consideration the fact that some people support dependent children while others do not” (McCabe & Popp Berman, 2016). Under this logic, income supplements are perceived as appropriate because they are tied to family size, not employment or tax liability.
Ultimately, new, transformational narratives are needed to garner support for and sustainability of changes in policies and systems outlined in this paper. Public awareness efforts could improve the public’s understanding of the structural policies (e.g., housing, labor, education policies) and processes that lead to living conditions (e.g., poverty, housing instability, incarceration) that create toxic stress for children, how these conditions affect children’s brain architecture and function, and how ultimately these can impact health, education, employment, income, and future generations. Describing the structural circumstances that contribute to adversity can also promote greater compassion for families. Greater compassion could amount to reductions in pathologizing behavioral patterns caused by structural determinants and increased understanding and support for diagnosing and addressing structural determinants of health. Indeed, studies have shown that communication of societal factors can change the public’s understanding of a problem and increase support for social and economic policies (Barry et al., 2013, Bostrom, 2004, Niederdeppe et al., 2014, Young et al., 2016). In addition to public health campaigns (a top-down approach), community organizing (a bottom-up approach) is a strategy that can be effective at building community or collective power to change narratives that support local solutions for social and economic conditions contributing to inequitable burdens of injury and disease (Speer et al., 2020).
The pandemic has changed our perception of a “struggling family,” as millions of parents across the country filed unemployment claims and lost reliable income, and provides an opportunity for a less hyper-individualized narrative around well-being. As COVID-19 changes the landscape for families raising children, shifting from a narrative that emphasizes individual responsibility to one normalizing that all families could find themselves in need is necessary to facilitate the type of transformation needed to respond to the impacts of the crisis and strengthen families in recovery. A narrative that encourages connectedness, empathy, utilization of services, and acknowledges the difficulty of child-rearing in the context of structural and social conditions is needed to reduce not only the stigma of help-seeking, but to change the way we view and support families. This is not an individual problem. This is a societal problem, and as such, all sectors need to come together with a shared vision for the health and well-being of our children, and the future prosperity of our nation. New narratives such as these could increase political will to build a new public child and family well-being system that does not return us to the pre-pandemic status quo.
4 Adoption of policies that assure healthy conditions for families
On a societal level, the pandemic has exposed serious flaws in federal, state, and local policies and highlighted the failure of federal and state governments to adequately strengthen supports for American families before crisis occurs. While not comprehensive, the following section discusses critical buckets of policies that should be considered by policymakers for helping families through the crisis and recovery, and in preparation for future crises.
4.1 Workplace policies
Paid family leave provides working parents the ability to spend time away from work caring for their children and themselves without stress related to unemployment or long periods without pay. The ability to take time off work without fear of losing income can reduce the burden of choosing between health and safety and being able to provide for children. In the case of COVID-19, a parent can stay home and reduce their risk of transmitting the disease. Paid leave policies have demonstrated effectiveness at improving child wellness, reducing family and maternal stress and depression (Association of State and Territorial Health Officials). For example, after California passed paid family leave legislation in 2004, the state saw a reduction in rates of pediatric abuse head trauma, the leading cause of fatal child abuse among young children (Klevens, Luo, Xu, Peterson, & Latzman, 2016). Given stay at home orders and guidelines to quarantine among those exposed to COVID-19, our public health response requires universal access to paid leave.
Other critical policies include predictable work schedules and the availability of quality, affordable childcare so parents can work. Flexible and predictable work schedules allow parents to manage work and caregiving responsibilities, including finding reliable childcare. Childcare is increasingly expensive and unaffordable for many families in this country, and costs will surely increase given the large numbers of childcare centers that were forced to close during the pandemic. Most low-income families rely on programs such as Head Start and childcare subsidies to be able to obtain and afford childcare, which benefit both parents and children. For example, parents who receive childcare subsidies are more likely to work standard hours (Press, Fagan, & Laughlin, 2006), work full time (Marshall, Robeson, Tracy, Frye, & Roberts, 2013), and earn more (Ha, 2009). Moreover, the receipt of childcare subsidies has been linked to fewer investigations of abuse and neglect (Yang, Maguire-Jack, Showalter, Kim, & Slack, 2019). While childcare subsidies have been associated with positive outcomes, parents would not need to rely on them if childcare was more affordable and universally available, as is the case in other wealthy nations with more universal childcare policies (e.g., Germany, Norway).
4.2 Income policies
More than a decade of research has shown that anti-poverty policies (e.g., Earned Income Tax Credit [EITC]) are effective at lifting families out of poverty, and are linked to several well-being outcomes (Averett and Wang, 2016, Klevens et al., 2017, Kovski et al., 2021, Rostad et al., 2020). Increasing the generosity of state-level EITCs, the largest anti-poverty program in the United States, has been associated with decreases in reported neglect, particularly among young children (Kovski et al., 2021). State EITC policies have also been linked to fewer abusive head traumas among young children (Klevens et al., 2017) and reductions in foster care entries (Rostad et al., 2020), but only when they are refundable and benefit families with little to no tax liability.
Anti-poverty tax credit policies like the EITC and child tax credit are contingent on employment and operate so that as income rises, the benefits of the policy diminish. Further, a recent report demonstrates the potential for unintended consequences as low-income families who claimed the EITC were more likely to be audited by the Internal Revenue Service and audits were concentrated in areas with a large Black population (Mock, 2019). While tax credits can be helpful, they are only helpful to those within a limited income range, and even then, there are associated risks for taking advantage of that policy. Economic policies need to go above and beyond tax liability, because stronger economic supports for lower income families would not only keep families out of poverty, but would also allow families to save money for times of crises, and perhaps, the current pandemic would not feel so catastrophic for so many.
Raising minimum wages can improve the financial situation of those in poverty, narrow the poverty gap, and break the poverty threshold (Raissian & Bullinger, 2017), yet the federal minimum wage has remained stagnant since 2009. In addition, minimum wages do not always equate to a livable wage. According to economists, families must make $52 k to $156 k to feel financially stable and enjoy the perks of middle-income (Kochhar, 2018). This requires a basic income that in many cities far exceeds $15 an hour. Still, just a $1 increase in the minimum wage has been linked to reductions in neglect reports, and the effect was strongest for young and school-aged children (Raissian & Bullinger, 2017). Of additional importance is that benefit cliffs have not been adjusted to reflect how much families actually need to survive (National Conference on State Legislatures, 2019). As such, earning $15 an hour keeps families in poverty, while excluding them from taking advantage of social supports like SNAP that they still need. Accordingly, policies may be effectively reducing the number of families in poverty, but less so when it comes to conferring the economic security needed in a public health crisis. As such, income policies like basic income and livable wages in addition to efforts to address benefits cliffs are needed so that families are able to live and save for unknown crises.
4.3 Wealth-building policies
The largest contributor to wealth in the U.S. is owning a home (Mitnik & Grusky, 2015), and families of color are far less likely to own homes than their white counterparts, which, accordingly, mirrors racial disparities in wealth. Black families are more likely to rent and live in public housing in smaller homes that may be overcrowded – a setting that makes families particularly vulnerable during an infectious disease outbreak. To understand the increased risk of COVID infection and death in communities of color, it is important to acknowledge the decades of federal policies that reduced their access to homeownership, pushed them to poor quality housing at the margins of communities, and perpetuate the wealth gap that contributes to the health inequities made salient by the pandemic. Further, housing-related concerns (e.g., related to instability, safety, and insecurity) are a common reason for child welfare involvement and the removal of children from their families (Fowler et al., 2013, Zlotnick, 2009). Opportunities for homeownership may be an important means to helping families obtain housing security and build their financial security as they recover from the crisis. Policies to improve homeownership may include inclusionary zoning and other government investments in building and subsidizing housing akin to efforts post World War II. The New Deal provided opportunities for middle-income white families to take advantage of affordable, suburban homes and allowed them to build wealth to recover from the Great Depression (Rothstein, 2017). Similar efforts could be replicated and made available to those families most vulnerable – racial and ethnic minorities. Homeownership can support stability for children (Fowler & Farrell, 2017; Warren & Font, 2015) and help families grow equity—a necessary step towards generational wealth and economic mobility, and may ultimately reduce inequities in children’s exposure to CAN and other ACEs.
Opportunities to build wealth and move up the socioeconomic ladder may also include child savings accounts (CSAs), which aim to improve financial security and capabilities, as well as educational outcomes, for low- and moderate-income families. Although CSAs vary widely in design, the intent is to establish a savings account in a child’s name with an initial “seed” deposit from a community organization, private institution, or government. CSAs have been associated with greater savings over time (Butrica, Carasso, Steuerle, & Toohey, 2008), better educational outcomes (Elliott & Beverly, 2011), and have the potential to help children build wealth over their lifetime. CSAs have improved parents’ educational expectations for their children, reduced depression symptoms among mothers, and increased likelihood of attending and graduating college (Markoff, Loya, & Santos, 2018).
4.4 Implementing equitable policies
Hundreds of years of federal and state policies have systematically ensured inequality – in income, housing, healthcare, education – and inequitable access to opportunities to better oneself, their families, and future generations. We have designed our systems to penalize poverty, and thereby, race – this is especially salient in child welfare and law enforcement. Neglect is the leading form of child maltreatment (Child Welfare Information Gateway, 2021), but neglect is often used as a guise for poverty (Milner & Kelly, 2020). We understand that neglecting the basic needs of children is harmful; we also understand that removing children from their homes is harmful. And yet we continue to implement solutions that harm children and widen inequalities. Children of color are removed from their homes and communities, and placed in congregate care settings at much higher rates than white children; they also stay longer in those settings, and are less likely to receive the services they need (Roberts, 2002). There is no excuse for a child to wait in the system for more than six months until they find a home, particularly if the only barrier to reunification is a lack of necessary economic and housing supports. At the same time, there is no excuse for a person of color who cannot afford bail to sit in a jail cell for months and even years awaiting trial for a crime that is yet to be adjudicated.
Law enforcement policies that concentrate police in low-income neighborhoods of color, allow stop and frisk, and value law and order above all else have led to the mass incarceration of people of color – greatly diminishing their own access to opportunity upon reentry, and consequently, their children’s. Reentry following incarceration is difficult, but the collateral consequences make it near impossible to better oneself and avoid recidivism. Those convicted of a felony (e.g., marijuana possession) have limited access to employment opportunities as many employers are not willing to hire someone with a criminal record; have limited access to affordable housing as they are not eligible for housing programs, and consequently, may not be able to live with their families upon release; and in most states, do not have the right to vote or the opportunity to participate in civil society (Finzen, 2005). Adding insult to injury, many people of color are convicted of felonies for possession of marijuana, which is now legal in several states and offers a lucrative business opportunity for those without criminal records (Union, 2020).
We must decriminalize poverty. The pandemic and civil unrest have highlighted how socioeconomic and racial disparities—as a result of conditions that inequitably distribute access to opportunity—contribute to gross inequities in burden of disease and violence. As a country, we must intentionally and urgently address segregation and the lack of opportunity for families of color as a result of decades of racist policies. The continued segregation of communities has major implications for the civil unrest of today, as well as the burden of COVID in communities of color. We need policies that help desegregate neighborhoods, narrow health inequities, and address hostile relations between police and communities of color.
Ultimately, we need policies that change the context for families and mitigate the impacts of the current crisis, including policies that reduce risk for CAN, and strategies to address the disproportionate burden of COVID on communities of color. The inevitable economic devastation for thousands of families requires proactive policies that help them recover and rebuild their financial security to ensure the health and well-being of society. Moving forward, we need to sustain innovations that are found to make the default “choice” a healthy-one.
5 Investing in children today for a more prosperous and equitable tomorrow
The U.S. operates under the assumption that tax cuts to corporations and the wealthy will trickle down to benefit every-one else. In reality, wages have remained stagnant, income inequality has grown exponentially, and opportunities for upward mobility for the lower and middle classes are disappearing. We have also seen “deaths of despair” rise, as deaths attributable to alcohol, drug use, and suicide are increasing (Case & Deaton, 2020). These deaths represent a landscape in the U.S. in which access to resources and opportunities are concentrated at the top, while the struggles associated with financial insecurity increasingly undermine health and contribute to premature death. The pandemic has shed a bright light on the downfalls of this type of economy and the policies that uphold it; it is time to declare this is not a viable strategy for a country that aims to ensure the right to “life, liberty, and the pursuit of happiness” to all of its citizens.
The past several months have seen sweeping changes in the policy landscape in the U.S. in response to the pandemic. Indeed, we have never seen this level of generosity in policy and it’s this generosity that has kept many low- and middle-income families from moving (deeper) into poverty. The American Rescue Plan Act (ARP) included major policy changes that greatly benefit families and children, particularly children of color, the most powerful of which were changes to the Child Tax Credit (CTC). The CTC was introduced in 1997 as a nonrefundable tax credit to help offset the costs of raising children, later expanded to be partially refundable for families earning above a certain income threshold (in 2001, $10,000 or more; in 2009, $3,000 or more); in 2018, the amount of the credit was increased from $1000 to $2000 per eligible child, with up to $1400 of the credit refundable. Still, because the credit was partially refundable, it did not reach the families that needed it most as it excluded families with too little earnings to be eligible for the full credit.
The ARP increased the maximum credit to $3600 for every young child (0–6 years) and $3000 for every older child (6–17 years) in the household and made the credit fully refundable so that families with little to no tax liability are eligible for the full credit. These changes to the CTC are projected to cut child poverty by almost half, with larger reductions for Black, Hispanic, and Native American children whose families had previously been excluded from CTC benefits (Center on Poverty and Social Policy, 2021). The ARP CTC also allows families to receive monthly advance payments through December 2021 so they do not have to wait until next year to receive the benefit. Families started to receive payments in July, which have already had significant impacts on their economic well-being, as fewer families reported food insufficiency and difficulty paying their bills after the first CTC payments were issued (U.S. Census Bureau, 2021), and approximately 3.5 million children have been kept out of poverty by both the July and August payments (Parolin & Curran, 2021).
The ARP included several other provisions that have helped families weather the pandemic, including increased funding for housing assistance, the Supplemental Nutrition Assistance Program (SNAP), and childcare. Billions of dollars were allocated to mitigate housing issues related to the pandemic, which included emergency rental assistance, homeowners’ assistance, emergency housing vouchers for families experiencing or at risk for homelessness, and utility assistance. These programs along with state and federal eviction moratoriums are expected to reduce housing insecurity, homelessness, and help families avoid displacement. Given that housing insecurity has been linked to the spread of COVID-19, these programs should also help reduce infection transmission (Benfer et al., 2021). The ARP also included billions of dollars for nutritional assistance, including an extension of the 15 % increase in SNAP benefits and investments in the WIC program, to combat hunger and provide families greater access to healthier eating options. This investment was not only good for families; increases in SNAP benefits also stimulate economic recovery by freeing up families’ income to be spent on goods and services that otherwise would have been spent on food (Rosenbaum, Dean, & Neuberger, 2020). Finally, in response to the childcare crisis the pandemic has laid bare, the ARP included the greatest investment in childcare in decades. Billions of dollars were invested in childcare assistance to increase childcare access and reduce costs for more families, including essential workers, and help childcare providers stay afloat during the pandemic.
It is hard to imagine a more significant policy change in US history than what was achieved with the CTC under the ARP – indeed, making the credit fully refundable and allowing monthly payments mirrors the child allowances in other wealthy nations that have historically been more generous with their income supports for families. Despite the significant benefits of the ARP’s CTC to children and families and its potential to narrow race inequities, these changes are currently set to expire in 2022. While the Biden administration is trying to extend these changes as part of their Build Back Better Plan, it faces considerable roadblocks in a highly polarized political climate. Opponents to CTC permanency argue that the program is too costly and dis-incentivizes work. However, the research does not support these arguments. Recent research estimates that a fully refundable CTC of ARP generosity levels allowing for monthly payments would cost around $100 billion, but produce nearly $800 billion in benefits to society through investments in children (Garfinkel, Sariscsany, Ananat, Collyer, & Wimer, 2021). There is also evidence from Canada that a fully refundable CTC could boost employment, not reduce it (Koebel & Schirle, 2016). Further, it cannot be emphasized enough that this single policy could slash child poverty and drastically narrow race inequities (Center on Poverty and Social Policy, 2021). Accordingly, advocacy efforts should lift up the potential of this investment (in addition to investments in childcare and affordable housing outlined in the Build Back Better Plan) to improve the well-being and health of children, particularly young children and children of color, and reduce their risk for exposure to early adversity, including CAN.
Despite the current political polarization in the U.S., we all have in common the desire for health and well-being for our children, our families, and ourselves. Moreover, people are realizing that the status quo no longer provides a reasonable path towards achieving the 'American Dream'. As such, advocacy for solutions that improve social and economic conditions for families, such as the expanded CTC under ARPA, should capitalize on our commonalities and message across both sides of the aisle. Developing new, positive narratives and disseminating action-oriented messages that highlight evidence-based policies, empirical facts, and uplift lived experiences are powerful advocacy strategies for garnering widespread support for the types of policies that actually work for American families (FrameWorks Institute, 2004, Speer et al., 2020). When people understand the root causes of poverty and adversity, they can better understand population-level solutions, and they are better equipped to vote for solutions and the change we need to see to ensure a healthier and more prosperous tomorrow. Both public health campaigns and community organizing can bolster the collective power needed to change existing, harmful narratives and advocate for local and national solutions that provide the conditions for all children and their families to thrive.
We need to reimagine a new economy that invests early in children’s lives to promote conditions for families that are supportive of health and well-being across the lifespan. In this reimagined economy, universal health promotion is prioritized, and families have what they need to help their children thrive, including economic supports, safe and affordable housing, access to resources, and family friendly work policies. Social distancing can be achievable for all families and will curtail pandemics. CAN will be rare as families will have the necessary resources to meet their children’s basic needs, freeing up bandwidth to attend to children’s social and emotional needs, and parents will have less stress because they will have fewer concerns about their financial security. Promoting supportive conditions for families in their children’s formative years will set children on trajectories to becoming productive adults that contribute to the economy through their employment and ability to purchase goods across the lifespan. And indeed, the research tells us this is so: children who are exposed to less adversity do better in school, take fewer sick days at work, are less likely to be poor, are less likely to develop substance use disorders, and are less likely to develop chronic health conditions that can lead to premature death (Merrick et al., 2019, Metzler et al., 2017). Ultimately, this contributes to a population that is more resilient in the face of pandemics and economic crisis.
Despite the robust science outlining the critical importance of child well-being, the prioritization of primary prevention efforts to address CAN and other adversities have been severely lacking. By almost every indicator of health and well-being, children in the U.S. are faring worse than children in other developed nations, in large part because we do not invest as much in children and families (Edelman, 2016). The gap between what we know and what we choose to do is alarming. If we invested in the prevention of CAN in the same manner that we invested in preventing infectious disease, cancer, or smoking, we could significantly reduce children’s exposure to CAN, and, consequently, also reduce chronic and infectious disease and other harmful outcomes (Merrick et al., 2019). Importantly, the costs of pandemics to child well-being would be minimized. Communities, states, and the federal government must develop a shared vision – to ensure that all children thrive and reach their full potential. All sectors – government, business, philanthropy, community-based organizations, and faith institutions – have a role to play in primary prevention and promoting well-being. Budgets should reflect a priority on upstream prevention and an investment in children. COVID-19 has exposed that the U.S. does not value the well-being of children and families. But it has also provided an unprecedented opportunity to ensure that, in the future, policies reflect a society that understands that it takes a village to raise a healthy child.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
==== Refs
References
American Civil Liberties Union. (2020). A Tale of Two Countries: Racially Targeted Arrests in the Era of Marijuana Reform. Retrieved from https://www.aclu.org/report/tale-two-countries-racially-targeted-arrests-era-marijuana-reform.
Anda R.F. Felitti V.J. Bremner J.D. Walker J.D. Whitfield C. Perry B.D. …Giles W.H. The enduring effects of abuse and related adverse experiences in childhood: A convergence of evidence from neurobiology and epidemiology European Archives of Psychiatry and Clinical Neuroscience 256 3 2006 174 186 10.1007/s00406-005-0624-4 16311898
Anderson M. Perrin A. Nearly one-in-five teens can't always finish their homework because of the digital divide 2018 Pew Research Center Washington, D.C.
Association of State and Territorial Health Officials. Essentials for Childhood Policy Guide. Retrieved from https://www.astho.org/Prevention/Essentials-for-Childhood-Policy-Guide/.
Averett S. Wang Y. Effects of Higher EITC Payments on Children’s Health, Quality of Home Environment, and Noncognitive Skills Public Finance Review 46 4 2016 519 557 10.1177/1091142116654965
Barry C.L. Brescoll V.L. Gollust S.E. Framing childhood obesity: How individualizing the problem affects public support for prevention Political Psychology 34 3 2013 327 349
Bellis M.A. Hughes K. Ford K. Ramos Rodriguez G. Sethi D. Passmore J. Life course health consequences and associated annual costs of adverse childhood experiences across Europe and North America: A systematic review and meta-analysis The Lancet Public Health 4 10 2019 e517 e528 10.1016/s2468-2667(19)30145-8 31492648
Benfer E.A. Vlahov D. Long M.Y. Walker-Wells E. Pottenger J.L. Jr. Gonsalves G. Keene D.E. Eviction, Health Inequity, and the Spread of COVID-19: Housing Policy as a Primary Pandemic Mitigation Strategy Journal of Urban Health 98 1 2021 1 12 10.1007/s11524-020-00502-1
Berchick, E. R., Barnett, J. C., & Upton, R. D. (2019). Health Insurance Coverage in the United States Current Population Reports 2018 D.C Washington
Bostrom, M. (2004). Together for success: Communicating low-wage work as economy, not poverty: Economy that Works Project of the Ford Foundation. Retrieved from http://www.economythatworks.org/reports.htm.
Brady D. Fullerton A.S. Cross J.M. Putting poverty in political context: A multi-level analysis of adult poverty across 18 affluent democracies Social Forces 88 1 2009 217 299
Brown D.W. Kowalski A.E. Lurie I.Z. Medicaid as an investment in children: What is the long-term impact on tax receipts? NBER Working Paper No. 20835 2015
Butrica B.A. Carasso A. Steuerle C.E. Toohey D.J. Children's Savings Accounts: Why Design Matters 2008 Urban Institute Washington, DC
Case A. Deaton A. Deaths of Despair and the Future of Capitalism 2020 Princeton University Press Princeton, NJ
Center on Poverty and Social Policy, Columbia University. (2021). “A Poverty Reduction Analysis of the American Family Act.” Poverty and Social Policy Fact Sheet. https://www.povertycenter.columbia.edu/news-internal/2019/3/5/the-afa-and-child-poverty.
Child Welfare Information Gateway. (2021). Child Maltreatment 2019: Summary of key findings. U.S. Department of Health and Human Services, Administration for Children and Families, Children's Bureau. https://www.childwelfare.gov/ pubs/factsheets/canstats/.
Cohodes S. Grossman D. Kleiner S. Lovenheim M.F. The effect of child health insurance access on schooling: Evidence from public insurance expansions NBER Working Paper No. 20178 2014
Desilver D. As coronavirus spreads, which U.S. workers have paid sick leave - and which don't? 2020 Pew Research Center Washington, D.C. Retrieved from
Dettlaff A.J. Rivaux S.L. Baumann D.J. Fluke J.D. Rycraft J.R. James J. Disentangling substantiation: The influence of race, income, and risk on the substantiation decision in child welfare Children and Youth Services Review 33 9 2011 1630 1637 10.1016/j.childyouth.2011.04.005
Edelman M.W. Ending childhood poverty in America Academic Pediatrics 16 3 Suppl 2016 S6 S7 10.1016/j.acap.2016.02.006 27044703
Elliott W. Beverly S. The role of savings and wealth in reducing 'wilt' between expectations and college attendance Journal of Children & Poverty 17 2 2011 165 185
Finzen M.E. Systems of oppression: The collateral consequences of incarceration and their effects on Black communities Georgetown Journal on Poverty Law & Policy, XI I 2 2005 299 324
Fowler P.J. Farrell A.F. Housing and Child Well Being: Implications for Research, Policy, and Practice American Journal of Community Psychology 60 1–2 2017 3 8 10.1002/ajcp.12154 28792062
Fowler P.J. Henry D.B. Schoeny M. Landsverk J. Chavira D. Taylor J.J. Inadequate Housing Among Families Under Investigation for Child Abuse and Neglect: Prevalence from a National Probability Sample American Journal of Community Psychology 52 1–2 2013 106 114 10.1007/s10464-013-9580-8 23702790
Institute FrameWorks Making the Public Case for Child Abuse and Neglect Prevention: A FrameWorks Message Memo Retrieved from https://www.frameworksinstitute.org/wp-content/uploads/2020/06/MakingthePublicCaseforChildAbuseandNeglectPrevention_2004.pdf 2004
Garfinkel, I., Sariscsany, L., Ananat, E., Collyer, S., & Wimer, C. (2021). The costs and benefits of a child allowance. Poverty & Social Policy Brief (Vol. 5, pp. 1-10): Center on Poverty and Social Policy, Columbia University.
Ha Y. Stability of Child-Care Subsidy Use and Earnings of Low-Income Families Social Service Review 83 4 2009 495 525 10.1086/650352
Institute of Medicine Access to health care in America 1993 National Academies Press Washington, DC
Jenkins, A. (2018). Shifting the narrative: What it takes to reframe the debate for social justice in the US. Retrieved from: https://haasinstitute.berkeley.edu/shifting-narrative.
Jubitana C. Broken hearts and broken homes: How mass incarceration impacts children and families 2019 The Kenan Institute for Ethics at Duke University Durham, NC Retrieved from: Https://kenan.ethics.duke.edu/broken-hearts-and-broken-homes-how-mass-incarceration-impacts-children-and-families-may/
Klevens J. Luo F. Xu L. Peterson C. Latzman N.E. Paid family leave's effect on hospital admissions for pediatric abusive head trauma Injury Prevention 22 6 2016 442 445 10.1136/injuryprev-2015-041702 26869666
Klevens, J., & Metzler, M. (2019). Bringing a Health Equity Perspective to the Prevention of Child Abuse and Neglect. In B. Lonne, D. Scott, D. Higgins & T. Herrenkohl (Eds.), Re-Visioning Public Health Approaches for Protecting Children. Child Maltreatment (Contemporary Issues in Research and Policy), vol 9. Springer, Cham. 10/1007/978-3-030-05858-6_13.
Klevens J. Schmidt B. Luo F. Xu L. Ports K.A. Lee R.D. Effect of the earned income tax credit on hospital admissions for pediatric abusive head trauma, 1995–2013 Public Health Reports 132 4 2017 505 511 10.1177/0033354917710905 28609181
Kochhar R. The American middle class is stable in size, but losing ground financially to upper-income families 2018 Pew Research Center Washington, D.C. Retrieved from https://www.pewresearch.org/fact-tank/2018/09/06/the-american-middle-class-is-stable-in-size-but-losing-ground-financially-to-upper-income-families/
Koebel K. Schirle T. The differential impact of universal child benefits on the labour supply of married and single mothers Canadian Public Policy 42 1 2016 49 64
Kovski N.L. Hill H.D. Mooney S.J. Rivara F.P. Morgan E.R. Rowhani-Rahbar A. Association of State-Level Earned Income Tax Credits With Rates of Reported Child Maltreatment, 2004–2017 Child Maltreatment 1077559520987302 2021 10.1177/1077559520987302
Lindsay D. The Welfare of Children 2004 Oxford University Press New York
London A.S. Myers N.A. Race, incarceration, and health: A life-course approach Research on Aging 28 3 2006 409 422 10.1177/0164027505285849
Markoff S. Loya R. Santos J. Quick Guide to CSA Research: An Overview of Evidence on Children's Savings Accounts 2018 Prosperity Now Washington, D.C.
Marshall N.L. Robeson W.W. Tracy A.J. Frye A. Roberts J. Subsidized child care, maternal employment and access to quality, affordable child care Early Childhood Research Quarterly 28 4 2013 808 819
McCabe J.T. Popp Berman E. American exceptionalism revisited: Tax relief, poverty reduction, and the politics of Child Tax Credits Sociological Science 3 24 2016 540 567
Merrick M.T. Ford D.C. Ports K.A. Guinn A. Chen J. Klevens J. …Mercy J.A. Vital Signs: Estimated Proportion of Adult Health Problems Attributable to Adverse Childhood Experiences and Implications for Prevention – 25 States, 2015–2017 Morbidity and Mortality Weekly Report (MMWR) 68 2019 999 1005 31697656
Metzler M. Merrick M.T. Klevens J. Ports K.A. Ford D.C. Adverse childhood experiences and life opportunities: Shifting the narrative Children and Youth Services Review 72 2017 141 149 10.1016/j.childyouth.2016.10.021
Milner J. Kelly D. It’s Time to Stop Confusing Poverty with Neglect Retrieved from Children’s Bureau Express 20 10 2020 https://cbexpress.acf.hhs.gov/index.cfm?event=website.viewArticles&issueid=212§ionid=2&articleid=5474
Mitnik P.A. Grusky D.B. Economic Mobility in the United States 2015 The PEW Charitable Trusts Washington, D.C.
Mock B. IRS Audits Target the Wealth-Depleted Southern Black Belt 2019 Retrieved from Bloomberg CityLab https://www.bloomberg.com/news/articles/2019-04-03/map-reveals-more-irs-audits-in-southern-black-belt
National Conference on State Legislatures Moving on up: Helping families climb the economic ladder by addressing benefits cliffs Retrieved from https://www.ncsl.org/Portals/1/Documents/cyf/Benefits-Cliffs_v03_web.pdf 2019
Niederdeppe J. Bu Q.L. Borah P. Kindig D.A. Robert S.A. Message design strategies to raise public awareness of social determinants of health and population health disparities. [Research Support, Non-U.S. Gov't Review] Milbank Quarterly 86 3 2008 481 513 10.1111/j.1468-0009.2008.00530.x 18798887
Niederdeppe J. Shapiro M.A. Kim H.K. Bartolo D. Porticella N. Narrative persuasion, causality, complex integration, and support for obesity policy Health Communication 29 5 2014 431 444 23829316
Pager D. Western B. Sugie N. Sequencing disadvantage: Barriers to employment facing young black and white men with criminal records Annals of the American Academy of Political and Social Sciences 623 May 2009 195 213
Parolin Z. Curran M.A. Expanded Child Tax Credit Leads to Further Decline in Child Poverty in August 2021 2021 Center on Poverty and Social Policy, Columbia University Poverty and Social Policy Fact Sheet Access at: https://static1.squarespace.com/static/5743308460b5e922a25a6dc7/t/6152076712fd4d05c0e6e5c7/1632765893051/Monthly-poverty-with-CTC-August-CPSP-2021.pdf
Pew Center on the States Collateral Costs: Incarceration's Effect on Economic Mobility 2010 The Pew Charitable Trusts Washington, DC
Press J.E. Fagan J. Laughlin L. Taking pressure off families: Child-care subsidies lessen mothers' work-hour problems Journal of Marriage and Family 68 1 2006 155 171
Raissian K.M. Bullinger L.R. Money matters: Does the minimum wage affect child maltreatment rates? Children and Youth Services Review 72 2017 60 70 10.1016/j.childyouth.2016.09.033
Roberts D. Shattered Bonds: The Color of Child Welfare 2002 Basic Civitas Books New York, NY
Rosenbaum D. Dean S. Neuberger Z. The Case for Boosting SNAP Benefits in Next Major Economic Response Package 2020 Retrieved from Center on Budget and Policy Priorities https://www.cbpp.org/research/food-assistance/the-case-for-boosting-snap-benefits-in-next-major-economic-response
Rossen, L. M., Ahmad, F. B., Anderson, R. N., Branum, A. M., Du, C., Krumholz, H. M., . . . Faust, J. S. (2021). Disparities in Excess Mortality Associated with COVID-19 - United States, 2020. Morbidity and Mortality Weekly Report (MMWR), 70(33), 1114-1119. 10.15585/mmwr.mm7033a2.
Rostad W.L. Ports K.A. Tang S. Klevens J. Reducing the Number of Children Entering Foster Care: Effects of State Earned Income Tax Credits Child Maltreatment 25 4 2020 393 397 10.1177/1077559519900922 31973550
Rothstein R. The Color of Law: A Forgotten History of How Our Government Segregated America 2017 Liveright Publishing Corporation New York
Speer, P. W., Gupta, J., & Haapanen, K. (2020). Developing Community Power for Health Equity: A Landscape Analysis of Current Research and Theory. White Paper. Retrieved from https://static1.squarespace.com/static/5ee2c6c3c085f746bd33f80e/t/5f89f1325e27a51436c97b74/1602875699695/Landscape+-+Developing+Community+Power+for+Health+Equity+%281%29.pdf.U.S. Census Bureau (2021). Household Pulse Survey Collected Responses Just Before and Just After the Arrival of the First CTC Checks. Retrieved from https://www.census.gov/library/stories/2021/08/economic-hardship-declined-in-households-with-children-as-child-tax-credit-payments-arrived.html.Warren, E. J., & Font, S. A. (2015). Housing insecurity, maternal stress, and child maltreatment: An application of the family stress model. Social Service Review, 89(1), 9-39.
Wherry L.R. Meyer B.D. Saving teens: Using a policy discontinuity to estimate the effects of Medicaid eligibility Journal of Human Resources 51 3 2015 556 568
Wilson W. The truly disadvantaged: The inner city, the underclass and public policy 1987 University of Chicago Press Chicago, IL
Yang M.-Y. Maguire-Jack K. Showalter K. Kim Y.K. Slack K.S. Child care subsidy and child maltreatment Child & Family Social Work 24 4 2019 547 554 10.1111/cfs.12635
Young R. Hinnant A. Leshner G. Individual and social determinants of obesity in strategic health messages: Interaction with political ideology Health Communication 31 7 2016 903 910 26698295
Zlotnick C. What research tells us about the intersecting streams of homelessness and foster care American Journal of Orthopsychiatry 79 3 2009 319 325 10.1037/a0017218 19839669
| 36466794 | PMC9704498 | NO-CC CODE | 2022-12-05 23:15:16 | no | Child Youth Serv Rev. 2023 Jan 28; 144:106745 | utf-8 | Child Youth Serv Rev | 2,022 | 10.1016/j.childyouth.2022.106745 | oa_other |
==== Front
Infect Dis Clin North Am
Infect Dis Clin North Am
Infectious Disease Clinics of North America
0891-5520
1557-9824
Elsevier Health Sciences Division
S0891-5520(22)00079-4
10.1016/S0891-5520(22)00079-4
Article
After the COVID-19 Crisis: Update on Complex Infectious Disease Issues in the Intensive Care Unit
O’Grady Naomi P. Editor
Kadri Sameer S. Editor
31 10 2022
12 2022
31 10 2022
36 4 ii
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
==== Body
pmcINFECTIOUS DISEASE CLINICS OF NORTH AMERICA
www.id.theclinics.com
Consulting Editor
HELEN W. BOUCHER
December 2022 • Volume 36 • Number 4
| 0 | PMC9704818 | NO-CC CODE | 2022-12-01 23:19:05 | no | Infect Dis Clin North Am. 2022 Dec 31; 36(4):i | utf-8 | Infect Dis Clin North Am | 2,022 | 10.1016/S0891-5520(22)00079-4 | oa_other |
==== Front
Lancet Rheumatol
Lancet Rheumatol
The Lancet. Rheumatology
2665-9913
The Author(s). Published by Elsevier Ltd.
S2665-9913(22)00357-5
10.1016/S2665-9913(22)00357-5
Corrections
Correction to Lancet Rheumatol 2022; 4: e853–63
28 11 2022
28 11 2022
© 2022 The Author(s). Published by Elsevier Ltd.
2022
Elsevier Ltd
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
==== Body
pmcRussell MD, Galloway JB, Andrews CD, et al. Incidence and management of inflammatory arthritis in England before and during the COVID-19 pandemic: a population-level cohort study using OpenSAFELY. Lancet Rheumatol 2022; 4: e853–63—In the Summary of this Article, the Funding line should have read “None”, and in the Methods, the Role of the funding source section should have read “There was no direct funding source for this study”. The first sentence of the Acknowledgments was changed to “KB received funding from Versus Arthritis and Pfizer Global Medical Grants for Quality Improvement in Rheumatology Practice (68033839)”, and the eighth sentence to “No funding bodies had any role in study design, data collection, analysis or interpretation, manuscript writing, or in the decision to submit the article for publication”. These corrections have been made to the online version as of Nov 28, 2022.
| 36466787 | PMC9704819 | NO-CC CODE | 2022-12-01 23:19:05 | no | Lancet Rheumatol. 2022 Nov 28; doi: 10.1016/S2665-9913(22)00357-5 | utf-8 | Lancet Rheumatol | 2,022 | 10.1016/S2665-9913(22)00357-5 | oa_other |
==== Front
Lancet Oncol
Lancet Oncol
The Lancet. Oncology
1470-2045
1474-5488
Elsevier Ltd.
S1470-2045(22)00505-8
10.1016/S1470-2045(22)00505-8
Policy Review
International recommendations for plasma Epstein-Barr virus DNA measurement in nasopharyngeal carcinoma in resource-constrained settings: lessons from the COVID-19 pandemic
Lee Victor Ho-Fun MD ac*
Adham Marlinda MD d
Ben Kridis Wala Prof MD e
Bossi Paolo MD fg
Chen Ming-Yuan Prof MD h
Chitapanarux Imjai Prof MD i
Gregoire Vincent Prof MD j
Hao Sheng Po Prof MD k
Ho Cheryl MD l
Ho Gwo Fuang Prof MD m
Kannarunimit Danita MD n
Kwong Dora Lai-Wan Prof MD ac
Lam Ka-On MBBS ac
Lam Wai Kei Jacky FRCS op
Le Quynh-Thu Prof MD q
Lee Anne Wing-Mui Prof MD ac
Lee Nancy Y Prof MD r
Leung To-Wai MD ac
Licitra Lisa Prof MD gs
Lim Darren Wan-Teck MRCP t
Lin Jin-Ching Prof MD v
Loh Kwok Seng MBBS w
Lou Pei-Jen Prof MD xy
Machiels Jean-Pascal Prof MD z
Mai Hai-Qiang Prof MD h
Mesía Ricard MD aa
Ng Wai-Tong Prof MD ac
Ngan Roger Kai-Cheong Prof MBBS ac
Tay Joshua K PhD w
Tsang Raymond King-Yin MS bw
Tong Chi-Chung MBChB a
Wang Hung-Ming Prof MD ab
Wee Joseph T MBBS u
a Department of Clinical Oncology, School of Clinical Medicine, LKS Faculty of Medicine, The University of Hong Kong, Hong Kong Special Administrative Region, China
b Department of Surgery, School of Clinical Medicine, LKS Faculty of Medicine, The University of Hong Kong, Hong Kong Special Administrative Region, China
c Clinical Oncology Center, The University of Hong Kong–Shenzhen Hospital, Shenzhen, China
d Department of Otorhinolaryngology–Head and Neck Surgery, Faculty of Medicine, Universitas Indonesia–Cipto Mangunkusumo Hospital, Jakarta, Indonesia
e Department of Medical Oncology, Habib Bourguiba Hospital, University of Sfax, Sfax, Tunisia
f Medical Oncology Unit, Department of Medical and Surgical Specialties, Radiological Sciences and Public Health–Medical Oncology, University of Brescia, ASST–Spedali Civili, Brescia, Italy
g Head and Neck Medical Oncology Unit, Fondazione IRCCS Istituto Nazionale Dei Tumori, Milan, Italy
h Department of Nasopharyngeal Carcinoma, Sun Yat–sen University Cancer Center, State Key Laboratory of Oncology in South China, Collaborative Innovation Centre for Cancer Medicine, Guangdong Key Laboratory of Nasopharyngeal Carcinoma Diagnosis and Therapy, Guangzhou, China
i Division of Radiation Oncology, Faculty of Medicine, Chiang Mai University, Chiang Mai, Thailand
j Department of Radiation Oncology, Centre Léon Bérard, Lyon, France
k Department of Otolaryngology, Shin Kong Wu Ho–Su Memorial Hospital, Fu Jen Catholic University, New Taipei City, Taiwan
l Department of Medicine, Faculty of Medicine, University of British Columbia, Vancouver, BC, Canada
m Clinical Oncology Unit, University Malaya Cancer Centre, University of Malaya, Kuala Lumpur, Malaysia
n Division of Radiation Oncology, Department of Radiology, Faculty of Medicine, Chulalongkorn University, Bangkok, Thailand
o Li Ka Shing Institute of Health Sciences, The Chinese University of Hong Kong, Shatin, Hong Kong Special Administrative Region, China
p Department of Chemical Pathology, State Key Laboratory of Translational Oncology, and Department of Otorhinolaryngology, Head and Neck Surgery, The Chinese University of Hong Kong, Prince of Wales Hospital, Shatin, Hong Kong Special Administrative Region, China
q Department of Radiation Oncology, Stanford University, Stanford, CA, USA
r Department of Radiation Oncology, Memorial Sloan Kettering Cancer Center, New York, NY, USA
s Department of Oncology and Hemato–Oncology, University of Milan, Milan, Italy
t Division of Medical Oncology, National Cancer Centre Singapore, Singapore
u Division of Radiation Oncology, National Cancer Centre Singapore, Singapore
v Department of Radiation Oncology, Changhua Christian Hospital, Changhua, Taiwan
w Department of Otolaryngology–Head & Neck Surgery, National University of Singapore, Singapore
x Department of Otolaryngology, National Taiwan University Hospital and National Taiwan University College of Medicine, Taipei, Taiwan
y Graduate Institute of Anatomy and Cell Biology, National Taiwan University College of Medicine, Taipei, Taiwan
z Service d'Oncologie Médicale, Institut Roi Albert II, Cliniques Universitaires Saint-Luc, Brussels, Belgium
aa Medical Oncology Department, Catalan Institute of Oncology–Badalona, B–ARGO Group, IGTP, Badalona, Spain
ab Division of Hematology-Oncology, Department of Internal Medicine, Chang Gung Memorial Hospital, Chang Gung University, Taoyuan, Taiwan
* Correspondence to:Dr Victor Ho-Fun Lee, Department of Clinical Oncology, School of Clinical Medicine, LKS Faculty of Medicine, The University of Hong Kong, Hong Kong Special Administrative Region, China
28 11 2022
12 2022
28 11 2022
23 12 e544e551
© 2022 Elsevier Ltd. All rights reserved.
2022
Elsevier Ltd
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
The effects of the COVID-19 pandemic continue to constrain health-care staff and resources worldwide, despite the availability of effective vaccines. Aerosol-generating procedures such as endoscopy, a common investigation tool for nasopharyngeal carcinoma, are recognised as a likely cause of SARS-CoV-2 spread in hospitals. Plasma Epstein-Barr virus (EBV) DNA is considered the most accurate biomarker for the routine management of nasopharyngeal carcinoma. A consensus statement on whether plasma EBV DNA can minimise the need for or replace aerosol-generating procedures, imaging methods, and face-to-face consultations in managing nasopharyngeal carcinoma is urgently needed amid the current pandemic and potentially for future highly contagious airborne diseases or natural disasters. We completed a modified Delphi consensus process of three rounds with 33 international experts in otorhinolaryngology or head and neck surgery, radiation oncology, medical oncology, and clinical oncology with vast experience in managing nasopharyngeal carcinoma, representing 51 international professional societies and national clinical trial groups. These consensus recommendations aim to enhance consistency in clinical practice, reduce ambiguity in delivering care, and offer advice for clinicians worldwide who work in endemic and non-endemic regions of nasopharyngeal carcinoma, in the context of COVID-19 and other airborne pandemics, and in future unexpected settings of severe resource constraints and insufficiency of personal protective equipment.
==== Body
pmcIntroduction
The COVID-19 pandemic is estimated by WHO to have claimed more than 6·4 million lives since December, 2019.1 Various types of COVID-19 vaccines against the original strain of SARS-CoV-2 and the alpha (B.1.1.7) variant that emerged in January, 2021 have been shown to be safe and effective, as evidenced by clinical trials and real-world data.2, 3, 4, 5, 6, 7, 8 However, SARS-CoV-2, like other RNA viruses, is very prone to genetic evolution as it adapts to new human hosts.1 So far, at least ten variants have been identified worldwide, of which the delta (B.1.617.2) variant and the more recent omicron (B.1.1.529) variant, in particular, have been jeopardising millions of people's lives since late November, 2021. The currently licensed and approved COVID-19 vaccines might not effectively prevent infection with these new and other imminent subvariants (eg, BA.2, BA.3, BA.4, and BA.5).9, 10, 11, 12, 13 Although the pandemic is seemingly contained in Europe and North America, at the time of writing, it continues to be widely active in southeast Asia, the Western Pacific region, and Africa. The more contagious omicron variants are still spreading relentlessly in these regions, which is leading to acute demand and subsequent scarcity of intensive care resources, quarantine and isolation facilities, and rapid antigen testing kits. Although international recommendation guidelines and consensus statements on risk stratification and treatment of head and neck cancers have been published,14, 15, 16, 17, 18, 19 none exist for nasopharyngeal carcinoma, a malignancy that is endemic to the aforementioned geographical regions. Nasopharyngeal carcinoma is a unique, distinct disease entity distinguished from head and neck squamous cell carcinoma by its strong association with Epstein-Barr virus (EBV).20 Concerted efforts, over the past two decades, to improve and internationally harmonise detection limits have made plasma EBV DNA the most sensitive and specific biomarker for the screening, diagnosis, risk stratification, treatment response evaluation, relapse surveillance, and prognostication of this deadly malignancy.21 More detailed and sophisticated investigative tools with MRI with diverse sequences (compared with CT for head and neck squamous cell carcinoma), PET with integrated CT, and nasoendoscopy are very often indicated in the routine clinical management of nasopharyngeal carcinoma.22, 23, 24, 25 However, nasoendoscopy has been classified, with a high degree of agreement, as an aerosol-generating procedure with a high risk for viral aerosolisation.15, 26 Unfortunately, provision of these imaging and endoscopy services and typical face-to-face doctor–patient consultations had to be suspended or were severely delayed during the pandemic. Given the accuracy and reliability of plasma EBV DNA, and to address the urgent need for aerosol-free procedures to manage nasopharyngeal carcinoma in the face of the potential for future waves of the COVID-19 or other airborne pandemics, we used a modified online Delphi process, with representation from nasopharyngeal carcinoma experts worldwide, to develop consensus recommendations on the use of plasma EBV DNA for the management of nasopharyngeal carcinoma during the COVID-19 pandemic or other circumstances in which personnel and resources are severely constrained in an acute setting.
Data collection, survey design, and participant recruitment
The steering committee (VH-FL, AW-ML, and W-TN) collated all published literature on plasma EBV DNA in nasopharyngeal carcinoma management, designed the online survey, and piloted the questions for the survey. During August, 2021, we invited 33 nasopharyngeal carcinoma experts from the fields of otorhinolaryngology or head and neck surgery, radiation oncology, medical oncology, and clinical oncology. Experts from four continents (Asia, North America, Europe, and Africa) were invited, specifically from China, Hong Kong, Taiwan, Thailand, Malaysia, Singapore, Indonesia, Italy, France, Belgium, Spain, the USA, Canada, and Tunisia. Experts were considered if they had practised as full-time clinicians in tertiary academic institutions and university teaching or affiliated hospitals for at least 10 years, managed and treated more than 200 patients with nasopharyngeal carcinoma in total, authored or coauthored at least 20 journal articles on nasopharyngeal carcinoma or plasma EBV DNA, and had clinical experience in using plasma EBV DNA in both clinical trials and routine clinical practice. The invited experts provided representation from 51 international professional societies and national clinical trial groups (panel ).Panel Bodies and societies represented by the experts in this consensus development committee
• American Academy of Otolaryngology Head and Neck Surgery Foundation
• American Association for Cancer Research
• American Association for the Advancement of Science
• American Head and Neck Society
• American Joint Committee on Cancer
• American Radium Society
• American Society for Radiation Oncology
• American Society of Clinical Oncology
• Asian Society of Head and Neck Oncology
• BC Cancer Head and Neck Tumour Group
• Catalan-Balearic Society for Oncology
• China Anti-Cancer Association
• Chinese Society of Clinical Oncology
• European Head and Neck Society
• European Organisation for Research and Treatment of Cancer
• European Society for Medical Oncology
• European Society for Radiotherapy and Oncology
• Federations of Asian Organizations for Radiation Oncology
• Head and Neck Cancer International Group
• Hong Kong Cancer Therapy Society
• Hong Kong College of Otorhinolaryngologists
• Hong Kong College of Radiologists
• Hong Kong Head and Neck Society
• Hong Kong Nasopharyngeal Carcinoma Study Group
• Hong Kong Society of Clinical Oncology
• Hong Kong Society of Otorhinolaryngology, Head and Neck Surgery
• Indonesian Otorhinolaryngology Head and Neck Society
• International Academy of Oral Oncology
• International Commission on Radiation Units and Measurements
• International Federation of Head and Neck Oncologic Societies
• International Head and Neck Scientific Group
• Italian Association of Head and Neck Oncology
• Italian Association of Medical Oncology
• Malaysian Oncological Society
• National Cancer Staging Committee of China
• North-West Oncological Italian Group
• NRG Oncology
• Pan-Pearl River Delta Radiation Oncology Conference Committee
• Radiological Society of North America
• Royal College of Radiologists
• Singapore Radiological Society
• Singapore Society of Oncology
• Sociedad Española de Oncología Médica
• Société Française Radiation Oncology
• South East Asian Radiation Oncology Group
• Spanish Head and Neck Cancer Group
• Taiwan Head and Neck Society
• Taiwan Oncology Society
• Taiwan Society for Therapeutic Radiology and Oncology
• Thai Association of Radiation Oncology
• Union for International Cancer Control
Consensus establishment
The collected published literature and article of recommendations on the use of plasma EBV DNA in the management of nasopharyngeal carcinoma21 was shared with all participating experts. The article summarises all the published literature and provides a comprehensive overview and recommendation on the use of plasma EBV DNA in various routine clinical settings. The experts could comment on the literature and survey questions before the commencement of the online survey. Upon acceptance of invitation, they were given the link to complete the online questionnaire. Consensus statements were developed with an online, modified Delphi process done over three rounds (figure ). The online survey consisted of two parts: the first on the use of plasma EBV DNA in diagnosis, pretreatment investigations, staging, response monitoring during radical treatment, surveillance, response monitoring for recurrent or metastatic nasopharyngeal carcinoma in a normal (non-pandemic) setting (questions 5–10); and the second on the use of plasma EBV DNA in the same clinical circumstances but in the context of severe personnel and resource constraints and risk of COVID-19 transmission as a result of a biopsy or the aerosol-generating nasoendoscopy procedure (questions 11–26; appendix pp 4–7).Figure Modified Delphi process of the online survey
Participants were invited by email to complete each round of the survey, which was open for a period of 3 weeks. A reminder email was sent 2 days and 1 day before the deadline to prompt participants who had not yet submitted their responses to complete the survey. When completing the survey, participants were expected to consider an extremely constrained setting in terms of capacity and resources (including, but not limited to, a severe dearth in health-care professionals and other clinical clerical staff members, operating and endoscopy room capacity, inpatient and intensive care bed capacity, ventilation facilities, and rapid COVID-19 diagnostic tests) compared with baseline before and during the initial waves of the COVID-19 pandemic. Following each round of the survey, the steering committee analysed the results and applied the following predetermined criteria for agreement, which were based on the RAND method27 and set in the protocol before the start of the project: 80% or higher agreement was classified as strong agreement for a statement, whereas 20% or lower indicated a strong disagreement. For each statement, the Delphi process was stopped either when strong agreement was reached or after completion of three rounds of the survey, whichever occurred first. Items that reached strong agreement were not included again in subsequent rounds of the survey. After the third round, statements that did not reach the strong agreement threshold but that reached a threshold of agreement of 67% of higher were considered to have reached agreement.18 Results from the first and second rounds were emailed to participants for review before the next round commenced. Participants were reminded that they could change their responses in the subsequent round for questions that had not yet reached strong agreement. When necessary, questions would be iteratively revised between rounds before being asked again, and new questions would be introduced to provide more granularity to the topic or to reduce ambiguity on the phrasing of a previous question. The consensus statement development process and the online survey were approved by the Institutional Review Board of the University of Hong Kong and of the Hong Kong West Cluster Hospital Authority.
Findings
The first four questions appeared only in the first round of the survey and asked participants for their name (question 1), specialty (otorhinolaryngology or head and neck surgery, radiation oncology, medical oncology, and clinical oncology; question 2), all affiliations with any local, regional, national, or international professional surgery or oncology organisations (question 3), and any membership in international surgery or oncology organisations (question 4; appendix pp 1–3).
A total of 22 questions pertaining to the objective of this consensus recommendation were asked in this survey. 22 questions were asked in the first round. A summary of the results of the three rounds of the survey is available in the appendix (pp 4–8) and in the table .Table Consensus recommendations for the use of plasma EBV DNA in the management of nasopharyngeal carcinoma
Agreement or disagreement level
Part 1: use of plasma EBV DNA in typical (non-pandemic) circumstances
For pretreatment investigation of newly diagnosed nasopharyngeal carcinoma Strong agreement
To monitor response during radical treatment Strong agreement
After completion of radical treatment Strong agreement
For disease surveillance during subsequent follow-up for early detection of relapse Strong agreement
To monitor response to salvage treatment for recurrent or metastatic nasopharyngeal carcinoma Strong agreement
Part 2: use of plasma EBV DNA in a setting of acute and severe personnel and resource constraints (eg, during the COVID-19 pandemic)
As the only screening and diagnostic tool for nasopharyngeal carcinoma, replacing nasoendoscopy, biopsy, and other diagnostic or imaging tools Strong disagreement
In combination with IgA anti-VCA as the only screening and diagnostic tool for nasopharyngeal carcinoma, replacing nasoendoscopy, biopsy, and other diagnostic or imaging tools Strong disagreement
For repeat testing (eg, 4 weeks later) as the only screening and diagnostic tool for a patient with clinical suspicion of nasopharyngeal carcinoma if the initial plasma EBV DNA titre is undetectable, replacing nasoendoscopy, biopsy, and other diagnostic or imaging tools in clinic Strong disagreement
A high plasma EBV DNA concentration together with salient clinical symptoms but without histological confirmation is sufficient for the diagnosis of nasopharyngeal carcinoma Strong disagreement
In combination with IgA anti-VCA with imaging tools (eg, ultrasonography, CT, MRI, bone scintigraphy, and PET–CT) as the only staging investigations for nasopharyngeal carcinoma, replacing nasoendoscopy and biopsy Disagreement
In combination with IgA anti-VCA without imaging tools (except CT scan for radiotherapy planning purpose) as the only staging investigations for nasopharyngeal carcinoma, replacing nasoendoscopy and biopsy in clinic Strong disagreement
As the only test to replace clinical consultations to monitor the progress of nasopharyngeal carcinoma in the course of radical treatment Disagreement
As the only test without imaging tools (eg, ultrasonography, CT, MRI, and PET-CT), to replace nasoendoscopy and biopsy and other diagnostic tools in clinic to confirm complete local and regional remission after completion of radical treatment Strong disagreement
In combination with imaging tools (eg, ultrasonography, CT, MRI, bone scintigraphy, and PET-CT) as the only surveillance tools for relapse of nasopharyngeal carcinoma after completion of radical treatment, replacing nasoendoscopy, biopsy, and other diagnostic tools Agreement
As the only surveillance test without imaging tools (eg, ultrasonography, CT, MRI, bone scintigraphy, and PET-CT), replacing nasoendoscopy, biopsy, and other diagnostic tools for relapse of nasopharyngeal carcinoma after completion of radical treatment Strong disagreement
As the only surveillance test, replacing clinical consultations, nasoendoscopy, biopsy, and other imaging or diagnostic tools for relapse of nasopharyngeal carcinoma after radical treatment Strong disagreement
In combination with imaging tools (eg, ultrasonography, CT, MRI, and PET-CT), replacing nasoendoscopy, biopsy, and other diagnostic tools to diagnose clinically suspicious recurrent nasopharyngeal carcinoma Disagreement
As the only test without imaging tools (eg, ultrasonography, CT, MRI, and PET-CT), replacing nasoendoscopy, biopsy, and other diagnostic tools to diagnose clinically suspicious recurrent nasopharyngeal carcinoma Strong disagreement
An increased or progressively rising titre of plasma EBV DNA alone, without histological and imaging results, is sufficient for the diagnosis of recurrent nasopharyngeal carcinoma Strong disagreement
As the only test without other diagnostic and imaging tools to monitor the tumour response during and after treatment for recurrent or metastatic nasopharyngeal carcinoma Disagreement
Strong agreement corresponds to a threshold of 80% or higher agreement. Strong disagreement corresponds to a threshold of 20% or lower agreement. Agreement or disagreement corresponds to a threshold of 67% or higher after the third round for statements that did not reach a strong agreement or strong disagreement in previous rounds. EBV=Epstein-Barr virus. VCA=viral capsid antigen.
Consensus statement on measuring plasma EBV DNA in routine clinical management
All participants agreed that plasma EBV DNA should be routinely measured during pretreatment investigations. Although consensus was reached on the use of plasma EBV DNA for disease surveillance to rule out early relapse and to monitor treatment response during salvage treatment for recurrent or metastatic nasopharyngeal carcinoma, two (6%) participants queried its sensitivity to detect small, locally recurrent disease or early relapse. The question that did not reach any consensus or agreement (64%, question 5) was whether plasma EBV DNA could be considered for nasopharyngeal carcinoma screening in endemic regions, and whether its diagnostic sensitivity and cost-effectiveness could be further improved by combination with EBV IgA anti-viral capsid antigen testing. Slightly divided opinions on the use of plasma EBV DNA for screening appeared in question 5: although most of the experts commented that plasma EBV DNA and serology was cost-effective for nasopharyngeal carcinoma screening, others argued that neither plasma EBV DNA alone, or in combination with EBV IgA anti-viral capsid antigen or IgA early antigen, or both can further improve sensitivity and specificity, especially for small early-stage disease.
Measuring EBV DNA during resource constraints and to avoid aerosol generation
Screening and diagnosis
The participants were strongly against the use and repeated use (eg, 4 weeks after an initial undetectable titre) of plasma EBV DNA exclusively (88% disagreement, questions 11 and 13) or in combination with EBV IgA anti-viral capsid antigen (94% disagreement, question 12) to replace nasoendoscopy and other diagnostic or imaging tools in the clinic to screen for and diagnose nasopharyngeal carcinoma during the pandemic, when resources were severely low and aerosol-generating procedures were to be avoided. Similarly, a consensus strong disagreement was also reached regarding the use of plasma EBV DNA without histological confirmation to diagnose nasopharyngeal carcinoma (97% disagreement, question 14), even for patients who present with salient clinical symptoms. Most of the participants commented that nasoendoscopic examination, biopsy (to rule out other EBV-associated malignancies, such as natural killer T-cell lymphoma, which is common in Asia and in Latin America), and imaging were still essential to nasopharyngeal carcinoma diagnosis, even in the setting of restricted resources. All experts who disagreed with plasma EBV DNA as the only investigation tool urged for the use of full personal protective equipment to conduct risky endoscopic procedures. They further commented that a histological diagnosis of nasopharyngeal carcinoma was necessary before starting any treatment.
There was unanimous disagreement (100%) regarding the use of plasma EBV DNA and IgA anti-viral capsid antigen, without imaging tools, to replace nasoendoscopy and biopsy as the only staging investigations for nasopharyngeal carcinoma during the COVID-19 pandemic (question 16). Even with the availability of imaging tools, the participants expressed disagreement against the use of plasma EBV DNA to replace nasoendoscopy and biopsy as the only staging investigation during the pandemic (73% disagreement, question 15). Again, the participants suggested an endoscopic examination to evaluate the local extent of the disease, and histological confirmation to establish the diagnosis and exclude other EBV-associated malignancies, even in resource-constrained settings.
Treatment response monitoring
The participants did not support the use of plasma EBV DNA to replace clinical consultations to monitor the progress of nasopharyngeal carcinoma in the course of its radical treatment during the pandemic (73% disagreement, question 17). Clinical evaluations in the form of face-to-face consultations, as commented by the participants, were needed to monitor for acute toxicity during radical treatment.
Confirmation of disease remission after radical treatment
Although 55% of the participants agreed on the use of plasma EBV DNA and imaging tools to replace nasoendoscopy, biopsy, and other diagnostic tools to confirm local and regional remission of nasopharyngeal carcinoma after radical treatment during the pandemic (question 18), an agreement could not be reached because other participants believed that nasoendoscopy with or without biopsy were still necessary to evaluate local remission after radical treatment, and for patients whose pretreatment plasma EBV DNA was undetectable. By contrast, most participants strongly disagreed with the use of plasma EBV DNA only, without any imaging tools, nasoendoscopy, or biopsy (91% disagreement, question 19), as the only way to confirm local and regional remission following radical treatment.
Disease surveillance after completion of radical treatment
An agreement (70% agreement, question 20) was reached on the use of plasma EBV DNA and imaging, to replace nasoendoscopy and biopsy, as the only surveillance tools for nasopharyngeal carcinoma relapse after completion of radical treatment during the pandemic. However, strong disagreement (82% disagreement, question 21) was observed if only plasma EBV DNA was used as the surveillance tool for nasopharyngeal carcinoma relapse, to replace nasoendoscopy, biopsy, and other imaging or diagnostic tools, and to replace clinical consultation, nasoendoscopy, biopsy, and other imaging or diagnostic tools in severely resource-constrained settings (85% disagreement, question 22).
Diagnosis of clinically suspicious recurrent disease
When asked whether plasma EBV DNA could replace nasoendoscopy and biopsy to diagnose clinically suspicious recurrent symptoms in patients who presented with symptoms highly suggestive of nasopharyngeal carcinoma recurrence, most participants disagreed, even when imaging scans were available (79% disagreement, question 23), and all strongly disagreed if imaging services were not provided at all due to resource constraints during the pandemic (100% disagreement, question 24). Similarly, most participants disagreed with increased or progressively rising titres of plasma EBV alone, without any histological and imaging correlation, as the only criterion to diagnose nasopharyngeal carcinoma recurrence during the pandemic (94% disagreement, question 25).
Monitoring treatment response in recurrent or metastatic disease
Most participants disagreed with measuring plasma EBV DNA only, without other imaging or diagnostic tools, to monitor the treatment response for recurrent or metastatic nasopharyngeal carcinoma during the pandemic (73% disagreement, question 26). The general opinion was that imaging assessment, even if done over a prolonged interval due to resource limitations, was still needed because it provided a clear presentation of the extent of the tumour response, which could not be accurately reflected by the change in titre of plasma EBV DNA.
Discussion
Endoscopic and imaging assessment are essential procedures for the clinical management of nasopharyngeal carcinoma. However, to our knowledge, the possibility of limited availability of these facilities and resources during a global pandemic had not before been considered, even in high-income countries and regions with well established health-care infrastructure and expertise. Plasma EBV DNA has been considered an accurate, reliable, and cost-effective tumour marker and biomarker of nasopharyngeal carcinoma, which has been widely used in screening, diagnosis, prognostication, surveillance for relapse, and treatment response evaluation during radical and salvage treatment.21, 28 Histological confirmation from a tumour biopsy is still the gold standard for the diagnosis of cancer, including nasopharyngeal carcinoma. However, whether liquid biopsies, which can contain cell-free tumour DNA or circulating tumour DNA, could replace conventional diagnostic methods such as tumour biopsies and endoscopy and imaging assessment is a question worth exploring in light of the COVID-19 pandemic. Nasoendoscopy has been classified with a high level of consensus as a possibly aerosol-generating procedure.15, 26 For this reason, and considering that the pandemic also occasioned a severe shortage of health-care personnel, personal protective equipment, and rapid diagnostic tests for SARS-CoV-2, we developed this global consensus recommendation on the use of plasma EBV DNA in routine clinical management of nasopharyngeal carcinoma, which aims to provide a guideline to health-care staff actively involved in the management of nasopharyngeal carcinoma and overwhelmed by the pandemic. These recommendations should not be considered as permanent, but as contingent on current conditions and in place for future pandemics or natural disasters. Furthermore, they should be interpreted in the context of global, national, regional, and local COVID-19 circumstances, which can drastically change in a short time.
Nasopharyngeal carcinoma is managed quite differently from head and neck squamous cell carcinoma. Nasoendocopy and nasoendoscopic biopsy, imaging with MRI with numerous scanning sequences, and PET-CT scans are essential components for the diagnosis of nasopharyngeal carcinoma and its differentiation from post-treatment changes and treatment-related complications.22, 23, 24, 25 Unlike head and neck squamous cell carcinoma, for which no accurate and reliable tumour biomarker exists, nasopharyngeal carcinoma typically results in circulating cell-free EBV DNA, which can be easily and reliably extracted as plasma EBV DNA and evaluated in routine blood taking. Following the concerted efforts on international harmonisation of the performance level of the assay, plasma EBV DNA has come into worldwide usage in routine clinical care and as an important risk stratification factor in international, multicentre, randomised controlled trials.21, 29, 30
During a pandemic, considering alternative emergency strategies for the management of nasopharyngeal carcinoma becomes imperative, as nasoendoscopy, imaging scans, and even in-person medical consultations pose substantial exposure risks to both health-care workers and patients, especially in situations of scarcity of front-line health-care workers and personal protective equipment.31, 32 The results of our survey suggest that, although plasma EBV DNA (supplemented, when possible, by imaging scans) can be used in most clinical circumstances in the context of resource limitations, it cannot completely replace nasoendoscopy and tumour biopsy. All experts who disagreed with the use of plasma EBV DNA as the only tool in nasopharyngeal carcinoma management considered that these procedures are essential and that the patients’ right to receive these standard investigations should not be suppressed, insofar as the health-care system allowed.
The invited experts represented a broad range of clinical disciplines relevant for the care of nasopharyngeal carcinoma, consisting of professionals experienced in managing this malignancy as front-line doctors from across four continents. They also hold leadership and membership positions in international and national surgical and oncological societies and come from both low-income and middle-income countries and high-income countries, being very familiar with their national, regional, and local clinical practice of nasopharyngeal carcinoma management. Our recommendations have taken into consideration various clinical scenarios from the joint perspectives of surgeons and non-surgeons, unlike previous recommendations published by surgeons and oncologists separately. In contrast to previous recommendations on flexible nasoendoscopy only for patients who are deemed to be at high risk of cancer recurrence and mortality from head and neck cancers and only when adequate personal protective equipment is available,17 our experts strongly disagreed with measuring plasma EBV DNA only to replace the conventional but essential nasoendoscopic examinations, even when imaging services were scarce. Plasma EBV DNA in combination with imaging scans can be considered an acceptable alternative in a very restricted clinical setting, but cannot replace face-to-face consultations and nasoendoscopy, and it cannot be used alone (without imaging) for the management of nasopharyngeal carcinoma. We believe that these recommendations, which are the concerted efforts and results of international collaboration and cooperation, can serve as references and be used in other acute settings or during natural disasters, where there is an unexpectedly high risk of health hazards to patients and health-care workers and a severe paucity of health personnel and resources. We also hope that our recommendations can further promote and invite international harmonisation and more affordable use of plasma EBV DNA for nasopharyngeal carcinoma and other accurate blood or liquid biomarkers for other cancers at institutions and hospitals in low-income and middle-income countries, so that patients treated in these locations are not underprivileged in the face of future pandemics or natural disasters.
Our recommendations have some limitations. They might not be fully applicable to countries or regions where nasopharyngeal carcinoma is sporadic, and where standardised and accurate plasma EBV DNA assays are not available. When asked to participate in this study, our experts were expected to be confronted with any eventuality and the most difficult situations when all the necessary resources and personnel were severely constrained and rationed, which might not actually occur in some countries or regions where the pandemic was better controlled. The time period over which this survey was conducted was the most critical and difficult for the Asian population affected by the omicron variant and subvariants, happening at a time when the most stringent infection control and physical distancing measures were gradually being lifted in North America and in Europe. A strong clinical network should be provided and duly modulated with continuous support to facilitate front-line clinicians when they are faced with extremely difficult and fluctuating situations in different regions of the world.
Conclusion
We are hoping that the pandemic is nearing its end, but global preparedness for a similar or different disruptor is highly advised. Although plasma EBV DNA has reached a crucial role in the clinical management of nasopharyngeal carcinoma, nasoendoscopic and imaging examinations and face-to-face consultations remain essential, even in the setting of acute resource and personnel limitations, such as during the COVID-19 pandemic. Our consensus recommendations illustrate an excellent example of international collaboration in times of current and imminent global challenges, which can be easily applied and adopted to suit different needs. Although these recommendations were specifically developed for the COVID-19 pandemic, they have the potential generalisability to be applied in other circumstances of severe shortage of health-care personnel and resources. Measures to ensure adequate protective equipment for the safety of patients and health-care workers and essential support for health facilities are urgently warranted before future pandemics or disasters arise.
Search strategy and selection criteria
To clarify uncertainty around the use of plasma Epstein-Barrvirus (EBV) DNA in the clinical management of nasopharyngeal carcinoma in the setting of acute personnel and resource constraints during the COVID-19 pandemic, we did a literature search of PubMed, Embase, MEDLINE, and Google Scholar that included grey literature, individual association correspondence, and guidelines on COVID-19 (or COVID or pandemic) and nasopharyngeal carcinoma. We used the search terms “nasopharyngeal carcinoma” or “nasopharyngeal cancer”, “COVID-19” or “COVID” or “SARS-CoV-2”, and “EBV DNA” or “cell-free DNA” and searched for articles published in English between Nov 1, 2019, and April 30, 2022. We prioritised peer-reviewed journal articles, large case-series papers, and guidelines and recommendations published by local, regional, national, or international bodies. All participating experts held discussions to identify areas of uncertainty in the literature search and selection criteria. We did not identify any article or paper that fulfilled our selection criteria. Therefore, we designed and conducted three rounds of online survey with modified Delphi method and established consensus statements. All domains (part 1 and 2 in the survey) and questions within each domain were piloted by the steering committee (VH-FL, AW-ML, and W-TN) for readability and validity (appendix pp 4–8).
Declaration of interests
VH-FL reports personal fees and grants from AstraZeneca; and personal fees from AQUILAB, Amgen, Boston Scientific, Eli Lilly, Merck Sharp and Dohme, Novartis, Pfizer, and Takeda, all outside this Policy Review. GFH reports fees paid to the institution from Eli Lilly, Regeneron, Merck Sharp and Dohme, AB Science, Astellas, Tessa Therapeutics, Roche, Arcus Bioscience, AstraZeneca, and Pfizer; and personal fees from Merck Sharp and Dohme, Novartis, Roche, AstraZeneca, Boehringer Ingelheim, Pfizer, Ipsen, Bristol Myers Squibb, Janssen, and Taiho, all outside the scope of this Policy Review. WKJL is a shareholder of Illumina/GRAIL. Q-TL reports personal fees from the RTOG Foundation and NRG Oncology, outside the scope of this Policy Review, and is supported by the US National Institutes of Health (2U10CA180868-06, 1R01DE029672-01A1, P30CA124435, and R01DE030894-01A1). NYL is supported by the US National Institutes of Health (R01 CA238392-02A1 and 5UG1CA233290-02). DW-TL reports institutional fees from Bristol Myers Squibb, Merck Sharp and Dohme, Boehringer Ingelheim, Janssen, and Novartis, outside the scope of this Policy Review. RM reports personal fees from Merck Pharmaceuticals, Merck Sharp and Dohme, Bristol Myers Squibb, Roche, Amgen, AstraZeneca, Nanobiotix, Seattle Genetics, and Boehringer Ingelheim, all outside the scope of this Policy Review. RK-CN reports personal fees from Pfizer, Novartis, Sanofi, AstraZeneca, Eli Lilly, Merck Sharp and Dohme, Zai Lab, Roche, Eisai, Merck Pharmaceutical, Astellas, and Nuance (China), outside the scope of this Policy Review. RK-YT is the President of Hong Kong Society of Otolaryngology, Head and Neck Surgery, and past President of Hong Kong Head and Neck Society. All other authors declare no competing interests.
Supplementary Material
Supplementary appendix
Acknowledgments
We thank Karina Cheung of the Department of Clinical Oncology, School of Clinical Medicine, LKS Faculty of Medicine, The University of Hong Kong (Hong Kong Special Administrative Region, China), for her clinical and secretarial support on this study.
Editorial note: the Lancet Group takes a neutral position with respect to territorial claims in published maps and institutional affiliations.
Contributors
VH-FL, AW-ML, and W-TN conceived and designed the study and piloted the survey questions. All authors participated in study development, data collection and interpretation, manuscript preparation, and approved the final manuscript of this Policy Review. The corresponding author had final responsibility for the decision to submit for publication.
==== Refs
References
1 Dejnirattisai W Huo J Zhou D SARS-CoV-2 omicron-B.1.1.529 leads to widespread escape from neutralizing antibody responses Cell 185 2022 467 484 e15 35081335
2 Polack FP Thomas SJ Kitchin N Safety and efficacy of the BNT162b2 mRNA Covid-19 vaccine N Engl J Med 383 2020 2603 2615 33301246
3 Voysey M Clemens SAC Madhi SA Safety and efficacy of the ChAdOx1 nCoV-19 vaccine (AZD1222) against SARS-CoV-2: an interim analysis of four randomised controlled trials in Brazil, South Africa, and the UK Lancet 397 2021 99 111 33306989
4 Baden LR El Sahly HM Essink B Efficacy and safety of the mRNA-1273 SARS-CoV-2 vaccine N Engl J Med 384 2021 403 416 33378609
5 Lopez Bernal J Andrews N Gower C Effectiveness of the Pfizer-BioNTech and Oxford-AstraZeneca vaccines on covid-19 related symptoms, hospital admissions, and mortality in older adults in England: test negative case-control study BMJ 373 2021 n1088
6 Hall VJ Foulkes S Saei A COVID-19 vaccine coverage in health-care workers in England and effectiveness of BNT162b2 mRNA vaccine against infection (SIREN): a prospective, multicentre, cohort study Lancet 397 2021 1725 1735 33901423
7 Thompson MG Burgess JL Naleway AL Interim estimates of vaccine effectiveness of BNT162b2 and mRNA-1273 COVID-19 vaccines in preventing SARS-CoV-2 infection among health care personnel, first responders, and other essential and frontline workers—eight U.S. locations, December 2020–March 2021 MMWR Morb Mortal Wkly Rep 70 2021 495 500 33793460
8 Dagan N Barda N Kepten E BNT162b2 mRNA Covid-19 vaccine in a nationwide mass vaccination setting N Engl J Med 384 2021 1412 1423 33626250
9 Abu-Raddad LJ Chemaitelly H Butt AA Effectiveness of the BNT162b2 Covid-19 vaccine against the B.1.1.7 and B.1.351 variants N Engl J Med 385 2021 187 189 33951357
10 Lopez Bernal J Andrews N Gower C Effectiveness of Covid-19 vaccines against the B.1.617.2 (delta) variant N Engl J Med 385 2021 585 594 34289274
11 Pouwels KB Pritchard E Matthews PC Effect of delta variant on viral burden and vaccine effectiveness against new SARS-CoV-2 infections in the UK Nat Med 27 2021 2127 2135 34650248
12 Andrews N Tessier E Stowe J Duration of protection against mild and severe disease by Covid-19 vaccines N Engl J Med 386 2022 340 350 35021002
13 Andrews N Stowe J Kirsebom F Covid-19 vaccine effectiveness against the omicron (B.1.1.529) variant N Engl J Med 386 2022 1532 1546 35249272
14 de Almeida JR Noel CW Forner D Development and validation of a surgical prioritization and ranking tool and navigation aid for head and neck cancer (SPARTAN-HN) in a scarce resource setting: response to the COVID-19 pandemic Cancer 126 2020 4895 4904 32780426
15 Maniakas A Jozaghi Y Zafereo ME Head and neck surgical oncology in the time of a pandemic: subsite-specific triage guidelines during the COVID-19 pandemic Head Neck 42 2020 1194 1201 32342541
16 Wu V Noel CW Forner D Considerations for head and neck oncology practices during the coronavirus disease 2019 (COVID-19) pandemic: Wuhan and Toronto experience Head Neck 42 2020 1202 1208 32338807
17 Mehanna H Hardman JC Shenson JA Recommendations for head and neck surgical oncology practice in a setting of acute severe resource constraint during the COVID-19 pandemic: an international consensus Lancet Oncol 21 2020 e350 e359 32534633
18 Thomson DJ Palma D Guckenberger M Practice recommendations for risk-adapted head and neck cancer radiation therapy during the COVID-19 pandemic: an ASTRO-ESTRO consensus statement Int J Radiat Oncol Biol Phys 107 2020 618 627 32302681
19 Thomson DJ Palma D Guckenberger M Practice recommendations for risk-adapted head and neck cancer radiotherapy during the COVID-19 pandemic: an ASTRO-ESTRO consensus statement Radiother Oncol 151 2020 314 321 32730830
20 Niedobitek G Young LS Sam CK Brooks L Prasad U Rickinson AB Expression of Epstein-Barr virus genes and of lymphocyte activation molecules in undifferentiated nasopharyngeal carcinomas Am J Pathol 140 1992 879 887 1314028
21 Lee AWM Lee VHF Ng WT A systematic review and recommendations on the use of plasma EBV DNA for nasopharyngeal carcinoma Eur J Cancer 153 2021 109 122 34153713
22 Lai V Li X Lee VH Lam KO Chan Q Khong PL Intravoxel incoherent motion MR imaging: comparison of diffusion and perfusion characteristics between nasopharyngeal carcinoma and post-chemoradiation fibrosis Eur Radiol 23 2013 2793 2801 23722897
23 Ai QY King AD Chan JSM Distinguishing early-stage nasopharyngeal carcinoma from benign hyperplasia using intravoxel incoherent motion diffusion-weighted MRI Eur Radiol 29 2019 5627 5634 30903340
24 Yang SS Wu YS Chen WC Benefit of [18F]-FDG PET/CT for treatment-naïve nasopharyngeal carcinoma Eur J Nucl Med Mol Imaging 49 2022 980 991 34468782
25 Xu J Wang J Bian X Deep learning for nasopharyngeal carcinoma identification using both white light and narrow-band imaging endoscopy Laryngoscope 132 2022 999 1007 34622964
26 Jackson T Deibert D Wyatt G Classification of aerosol-generating procedures: a rapid systematic review BMJ Open Respir Res 7 2020 e000730
27 Dalkey NC The Delphi method: an experimental study of group opinion 1969 RAND Corporation Santa Monica, CA
28 Miller JA Le QT Pinsky BA Wang H Cost-effectiveness of nasopharyngeal carcinoma screening with EBV PCR or serology in high-incidence populations worldwide J Natl Cancer Inst 113 2021 852 862 33351145
29 Lee Nancy NRG-HN001 randomized phase II and III studies of individualized treatment for nasopharyngeal carcinoma based on biomarker plasma Epstein-Barr virus (EBV) deoxyribonucleic acid (DNA) https://www.nrgoncology.org/Clinical-Trials/Protocol/nrg-hn001?filter=nrg-hn001
30 Chan ATC Hui EP Ngan RKC Analysis of plasma EBV DNA in nasopharyngeal cancer after chemoradiation to identify high-risk patients for adjuvant chemotherapy: a randomized-controlled trial J Clin Oncol 36 2018 3091 3100
31 Kang JJ Wong RJ Sherman EJ The 3 Bs of cancer care amid the COVID-19 pandemic crisis: “Be safe, be smart, be kind”—a multidisciplinary approach increasing the use of radiation and embracing telemedicine for head and neck cancer Cancer 126 2020 4092 4104 32639615
32 Matz M Allemani C van Tongeren M Excess mortality among essential workers in England and Wales during the COVID-19 pandemic J Epidemiol Community Health 76 2022 660 666 35470261
| 36455583 | PMC9704820 | NO-CC CODE | 2022-12-01 23:19:05 | no | Lancet Oncol. 2022 Dec 28; 23(12):e544-e551 | utf-8 | Lancet Oncol | 2,022 | 10.1016/S1470-2045(22)00505-8 | oa_other |
==== Front
Lancet Infect Dis
Lancet Infect Dis
The Lancet. Infectious Diseases
1473-3099
1474-4457
Elsevier Ltd.
S1473-3099(22)00756-3
10.1016/S1473-3099(22)00756-3
Comment
Role of immunity landscape in global risk assessment of re-emerging diseases
Pandey Abhishek a
Galvani Alison P a
a Center for Infectious Disease Modeling and Analysis (CIDMA), Yale School of Public Health, New Haven, CT 06520, USA
28 11 2022
28 11 2022
© 2022 Elsevier Ltd. All rights reserved.
2022
Elsevier Ltd
Elsevier has created a Monkeypox Information Center (https://www.elsevier.com/connect/monkeypox-information-center) in response to the declared public health emergency of international concern, with free information in English on the monkeypox virus. The Monkeypox Information Center is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its monkeypox related research that is available on the Monkeypox Information Center - including this research content - immediately available in publicly funded repositories, with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the Monkeypox Information Center remains active.
==== Body
pmcWith the continuous growth and connectedness of the human population, there is an increased threat of emerging and re-emerging diseases. These threats are often exacerbated by the environmental and sociopolitical crises facing humanity. The unprecedented speed at which the COVID-19 pandemic spread worldwide highlights the urgent need to address the issues that increase risks of disease emergence and spread. Such risks can be addressed through bolstering global surveillance and increasing concerted efforts to mitigate disease spread locally. After local emergence of any disease, an understanding of the geographical risks of widespread transmssion is fundamental for preparedness and resource allocation strategies to contain the disease before it becomes a global health emergency. For re-emerging diseases, it is crucial to evaluate the immunity landscape. In The Lancet Infectious Diseases, Juliana C Taube and colleagues1 use demographic modelling to generate a comprehensive and granular global immunity landscape for smallpox vaccination coverage. The authors developed an innovative approach that incorporates country-specific historical vaccination data with data on demographic growth to assess age-distributed pre-existing smallpox immunity, a proxy for orthopoxvirus immunity.
Since the identification of an initial cluster of monkeypox cases in the UK on May 14, 2022, monkeypox has been reported in more than 90 countries where it had not previously circulated.2 Several observational studies in past have shown that vaccination against smallpox might provide up to 85% protection against monkeypox for 3–5 years.3 Although routine vaccination against smallpox worldwide ceased by 1980, individuals who received smallpox vaccination might still have some protection against monkeypox. Therefore, understanding the smallpox vaccination coverage among the current population could be instrumental in identifying subpopulations most susceptible to infection and areas where widespread outbreaks are most probable. Accounting for the cross-immunity by smallpox vaccination and waning efficacy, Taube and colleagues1 combine country-specific data on smallpox vaccination coverage before cessation of the routine vaccination campaign in that country along with data on current demography to generate a global landscape of susceptibility to orthopoxviruses including for monkeypox.
Characterisation of the global spatial landscape for vulnerability to orthopoxviruses by Taube and colleagues1 showed that the age-specific coverage at the time of cessation of routine vaccination was the primary factor in determining the current spatial heterogeneity in susceptibility across the globe. Their finding was consistent with the observed median age of 37 (IQR 32–43) years for monkeypox infections globally,4 and suggested previous immunity through the smallpox vaccine might be protecting the older population from monkeypox infection. Similarly, the influenza A H1N1 pandemic in 2009 mostly affected the young population as the older population had previous immunity.5 An accurate understanding of the immunity landscape for H1N1, such as that generated by Taube and colleagues1 for smallpox, would have informed swift and optimal allocations of vaccines.6
The current multicountry monkeypox outbreak has disproportionately affected men who have sex with men. The rapid rollouts of vaccination with prioritisation of men who have sex with men in several countries have been effective in mitigating ongoing transmission.7 In addition to monkeypox, the risk of outbreaks from other orthopoxviruses, such as cowpox and buffalopox, has been increasing.8, 9 Beyond orthopoxviruses, global databases for immunity profiles of other vaccine-preventable diseases have become more crucial now than ever. Pandemic-driven restrictions, economic disruptions, and resource reallocations have caused reductions in routine immunisations for multiple diseases across the world.10 In 2021, 25 million children did not receive vaccines against preventable diseases such as measles and poliovirus.11 Moreover, misinformation about COVID-19 vaccines has exacerbated hesitancy towards vaccination more broadly.12 Although there remain only two poliovirus-endemic countries (Afghanistan and Pakistan), the recent identification of polio cases in non-endemic countries such as the USA and the UK underscores the importance of maintaining high vaccination coverage.13 Global assessment of the immunity landscape of vaccine-preventable diseases, as developed by Taube and colleagues,1 can identify locations that most need surveillance and proactive action.
Future extensions of the Taube and colleagues1 framework could capture the interplay of the immunity landscape with the shifting of viable geographical range for pathogens to identify areas that are at greatest risk of disease emergence and spread.14 Another dimension that could be incorporated in extensions of the framework is the sociopolitical factors that can affect transmission rates. For example, political destabilisation coinciding with drought from climate change has stoked a devastating cholera outbreak in Yemen, with disease fatality rates compounded by malnutrition.15 The foundational study by Taube and colleagues1 paves the way to develop novel approaches to understanding the complex interplay between immunological, climate, and sociopolitical systems that can improve prediction and proactively curb disease outbreaks worldwide.
IRF - Frederick 01/10/11 9:52:48 a 21000 7.0 80.0 Imaging #10-044-2 #10018 Lymph Node OSD 13E -881.32668 -38.949968 0.01 XpixCal=342.667 YpixCal=342.667 Unit=um ##fv3© 2022 Flickr - NIAID
2022
We declare no competing interests.
==== Refs
References
1 Taube JC Rest EC Lloyd-Smith JO Bansal S The global landscape of smallpox vaccination history and implications for current and future orthopoxvirus susceptibility: a modelling study Lancet Infect Dis 2022 Published online Nov 28, 2022. 10.1016/S1473-3099(22)00664-8
2 Reuters Factbox: monkeypox cases and deaths around the world Oct 28, 2022 Reuters https://www.reuters.com/business/healthcare-pharmaceuticals/monkeypox-cases-around-world-2022-05-23/
3 WHO Monkeypox https://www.who.int/news-room/fact-sheets/detail/monkeypox May 19, 2022
4 WHO Multi-country monkeypox outbreak: situation update https://www.who.int/emergencies/disease-outbreak-news/item/2022-DON393 June 17, 2022
5 US Centers for Disease Control and Prevention 2009 H1N1 Pandemic (H1N1pdm09 virus) https://www.cdc.gov/flu/pandemic-resources/2009-h1n1-pandemic.html June 11, 2019
6 Medlock J Galvani AP Optimizing influenza vaccine distribution Science 325 2009 1705 1708 19696313
7 Kupferschmidt K Monkeypox cases are plummeting. Scientists are debating why https://www.science.org/content/article/monkeypox-cases-are-plummeting-scientists-are-debating-why Oct 26, 2022
8 Eltom KH Samy AM Abd El Wahed A Czerny C-P Buffalopox virus: an emerging virus in livestock and humans Pathogens 9 2020 E676
9 Vorou RM Papavassiliou VG Pierroutsakos IN Cowpox virus infection: an emerging health threat Curr Opin Infect Dis 21 2008 153 156 18317038
10 UNICEF COVID-19 pandemic fuels largest continued backslide in vaccinations in three decades https://www.unicef.org/press-releases/WUENIC2022release July 14, 2022
11 Guglielmi G Pandemic drives largest drop in childhood vaccinations in 30 years Nature 608 2022 253 35883006
12 Opel DJ Brewer NT Buttenheim AM The legacy of the COVID-19 pandemic for childhood vaccination in the USA Lancet 2022 published online Oct 26. 10.1016/S0140-6736(22)01693-2
13 McKenna M Polio is back in the US and UK. Here's how that happened Aug 24, 2022 Wired https://www.wired.com/story/polio-is-back-in-the-us-and-uk-heres-how-that-happened/
14 Mora C McKenzie T Gaw IM Over half of known human pathogenic diseases can be aggravated by climate change Nat Clim Chang 12 2022 869 875 35968032
15 UN Yemen hit by world's worst cholera outbreak as cases reach 200,000 https://news.un.org/en/story/2017/06/560332 June 24, 2017
| 36455592 | PMC9704845 | NO-CC CODE | 2022-12-01 23:19:06 | no | Lancet Infect Dis. 2022 Nov 28; doi: 10.1016/S1473-3099(22)00756-3 | utf-8 | Lancet Infect Dis | 2,022 | 10.1016/S1473-3099(22)00756-3 | oa_other |
==== Front
Sci Afr
Sci Afr
Scientific African
2468-2276
The Author(s). Published by Elsevier B.V. on behalf of African Institute of Mathematical Sciences / Next Einstein Initiative.
S2468-2276(22)00385-4
10.1016/j.sciaf.2022.e01480
e01480
Article
Are Twitter sentiments during COVID-19 pandemic a critical determinant to predict stock market movements? A machine learning approach
Jena Pradyot Ranjan ⁎
Majhi Ritanjali
School of Humanities, Social Sciences and Management, National Institute of Technology Karnataka, Surathkal, Mangalore 575025, India
⁎ Corresponding author.
29 11 2022
3 2023
29 11 2022
19 e01480e01480
18 8 2022
22 11 2022
28 11 2022
© 2022 The Author(s)
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
The problem of stock market prediction is a challenging task owing to its complex nature and the numerous indirect factors at play. The sentiments regarding socio-political issues such as wars and pandemics can affect stock prices. The spread of the COVID-19 pandemic continues to take a toll on the economy and fluctuations in sentiment of the concerns about the health impacts of the disease can be captured from the microblogging platform, Twitter. We examined how these sentiments during the Covid-19 pandemic and the health impacts arising from the disease along with other macroeconomic indicators provide useful information to predict the stock indices in a more accurate manner. We developed a machine learning model namely, long-short term memory (LSTM) networks to predict the impact of the Covid-19 induced sentiments on the stock values of different sectors in the United States and India. We did the same predictions using the timeseries statistical models such as autoregressive moving average model and the linear regression model. We then compared the performance of the LSTM and the timeseries statistical models to find that the machine learning model has produced more accurate predictions of the stock indices. The performance of the models across the sectors and between the United States and India are compared to draw economic inferences.
Key words
Sentiment analysis
COVID-19
Stock market
LSTM
Twitter
Editor by Pradyot Ranjan Jena
==== Body
pmcIntroduction
The stock market has been observed to be a volatile signal and predicting these stock market movements has been the focus of analysts. Various models have been developed with the aim of capturing the patterns in stock market prices/indices Chen et al. [12] [10,11,13,[34], [35], [36]]). The most common approach in stock market analysis literature is the Technical analysis [1,2,11] which uses the stock prices or the derivatives of it as input. The underlying assumption is that the stock prices represent the macroeconomic indicators and news. Hence, the information embedded in stock prices is enough to predict the stock market. Macroeconomic variables like Gross Domestic Product, interest rates, currency exchange rates, customer price index, among others have been commonly used for stock indices forecasting [6,43]. Furthermore, text mining techniques are in use to include information such as financial news. However, with the advent of social network analysis more recently, using sentiment indices and other derived series as inputs, has proved useful for the stock forecasting ([4]; Barber and Terrance, [3].
Global Health Emergency of the COVID-19 outbreak was declared by the World Health Organization (WHO) on 30th, January 2020 [17,39]. With over 7 million confirmed cases and 400,000 deaths worldwide, the pandemic continues to have impact on the economy [45]. Some of these impacts include the spike in unemployment in the US which is over 20 million, where the prospect of a large number of marginalized people are pushed to poverty. In addition, COVID-19 has significant implication for the performances of the financial markets.
Twitter is an online microblogging platform and acts as a platform where an overall picture of the mood and general sentiment about specific topics can be observed. Twitter users across the world have been closely following the trends of the spread of the pandemic and share their views on the platform. Owing to the lockdown, more people are at home which explains the increased usage of the platform. In this paper, we study to what extent the twitter sentiment during the pandemic can help predict the stock indices in the US and India across a variety of sectors. Furthermore, the trends in healthcare conditions can help for useful information about the performance of the economy. The interpretation of the healthcare trends by people that produce expectations about the performance of the economy is also a crucial feature that we try to leverage.
The previous studies have shown the impact of monetary policy on stock movements. Sirucek [38] has used the DJIA index to represent stock market movement and M2 and MZM aggregates as the money supply to study the co-movement in the US. Li [23] showed how the money supply in Europe during the 2009 crisis had influenced the stock market. They used the Granger causality test and Vector Error Correction Model to shed light on both the short term as well as the long-term inter-relation. Studies have also examined the effect of floating foreign exchange rates on the stock market prices (Mechri et al., [24]). They used the monthly data on stock prices, gold prices, exchange rates and inflation rates for Tunisia from 2002 to 2017. Using the GARCH and multiple regression models, they found that the exchange rate does indeed have an effect on movement of the gold and oil prices and stock market prices. Zhou et al. [48] used the daily stock market (high frequency) data for the year 2016 and the Generative Adversarial Networks model using the Long Short-Term Memory (LSTM) and Convolutional Neural Networks, to predict the stock market prices in China.
Apart from these conventional economic determinants of the stock market performance, the sentiments and expectations have been used to predict stock indices. Bouktif et al. [5] have examined the use of sentiment features and provided an intricate pipeline to extract sentiment from tweets in the context of stock market. They used stock prices for 10 companies from NASDAQ from the year 2008 to 2018. A comparison between various models such as Logistic regression, Random Forest and Neural Networks is shown and the study highlighted how sentiment features help to increase the accuracy of the model. The complexity of the stock market signal is handled in Jin et al. [21] by using the Empirical Modal Decomposition method before feeding the features to the LSTM model which uses its memory units to optimize for timeseries data. Their work shows that the model also decreases the training time as well as increases prediction accuracy. The Stock tweets platform is used to scrape text and extract sentiment for the period from 2013 to 2018.
Furthermore, Twitter was used as the data source to monitor the public reaction and health during disasters e.g. hurricanes [22,37,47,48], floods [18], earthquakes [15], terrorist bombing Buntain et al. [9] public health related misinformation propagation [8,29,40] and others [19,42,44] and disease outbreaks [14,26,28,41].
Against this backdrop, our primary objectives in this paper are – (i) to study if Twitter sentiment and healthcare data during the pandemic periods are useful factors to predict stock market index movements, (ii) to compare the performance of different models for the prediction, and (iii) to study how the regression performance differs with respect to different sectors in United States (US) and India. The data used in this study are mainly of 4 kinds: Daily economic indicators, COVID-19 Health, Twitter Sentiment, Stock Index/ Price Related. Since our period of study is constrained to only 5 months, in order to have sufficient data points we use economic indicators that are available on a daily basis. This restricted us from using many other important indicators which do not have a daily frequency.
The results show that in this case, LSTM model shows better performance when compared to linear regression and ARIMA which may be due to its ability to capture nonlinear relationships between the features especially with respect to sentiment. When the sentiment-based features are used, the prediction ability of the model is improved and hence the sentiment features may provide additional useful information to model the stock price. The sentiment relationship with index is stronger in case of Finance and Consumer Goods sectors indices which have in general been more affected due to the pandemic unlike the Technology and Healthcare sectors. The relationship is in general stronger in US than in India which can be explained by greater percentage of people who are active on Twitter. The model performs best when the time lag is 1 day that is if we use the twitter sentiment of the previous day to predict the current stock price movement.
Materials and methods
The empirical analysis is carried out following the work-flow schematic presented in Fig. 1 .Fig. 1 Flowchart of the proposed methodology.
Fig 1
Data
The data is mainly collected for the period January - May 2020 (daily) as this was when the pandemic spread throughout the world and its effects on the economy were clearly evident.
Healthcare Data: The healthcare data was collected from [46] which is known to be a standard and accurate source of COVID-19 healthcare data. Apart from the global data, the dataset was grouped according to the country. The main variables that we use/ derive for our study are - daily number of active cases, number of deaths and number of recovered cases. We also calculate the corresponding daily rates for each of the mentioned variables. The testing coverage data is used to normalize the variables so as to level out the disparity in the number of cases which may be due to a low testing coverage in the country.
Economic: The Indian economy indicators used are mainly collected from the RBI's daily press releases on Money Market Operations, Weekly Statistical Supplement [31] and India Government Bonds India Government Bonds, [20] which provides the daily Bond Yield Rates. The corresponding data for the US are collected from [32]. Foreign exchange rate is one of the variables considered as India has a floating exchange rate which depends on the demand and supply and is a good indicator of the economy.
Apart from this, the other indicators that are used are mainly related to the Money Market. The central bank of most countries maintains the liquidity which is the cash available in the economy by using monetary policies. The various measures taken, affect the way in which Banks respond with changes in interest rates among others. It also helps manage the amount of inflation and activity in the economy. It has short term effects on financial markets. Repo combines lending-borrowing and sale-purchase transactions. The securities are sold in exchange for cash and these securities can then be repurchased by the original holder. The reverse repo is used to reduce or absorb liquidity. Thus, we consider those variables that vary on a daily basis and therefore do not include important rates such as the Repo Rate, Reverse Repo Rate among others. It should be noted that The Net Liquidity injections also include other operations done to change liquidity but which may not be done on a regular basis. The central bank uses bond to borrow money. Government bonds (G-secs/ Treasury) are considered to be very safe investments. The yield rate changes with the price of the bond. The rise and fall of yields capture the expectation of stakeholders about the future growth of the economy. For example, the 10Y bond yield rate in the US has decreased owing to the worsening of the expectations of the economy due to COVID-19. Many of these factors are interdependent on each other by being only correlated or by having a causal relationship with a certain time lag. Since the time lag cannot be estimated directly we try different combinations and experiment to get the best results. The foreign exchange rates depend on various variables including the overall economic activity, prospects, etc. several factors a nation's economic activity and growth prospects, interest rates, and geopolitical risk. Thus, the following variables are used: [Reverse Repo Operations (INR), Net Liquidity Injection (INR or $), Cash Balance (INR or $), 10Y Bond Yield Rates (%), 5Y Bond Yield Rates (%), 1Y Bond Yield Rates (%), Forex Rate ($/ INR)]
Stock Data: The data consists of the index values for major stock markets in the US and India collected from [33] and (NSE, [27]). The sector wise index (DJIA and Nifty) data is collected for the Financial, Consumer Goods, Healthcare and Technology sectors. Fig. 2 shows the timeseries data of Dow Jones U.S. Technology Index starting from 6 June 2019 to 6 June 2020.Fig. 2 Dow Jones U.S. Technology Index starting from 6 June 2019 to 6 June 2020.
Fig 2
Twitter Data: The twitter data was mainly collected from (PanaceaLab, [30]) and consisted only of tweet IDs. The tweets were categorized based on the location into Indian, US and others. The tweets were mainly filtered based on the keywords such as ’COVID-19′, ’Corona’, ’Wuhan’, etc. The number of tweets is around a million.
Methods
Data pre-processing
Data pre-processing consists of various operations. Better predictive models can be built more efficient in terms of convergence and prediction using pre-processing operations. Some of the techniques that are commonly used with respect to numerical type data (since our data is numeric type) have been discussed below:• Minmax scaling for numerical features using normalization, feature or sample wise (done by setting the axis of normalization). This is done by computing the overall minimum and maximum values of the feature taking all samples into consideration.
• Standard scaling (z-score normalization) numerical features using µ and σ computed on the training dataset i.e. subtracting mean and dividing by standard deviation. It is also called Zero-mean scaling.
• Bucketizing features (of numeric type) using quantiles.
• Filling in missing values in the data using the median (for numerical features), mean or mode (for categorical features).
• Computing the PCA of the input features to project the data into a lower dimensional space (with linearly dependent features) to reduce the input size and prevent overfitting.
The min-max normalization procedure is used in this study to convert the data in the range between 0 and 1. We apply data pre-processing techniques on all the stock price of the indices that we collected for Indian and U.S. sectors. The DJIA and Nifty sector-wise indices are normalised using the min max scaling. After this operation, the stock prices obtained are in the range 0 to 1. This technique performs better if the distribution is not Gaussian or the standard deviation is very small. The sentiments obtained from the sentiment analyser are already scaled from −1 to +1, hence they do not require any scaling as such. The health data for number of COVID cases is also scaled before using in the model.
Sentiment analysis
Using the tweet ids based on the location, we use the Twitter Search API to fetch the tweet contents. The sentiment is based on 3 dimensions: Valence, Arousal and Subjectivity which are all values from −1 to 1. By projecting the emotion along 3 different dimensions we can get useful information about the different aspects of the overall sentiment. Valence is a measure of how positive or negative the tweet is, Arousal is the intensity of the emotion and Subjectivity is used to ascertain if the text is an opinion or factual content. Each tweet must be pre-processed through an elaborate pipeline so as to get the best features and to remove noise. Some of the pre-processing includes:• Removal of stop words which removes some of the common words such as ‘in’, ‘a’, etc. which do not add useful information.
• Named Entity and Parts of Speech tagging.
• Many tweets carry many slang words and contracted forms of words. We use a slang word dictionary to replace slang words and expressions to appropriate meaningful words.
• The contracted words are expanded using a custom dictionary that maps the contracted versions of the word to the expanded version of the word(s).
• Correcting misspelled words and malformed sentences by replacing stray characters, URLs, username mentions, etc.
• We check if a particular tweet has a hashtag which corresponds to an emotional word and use this as well to predict sentiment.
• Sentences with lot of trailing shows high arousal thus this is taken into account.
• Emoticons are mapped to emotional states.
The Affective Norms for English Words (ANEW), consists of around 1000 words which are manually and accurately annotated a mean valence and arousal score, by experts [7]. The AFINN dataset, [16] also provides a similar list but with only Valence scores. These two lists were further extended by using the WordNet [25]. For every word in ANEW and AFINN, we add the same scores for all words in the synset of the word. This way we have a much larger list of EUs (Emotional Units). A valence and arousal score can be directly assigned to the words in the tweets that are present in the expanded ANEW and AFINN.
Principal component analysis
The Principal Component Analysis (PCA) is a linear method of dimensionality reduction that forms the underlying basis of multivariate data analysis based on projection methods. In PCA, the data is projected along the directions of in which there is maximum variance. Eigenvalues and eigenvectors of the covariance (see eqn. (1)) matrix of data are computed and each eigenvector corresponds to a principal component. The suitable number of features that can be extracted from this method is given by the number of the non-null eigenvalues or the number of principal components that can explain a statistically significant amount of variance in the data, generally kept as 99% explained variance. The functional form of PCA is –(1) Cov(X,Y)=1n−1Σi=1n(Xi−x¯)(Yi−y¯)
This method gives us a projection of the features (or high dimensional data points) in a vector space of a lower dimension. The principal components are mutually orthogonal. Fig. 3 shows the principal components for a dataset in 2 dimensions or a dataset with two features. In this work, we employ PCA to reduce the number of features to a few so that the correlated features can be minimized and only essential regressors can be considered for the prediction or classification model.Fig. 3 Principal Components.
Fig 3
Auto regressive integrated moving average (ARIMA) model
Time Series forecasting can be of tremendous commercial value. Businesses use it for predicting demand and sales to manage the procurement of resources. The regression function can be written as a variable defined as a linear combination of the others -(2) Y=β0+β1X1+β2X2...βnXn
Regression can be linear or non-linear, depending on the variables and their correlation with other considered features. When considering a time series type of data, regression is used with respect to forecasting. It can be primarily of two types:• Uni-variate Time Series Forecasting: The previous values of the time series alone are used to predict its future values.
• Multi-Variate Time Series Forecasting: Predictors other than the past values of the series (also called as exogenous variables) are used for forecasting.
For our analysis, we use the Auto Regressive Integrated Moving Average (ARIMA) model and the linear regression model, two of the most used regression models under the univariate and multivariate timeseries data respectively.
An ARIMA, is a type of forecasting algorithm model used for statistical analysis. It uses time-series data to decipher patterns so as to predict future trends. While seasonal or periodic data can be best described by exponential smoothing models, ARIMA models describe the autocorrelations in the data. Here, we limit our experimentation to only non-seasonal ARIMA models. To describe the model, we segregate the terms in it and study each term individually.
Autoregression model: Forecasting a variable using a regression of past values of the variable as regressors, hence the term “auto” regressors. Thus, an autoregressive model of order p can be written as:(3) yt=c+φ1yt−1+φ2yt−2+···+φpyt−p+εt,
where, εt is white noise and yt-I denotes the ith backward lag of yt. We refer to this as an AR(p) model, an autoregressive model of order p, also called as the lag order.
Moving average models: It uses past forecast errors in a regression-like fashion. It is referred to as an MA(q) model of order q, wherein q refers to the number of lag error values used.(4) yt=c+εt+θ1εt−1+θ2εt−2+···+θqεt−q,
Furthermore, stationarity is a desired property in timeseries data for it to be used in forecasting. A time series is called stationary if its properties do not depend on the time at which the series is observed. Often, the timeseries data in level are not stationary. Hence to convert them into stationary series, the raw observations are differenced. The symbol d is used to show the number of times that the raw data are differenced. It is also known as the degree of difference (d).
Giving the input parameters as p, q and d, we can estimate the regression coefficients and assess if past values of stock do indeed have an effect on the present value. Since past literature on the study of stocks with respect to this model suggests short term causations, we used d as 1, i.e., single-differencing of raw data and q as 0, i.e., no past error is used. However, the p-value depends on the stock being assessed. So, we estimated the p-value for each stock such that the stock behaves as stationary. Apart from the AR and MA terms, other independent variables are included in the ARIMA model estimated in this study.
Long-Short-Term-Memory (LSTM) networks
Traditional neural networks cannot learn from past events and are not designed to persist information to use as a formulation for understanding current or future events. This seems like a major shortcoming. Recurrent neural networks (RNN) address this issue as they are networked with loops in them, allowing information to persist. A recurrent neural network can be thought of as multiple copies of the same network, each passing a message to a successor.
However, sometimes, we only need to look at recent information to perform the present task instead of the total past events. This issue of Long-term dependencies is addressed by Long-Short-Term-Memory networks (LSTM's). LSTM can help us model non-linear dependencies. This is done by using a series of ‘’gates” (Fig. 4 ). These gates are contained in similar memory blocks which are connected layer by layer:Fig. 4 The architecture of LSTM.
Fig. 4:
An LSTM unit consists of three types of gates: Input Gate: Scales input to cell (write) Output Gate: Scales output to cell (read) Forget Gate: Scales old cell value (reset). Each gate can be thought of as a switch that controls the read or write, incorporating the long-term memory function into the model (Fig. 4).
The present study applies several inputs like twitter sentiments, economic indicators, healthcare indicators and stock market indicators to LSTM. Iterative LSTM parameter tuning was performed for data fitting. Several hyperparameters were evaluated to identify the optimal LSTM architecture that provides ideal assessment metrics. The hyperparameters include:1. number of LSTM layers;
2. number of nodes in each layer;
3. number of fully connected layers;
4. types of activation function;
5. number of dropout layers and percentage of dropout;
6. learning rate;
7. loss function;
8. optimizer;
9. batch size; and
10. number of epochs
The LSTM structure for the study comprises the following layers –1. LSTM with 512 nodes.
2. Fully connected layers with 64 nodes and ReLU activation function
3. Dropout with 0.4%.
4. Fully connected layers with 1 node and Linear activation function.
The final hyperparameters are:1. the learning rate is 0.001;
2. loss function of mean absolute error (MAE);
3. optimizer: ADAM;
4. epochs is 500; and
5. the batch size is 8
Though LSTM's and ARIMA might look the same in principle, both have their pros and cons. The advantages of using ARIMA over LSTM's are:• Simple to implement, no parameter tuning
• Easier to handle multivariate data
• Quick to run
However, LSTM's have proven to show better results more often and are hence more popular in time-series modeling for the following reasons:• No pre-requisites (stationarity, no level shifts)
• Can model non-linear function with neural networks
• Needs more data
Results and discussion
The sentiment analyser is able to accurately capture the trends which vary according to the healthcare events that dominate every user's Twitter feeds. In Fig. 5 , it can be seen that on the day the WHO declared COVID-19 as an international concern and a health emergency, there is a sudden dip in the emotion i.e. there is a sudden increase in negative sentiment (Fig. 6 ).Fig. 5 .Variation of Valence.
Fig. 5
Fig. 6 Sample Tweet.
Fig 6
The following is a sample run of the sentiment analyser.
Valence: −0.31
Arousal: +0.44
Subjectivity: 0.78
The analyser is able to accurately capture the negative sentiment (negative valence), the intensity of the emotion (high arousal) and the opinionated nature of the text (high subjectivity).
Next, we perform correlation analysis by computing the Pearson correlation coefficients between the different regressors. The value of the coefficient primarily captures the variation along the line of best fit. We use 16 samples to observe any trend between the features. From the analysis, we aim to determine if there exists a linear or direct correlation between the stock prices and the twitter sentiment. We get a value of 0.634 as the correlation coefficient, which might not be as too significant for small data size. It is possible that the same value of 0.6 would be highly significant when the size of the data is huge. However, since our sample size is small, we cannot derive any strong inference about the association between the sentiments and the stock prices and further analysis is needed.
It is also possible that the features might be non-linearly related and since the Pearson correlation can only capture linear association, it is not able to produce a statistically significant result. This indicates that LSTM might have an advantage over linear techniques like ARIMA and LR in predicting the stock, if there is any sort of correlation between the features (sentiment and stock prices) at all.
We discuss three potential models with respect to stock price prediction. However, for our analysis, it would be difficult to interpret and manage results from all these models. Hence, we decide to derive further results by choosing the most suitable model for our data. This was implemented by running a sample analysis and comparing the Root Mean-Squared Error from all models. At last, we choose the model with the least RMSE. Using all the predictors as input, we fit all the models for finding the model with best fit.
For ARIMA, we try with values of p from 0 to 6 and derive the best fit. Finally, we use the p, q, and d values as 1,1 and 0 respectively. Standard values of q and d are 1 and 0 respectively. For ARIMA, we use the stats model python library and multivariate ARIMA where we pass the other features as a numpy array argument to the function separately (exogenous variables).
For LSTM, its multiple input series method is used where all the features are given as inputs to form a 2D array, derived at the same time step (i.e.1day). A simple LSTM with the standard architecture is used without many LSTM layers. The used architecture is shown in the Fig. 4. The optimizer used is Adam and the loss function as Mean Squared Error (MSE). It is run for a total of 30 epochs as the model can be seen to converge by 10 epochs. the linear regression is run for simple OLS estimate.
As can be seen from Table 1 , the LSTM model has the highest accuracy in predicting both Indian and U.S. stock index values of the technology companies, as this model has the lowest RMSE value compared to both linear regression and ARIMA models. This is in conjunction with the past studies that prove the credibility of LSTM for its higher stock price prediction ability. Linear regression model performs worse than ARIMA, possibly because the ARIMA can predict based on previously observed values of the time series (though linear) and is able to map more complicated relations. The success of LSTM and ARIMA over LR depict that there is indeed some dependence of the stock price on its previous values. From here on, we show all results using the LSTM model.Table 1 RMSE for different models.
Table 1 ARIMA Linear Regression LSTM
US-Tech 1.047 1.248 0.8744
India-Tech 1.153 1.290 0.5364
Note: RMSE values are of order 103, & MSE of the order 106.
Further, to derive whether sentiments do have some correlation with the stock price, we fit the model with all the features and one without sentiment features. Tables 2 and 3 report the RMSE values from prediction of stock index values of finance and technology sectors respectively. We find that RMSE increases when the sentiment-based features are not used. This helps us to conclude that the twitter sentiments do have some impact on the stock price during this pandemic. This can be supported by the fact that due to social distancing more and more people expressed their opinions through social media platforms which is in turn affecting the sentiment of investors more profoundly. This effect is more evident in U.S. than in India, possibly because U.S. has a higher number of people expressing themselves over social media. This has been corroborated with two industries, namely Finance and Technology (Tables 2 and 3). We infer that the Tech industry shows lesser improvement in the RMSE than the Finance. This may indicate that the Finance industry is likely more affected in the pandemic and is more correlated to public's current sentiments.Table 2 RMSE of LSTM model for Finance Sector in India and U.S.
Table 2: India- Finance U.S.- Finance
Without Sentiment features 0.5142 0.5109
With Sentiment features 0.4017 0.3010
Principal components 0.4229 0.3119
Note: RMSE values are of order 103, & MSE of the order 106.
Table 3 RMSE of LSTM model For Technology Sector in India and U.S.
Table 3: Indian-Tech U.S. Tech
Without Sentiment features 0.9102 0.6917
With Sentiment features 0.8018 0.5075
Principal components 0.8744 0.5364
Note: RMSE values are of order 103, & MSE of the order 106.
Next, we performed the PCA to reduce the feature size to a manageable count. Initially, the feature set size is 16. After observing the features closely, 3 broad categories of features are found, namely healthcare, sentiment and macroeconomic indicators. Therefore, we perform PCA to reduce the feature set to 3. To support our assertion, the RMSE values of the prediction models before and after applying PCA are derived. As expected, the RMSE increases as the principal components are used instead of the original features as dimensionality reduction often leads to loss of desirable information. However, the increase in RMSE can be traded off for the lesser features.
We analyze how each sector in India and U.S. is affected by the pandemic, and more importantly how the twitter sentiments impact the same. As shown in Table 4 , in the Indian scenario, the stock indices of the Consumer goods and Finance sectors are much more determined by the COVID-19 impact as compared to the stock indices of the Pharma and Technology sectors, with consumer goods being the most affected. The RMSE value of the LSTM prediction model is lowest for the Consumer goods sector followed by Finance sector. Similarly, the R2 value, which is used as a measure of goodness of fit of a model, is highest in case of Consumer goods sector followed by the Finance sector. In the U.S scenario, as depicted by Table 5 , the Financial Sector is most impacted by COVID-19 sentiments. Furthermore, we examined with what time lag the Twitter sentiment must be used to achieve most accurate prediction and found out that a lag of 1 day in Twitter sentiment is most useful in predicting the current stock market movement. Thus, this study sheds light on several interesting aspects of movement of the economic market during the COVID-19 pandemic. Although the results here do not derive a strong relationship between the factors, they provide further evidence and show scope for providing useful information to help predict economic movement which can further be taken up in subsequent studies.Table 4 R2, RMSE and MSE Metrics of LSTM model for Indian Sectors.
Table 4: R2 RMSE MSE
Financial 0.5647 0.4229 0.1788
Pharma 0.5315 0.5654 0.3196
Tech 0.5019 0.8744 0.7645
Consumer Goods 0.5783 0.3723 0.1386
Note: RMSE values are of order 103, & MSE of the order 106.
Table 5 R2, RMSE and MSE Metrics of LSTM model for U.S. Sectors.
Table 5: R2 RMSE MSE
Financial 0.5710 0.3119 0.0972
Pharma 0.5475 0.4634 0.2147
Tech 0.5142 0.5364 0.2877
Consumer Goods 0.5333 0.4923 0.2423
Note: RMSE values are of order 103, & MSE of the order 106.
Conclusion
In this study, we evaluated the usefulness of the Twitter sentiment for improving the prediction of stock market indices of several sectors in the U.S. and India. We described different methods involved in extracting the sentiment, pre-processing and the techniques applied for using these models to predict the stock market indices of four sectors. Mainly, the recurrent neural network model, namely the LSTM and the conventional linear timeseries models such as ARIMA are compared for their prediction accuracy. The LSTM model is found to be more accurate across the sectors. The variation of the results across different sectors has been examined and it is found that Finance and Consumer Goods sectors have a stronger relationship with the health and Twitter sentiment features more than the Technology and Pharmaceutical sectors. This may be due to the fact that these sectors are more adversely affected due to the pandemic. Especially, the Consumer Goods sector felt a huge demand side shock as consumers across the globe either experienced a complete lockdown and hence, not allowed to visit shops and malls, or were subjected to movement restriction that resulted in a drastic reduction in demand. Supply chain disruptions due to the pandemic severely affected the movement of consumer goods from production to distribution centres further affecting the consumer sentiment. Similarly, the Finance sector was also affected by the negative sentiments of the investors who felt safe to hold on to their money rather than investing in different financial instruments. On the other hand, since the Pharmaceutical and Technology sectors were mostly seen as the ones that developed products that can help people overcome the pandemic effects. Hence, the negative sentiments for these sectors were significantly less. Furthermore, the relationship between twitter sentiments and stock market performance was stronger in the case of US which denotes the higher percentage of people in that country who actively engage with the Twitter platform.
The contribution of the current study is it establishes a significant link between non-monetary factor such as social media sentiments and stock market variations. Though, some of the past studies have examined this link, they mostly used the linear timeseries models. We developed a nonlinear neural network model that could use the nonlinearity in the data and predicts with higher accuracy. Future work will mainly involve studying more sectors and using stronger economic indicators. Moreover, the interpretability of the models can be improved by trying other neural network models.
Data availability statement information
The data that support the findings of this study are available from the corresponding author upon request.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
==== Refs
References
1 Atsalakis G. Valavanis K. Surveying stock market forecasting techniques - Part II: Soft computing methods Expert Syst. Appl. 36 3 PART 2 2009 5932 5941 10.1016/j.eswa.2008.07.006
2 Aguilar-Rivera R. Valenzuela-Rendón M. Rodríguez-Ortiz J.J. Genetic algorithms and Darwinian approaches in financial applications: a survey Expert Syst. Appl. 42 21 2015 7684 7697
3 Barber B. Terrance O. All that glitters: the effect of attention and news on the buying behavior of individual and institutional investors Rev. Financ. Stud. 21 2008 785 818
4 Bollen J. Mao H. Zeng X. Twitter mood predicts the stock market J. Comput. Sci. 2 1 2011 1 8
5 Bouktif S. Fiaz A. Awad M. Augmented textual features-based stock market prediction IEEE Access 8 2020 40269 40282 10.1109/ACCESS.2020.2976725
6 Boyacioglu M.A. Avci D. An adaptive network-based fuzzy inference system (ANFIS) for the prediction of stock market return: the case of the Istanbul stock exchange Expert Syst. Appl. 37 12 2010 7908 7912
7 Bradley M.M. Lang P.J. Affective Norms for English Words (ANEW): instruction manual and affective ratings IEEE Internet Comput. 12 5 2008 44 52 10.1109/MIC.2008.114
8 Broniatowski D.A. Jamison A.M. Qi S.H. AlKulaib L. Chen T. Benton A. Quinn S.C. Dredze M. Weaponized health communication: twitter bots and Russian trolls amplify the vaccine debate Am. J. Public Health 108 10 2018 1378 1384 10.2105/AJPH.2018.304567 30138075
9 Buntain C. Golbeck J. Liu B. LaFree G. Evaluating public response to the boston marathon bombing and other acts of terrorism through Twitter Proceedings of the International AAAI Conference on Web and Social Media 10 2021 555 558 10.1609/icwsm.v10i1.14774
10 Bustos O. Pomares-Quimbaya A. Stock market movement forecast: a systematic review Expert Syst. Appl. 156 2020 113464
11 Cavalcante R. Brasileiro R. Souza V. Nobrega J. Oliveira A. Computational intelligence and financial markets: a survey and future directions Expert Syst. Appl. 55 2016 194 211 10.1016/j.eswa.2016.02.006
12 Chen Qi Goldstein I. Jiang W. Price informativeness and investment sensitivity to stock price Rev. Financ. Stud. 20 3 2007 10.1093/rfs/hhl024
13 Choi N.Y. Sias R.W. Why does financial strength forecast stock returns? Evidence from subsequent demand by institutional investors Rev. Financ. Stud. 25 2012 1550 1587
14 Dredze M. Broniatowski D.A. Hilyard K.M. Zika vaccine misconceptions: a social media analysis Vaccine 34 30 2016 3441 3442 10.1016/j.vaccine.2016.05.008 27216759
15 Earle P.S. Bowden D.C. Guy M. Twitter earthquake detection: earthquake monitoring in a social world Ann. Geophys. 54 6 2011 708 715 10.4401/ag-5364
16 Finnrup Nielsen ’Making Sense of Micro posts’: big things come in small packages CEURWorkshop Proceedings 2011 93 98
17 Harapan H. Itoh N. Yufika A. Winardi W. Keam S. Te H. Megawati D. Hayati Z. Wagner A.L. Mudatsir M. Coronavirus disease 2019 (COVID-19): a literature review J. Infect Public Health 13 5 2020 667 673 10.1016/j.jiph.2020.03.019 32340833
18 Hirata E. Giannotti M.A. Larocca A.P.C. Quintanilha J.A. Flooding and inundation collaborative mapping – use of the Crowdmap/Ushahidi platform in the city of Sao Paulo, Brazil J. Flood Risk Manag. 11 2018 S98 S109 10.1111/jfr3.12181
19 Imran M. Mitra P. Castillo C. Twitter as a lifeline: human-annotated Twitter corpora for NLP of crisis-related messages Proceedings of the 10th International Conference on Language Resources and Evaluation, LREC 2016 2016 1638 1643
20 India Government Bonds - Investing.com India. (n.d.). Retrieved July 31, 2020, from https://in.investing.com/rates-bonds/india-government-bonds
21 Jin Z. Yang Y. Liu Y. Stock closing price prediction based on sentiment analysis and LSTM Neural Comput. Appl. 32 13 2020 9713 9729 10.1007/s00521-019-04504-2
22 Kabir M.Y. Madria S. CoronaVis: A Real-time COVID-19 Tweets Analyzer 2020 http://arxiv.org/abs/2004.13932
23 Li, Y. (2012). Empirical study on the relationship between money supply and stock market in Europe. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 7473 LNCS, 539–544. https://doi.org/10.1007/978-3-642-34062-8_70
24 MECHRI N. Ben Hamad S. de Peretti C. Charfi S. The impact of the exchange rate volatilities on stock markets dynamics: evidence from Tunisia and Turkey SSRN Electron. J. 2019 10.2139/ssrn.3304040
25 Miller G.A. WordNet: a lexical database for English Commun. ACM 38 11 1995 39 41 10.1145/219717.219748
26 Nagar R. Yuan Q. Freifeld C.C. Santillana M. Nojima A. Chunara R. Brownstein J.S. A case study of the New York City 2012-2013 influenza season with daily geocoded Twitter data from temporal and spatiotemporal perspectives J. Med. Internet Res. 16 10 2014 e236 10.2196/jmir.3416 25331122
27 NSE - National Stock Exchange of India Ltd. (n.d.). Retrieved July 31, 2020, from https://www1.nseindia.com/
28 Odlum M. Yoon S. What can we learn about the Ebola outbreak from tweets? Am. J. Infect. Control 43 6 2015 563 571 10.1016/j.ajic.2015.02.023 26042846
29 Oyeyemi S.O. Gabarron E. Wynn R. Ebola, Twitter, and misinformation: a dangerous combination? BMJ 349 October 2014 14 15 10.1136/bmj.g6178
30 PanaceaLab - COVID19 Twitter Dataset Homepage. (n.d.). Retrieved July 31, 2020, from http://www.panacealab.org/covid19/
31 Reserve Bank of India - Database. (n.d.). Retrieved July 31, 2020, from https://www.rbi.org.in/Scripts/Statistics.aspx
32 Resource Center. (n.d.). Retrieved July 31, 2020, from https://www.treasury.gov/resource-center/Pages/default.aspx
33 S&P Dow Jones Indices. (n.d.). Retrieved July 31, 2020, from https://www.spglobal.com/spdji/en/
34 Saleh T.A. Protocols for synthesis of nanomaterials, polymers, and green materials as adsorbents for water treatment technologies Environ. Technol. Innov. 24 2021 101821
35 Saleh T.A. Nanomaterials and hybrid nanocomposites for CO2 capture and utilization: environmental and energy sustainability RSC Adv. 12 37 2022 23869 23888 36093256
36 Saleh T.A. Global trends in technologies and nanomaterials for removal of sulfur organic compounds: clean energy and green environment J. Mol. Liq. 2022 119340
37 Sebastian A. Mobley W. Highfield W. Brody S.D. Leveraging Machine Learning and Twitter Data to Identify High Hazard Areas during Hurricane Harvey. c 2019 12549
38 Sirucek M. Munich Personal RePEc Archive Macroeconomic Variables and Stock Market : US Review Macroeconomic Variables and Stock Market : US review 39094 2012
39 Sohrabi C. Alsafi Z. O'Neill N. Khan M. Kerwan A. Al-Jabir A. Iosifidis C. Agha R. World Health Organization declares global emergency: a review of the 2019 novel coronavirus (COVID-19) Int. J.Surg. 76 February 2020 71 76 10.1016/j.ijsu.2020.02.034 32112977
40 Southwell B.G. Niederdeppe J. Cappella J.N. Gaysynsky A. Kelley D.E. Oh A. Peterson E.B. Chou W.-Y.S. Misinformation as a Misunderstood Challenge to Public Health Am. J. Prev. Med. 57 2 2019 282 285 10.1016/j.amepre.2019.03.009 31248741
41 Szomszor, M., Kostkova, P., & De Quincey, E. (2011). #Swineflu: twitter predicts swine flu outbreak in 2009. Lecture Notes of the Institute for Computer Sciences, Social-Informatics and Telecommunications Engineering, 69 LNICST (December 2009), 18–26. https://doi.org/10.1007/978-3-642-23635-8_3
42 Wang Z. Lam N.S.N. Obradovich N. Ye X. Are vulnerable communities digitally left behind in social responses to natural disasters? An evidence from Hurricane Sandy with Twitter data Appl. Geogr. 108 April 2019 1 8 10.1016/j.apgeog.2019.05.001
43 WEISBROD E. Stockholders’ unrealized returns and the market reaction to financial disclosures J. Finance LXXIV 2 2019 899 942
44 Wladdimiro D. Gonzalez-Cantergiani P. Hidalgo N. Rosas E. Disaster management platform to support real-time analytics Proceedings of the 2016 3rd International Conference on Information and Communication Technologies for Disaster Management, ICT-DM 2016 2017 10.1109/ICT-DM.2016.7857207
45 World Economic Outlook Update A Crisis Like No Other, An Uncertain Recovery June 2020
46 Worldometer Coronavirus CASES 2020 Worldometer 10.1101/2020.01.23.20018549V2
47 Yang J. Yu M. Qin H. Lu M. Yang C. A Twitter data credibility framework—Hurricane Harvey as a use case ISPRS Int. J. Geoinf. 8 3 2019 10.3390/ijgi8030111
48 Zhou X. Pan Z. Hu G. Tang S. Zhao C. Stock market prediction on high-frequency data using generative adversarial nets Math. Probl. Eng. 2018 10.1155/2018/4907423
| 36465525 | PMC9705006 | NO-CC CODE | 2022-12-06 23:15:06 | no | Sci Afr. 2023 Mar 29; 19:e01480 | utf-8 | Sci Afr | 2,022 | 10.1016/j.sciaf.2022.e01480 | oa_other |
==== Front
Psychoneuroendocrinology
Psychoneuroendocrinology
Psychoneuroendocrinology
0306-4530
1873-3360
The Authors. Published by Elsevier Ltd.
S0306-4530(22)00333-X
10.1016/j.psyneuen.2022.105992
105992
Article
Increases in stress hormone levels in a UK population during the COVID-19 pandemic: A prospective cohort study
Jia Ru a
Ayling Kieran a
Coupland Carol a
Chalder Trudie b
Massey Adam c
Nater Urs d
Broadbent Elizabeth e
Gasteiger Norina ef
Gao Wei g
Kirschbaum Clemens g
Vedhara Kavita a⁎
a Centre for Academic Primary Care, School of Medicine, University of Nottingham, University Park, Nottingham NG7 2RD, UK
b Department of Psychological Medicine, Institute of Psychiatry, Psychology & Neuroscience, King's College London, 16, De Crespigny Park, London SE5 8AF, UK
c Cortigenix, Cortigenix Laboratory, 6 Westhill Court, Walcott, Lincoln LN4 3BU, UK
d Department of Clinical and Health Psychology, University of Vienna, Liebiggasse 5, 1010 Vienna, Austria
e Department of Psychological Medicine, University of Auckland, Private bag 92019, Auckland, New Zealand
f School of Health Sciences, University of Manchester, Oxford Road, Manchester M13 9PL, UK
g Faculty of Psychology, TU Dresden, 01062 Dresden, Germany
⁎ Corresponding author.
29 11 2022
2 2023
29 11 2022
148 105992105992
9 8 2022
26 10 2022
24 11 2022
© 2022 The Authors
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Background
Research suggests that psychological factors may influence vulnerability to SARS-CoV-2 infection, although the mechanisms are unclear.
Purpose
We examined whether the hypothalamic-pituitary-adrenal axis may be a possible mechanism, by measuring the relationship between indices of psychological distress and cortisone in hair (hairE) in a UK cohort during the COVID-19 pandemic.
Methods
Participants (N = 827) provided two 3 cm hair samples over a 6-month period between April-September 2020. Samples reflected hairE in the 3 months prior to the collection date.
Results
HairE in the first samples (T1: commenced April 2020) did not differ significantly from pre-pandemic population norms. However, hairE in the second samples (T2: commenced July 2020) were significantly higher than T1 and pre-pandemic population norms, with a 23% increase between T1 and T2. Linear regressions, controlling for age and gender, demonstrated that at both timepoints, hairE levels were greatest in people with a history of mental health difficulties. In addition, stress reported at T1 predicted greater hairE at T2 and a greater change in hairE between T1 and T2.
Conclusions
These findings demonstrate that during the COVID-19 pandemic hairE was substantially elevated across a large community cohort, with greatest levels in those with a history of mental health difficulties and greatest changes in those reporting greatest levels of stress early in the pandemic. Further research is required with verified SARS-CoV-2 outcomes to determine whether the HPA axis is among the mechanisms by which a history of mental health difficulties and stress influence SARS-CoV-2 outcomes.
Keywords
Hair cortisol
Hair cortisone
Stress
COVID-19
SARS-CoV-2
Mental health
==== Body
pmc1 Introduction
The COVID-19 (Coronavirus, 2019) pandemic caused by the SARS-CoV-2 virus has resulted in unprecedented disruption to societies, health services, and economies. Increases in mental health difficulties (e.g., anxiety and depression) and risk factors associated with poorer mental health (e.g., loneliness) are now well-documented consequences of the pandemic (Jia et al., 2020, Kwong et al., 2020, Luchetti et al., 2020, Torales et al., 2020). In view of the established associations between adverse emotional experiences and physical health (Vedhara et al., 1999, Vedhara et al., 2010, Pantell et al., 2013), these findings raise important questions about the role of mental health, and related psychological constructs, as possible risk factors for SARS-CoV-2 infection and, critically, the mechanisms underlying any such effects. We provide evidence here regarding the potential role of the hypothalamic-pituitary-adrenal (HPA) axis.
The empirical basis for this hypothesis comes principally from viral challenge studies. These studies typically involve quarantining healthy volunteers for several days during which they are exposed to one or more respiratory viruses and followed up for evidence of infection and/or the presence of symptomatic illness. One of the first, and perhaps most well-known studies showed a dose response relationship between a composite measure of psychological stress (stressful life events, negative affect, and perceived stress) and the likelihood of viral infection and the severity of subsequent illness (Cohen et al., 1991). These results not only showed that increased levels of stress predicted an increased risk of developing a respiratory illness; but also that these effects occurred across a range of different viruses (rhinovirus type 2, 9, 14, respiratory syncytial virus and coronavirus type 229E). Since this ground-breaking work, several related studies have shown that psychological factors influence susceptibility to viral infections (Cohen et al., 1997, Cohen et al., 1998, Cohen, 1999, Cohen et al., 2003, Cohen et al., 2004, Cohen, 2005, Cohen et al., 2015). Evidence is emerging that these same pathways may be relevant in the context of SARS-CoV-2. For example, research from the UK biobank compared the risk of COVID-19 outcomes in participants diagnosed with a psychiatric disorder pre-pandemic, with those who had not (Yang et al., 2020). They observed that participants with a history of psychiatric disorder were at greater risk of COVID-19 infection, hospitalisation, and mortality. Further support comes from a recent systematic review of COVID-19-related mortality risk in people with severe mental illness (De Hert et al., 2021). Results from 13 studies suggested an association between severe mental illness and COVID-19 mortality (De Hert et al., 2021). Although the underlaying mechanisms for the increased risk of disease for those with psychiatric disorders are unclear, these studies point to a potential common pathway involving compromised immunity (Yang et al., 2020, De Hert et al., 2021).
One possibility for a common pathway that links psychological health to infection susceptibility may be the HPA axis and, specifically the hormone cortisol. Observational and experimental evidence has shown that psychological distress can dysregulate the HPA axis (Söndergaard and Theorell, 2003, Hsiao et al., 2011) and this in turn, through the immunomodulatory properties of cortisol, can compromise immune function (Ibar et al., 2021, O'Connor et al., 2021). The secretion of cortisol can suppress the activity of the natural killer (NK) cells and the transcription of proinflammatory cytokines via direct interactions with glucocorticoid receptors, which are present on many immune cells (Theoharides and Conti, 2020, Peters et al., 2021). Cohen and colleagues (2012) hypothesized that chronic increases in cortisol can lead to decreased sensitivity of immune cells to glucocorticoid hormones. This, in turn, can interfere with the production of proinflammatory cytokines in response to viral infections and promote an exaggerated response to infection (Cohen et al., 2012) as observed in some patients who become critically ill following COVID-19 infection (Peters et al., 2021). Evidence of an association between cortisol and SARS-CoV-2 outcomes is also emerging. For example, Tan et al. (2020) measured serum cortisol in 535 patients admitted to hospital during the first wave of COVID-19 infections in the UK. They reported that the risk of mortality increased significantly by 42% per doubling of cortisol concentrations, after adjusting for age, other comorbidities, and laboratory tests (Tan et al., 2020). These findings resonate with those of the RECOVERY trial (Horby et al., 2021) and a recent meta-analysis (Sterne et al., 2020) which demonstrated lower mortality in COVID-19 patients receiving synthetic corticosteroid treatments such as dexamethasone. These synthetic versions of cortisol alter the body’s own production of the hormone and, could, therefore, interfere with its capacity to dysregulate the immune system including its response to COVID-19 infection. Recent observational studies have also reported a relationship between cortisol and mental health during the COVID-19 pandemic. For example, Rajcani et al. (2021) and Marcil et al. (2021) reported that cortisol measured in hair during the COVID-19 pandemic increased by 22–27% among healthcare workers (Rajcani et al., 2021, Marcil et al., 2022). Ibar et al. (2021) found that health workers with burnout also had significantly higher hair cortisol levels during the pandemic.
Taken together, the evidence suggests that the increased risk of SARS-CoV-2 infection and poorer disease outcomes observed in people with a history of psychiatric illness, may be mediated by the HPA axis and an increase in the production of cortisol. To examine this further, we measured concentrations of cortisone (a metabolite of cortisol) in hair (hairE), in a general population sample of adults in the United Kingdom (UK), during the COVID-19 pandemic. Our aims were to: (i) report on whether and how levels of hairE changed in a UK cohort over a 6 month period early in the UK’s experience of the pandemic; (ii) compare these levels with existing pre-pandemic population data; and (iii) examine whether hairE levels differed significantly between people with or without a history of mental health difficulties as well as assess the relationship between hairE and levels of stress, anxiety and depression reported during the pandemic.
2 Methods
2.1 Patient and public involvement (PPI)
We convened a virtual PPI group to support this research, the aims of which were to advise on the development of the survey, the participant information sheet, and methods for optimising recruitment and retention. Individuals participated via Microsoft Teams in one-to-one or group-based discussions at the design phase of the research. These discussions informed the length and structure of the survey, language in the information sheet, and strategies for recruiting via media and social media. For example, the PPI group suggested using social media campaigns with daily interactive posts during the recruitment period. They also supported the idea of snowball recruitment. Snowball recruitment is a common sampling method in research where the researcher expands their pool of potential participants by encourages initial participants to reach out to their contacts to inform them about the research, and potentially participate (Marcus et al., 2017). The PPI group also advised on the frequency of providing feedback to participants and reporting study findings through the study website and between each wave of data collection.
2.2 Ethics, recruitment and eligibility
Ethical approval was granted from the University of Nottingham Faculty of Medicine & Health Sciences Research Ethics Committee (FMHS 506–2003) and recruitment commenced on 3 April 2020. Participants were recruited in the community via a social and mainstream media campaign involving, but not limited to, Facebook and Twitter. Dedicated social media accounts were created and engaging, interactive posts were posted daily through these accounts to encourage participation. We also sought to encourage the participation of healthcare workers and achieved this by seeking additional approvals through the Health Research Authority (HRA, approval number 20/HRA/1858). This enabled us to approach National Health Service (NHS) organisations and request that they advertise the research through their routine communications to staff (e.g., newsletters, emails). Recruitment continued until 30 April 2020.
All media and promotion directed potential participants to the study website through which they accessed the participant information sheet, consent form, online surveys, and instructions on how to take hair samples. Participants were informed through follow-up email communications that they would be entered into a prize draw of a £ 200 Amazon voucher at the end of the study if they completed all study surveys and provided two hair samples. Once recruited we used snowball recruitment (Marcus et al., 2017) to encourage rapid growth in the size of the cohort. This involved: (1) an email to existing participants to thank them for taking part and encouraging them to reach out to another 10 people they know (e.g., other members of the family or friends) to consider taking part. This email was sent to the first n = 335 participants one week after launching the study, and then new participants at the end of every week. (2) a similar thank you email to all participants alongside an encouragement to reach out to a further two people to consider taking part 10 days, three days, one day before closure of recruitment (i.e., 30th April 2020).
Eligibility criteria specified that participants should be: aged 18 and over; able to give informed consent; able to read English; residing in the UK at the time of completing the survey; and able to provide a sample of hair at least 1 cm long. The latter was collected to permit measurement of the hairE.
2.3 Procedures
Data were collected at three timepoints during 2020. Fig. 1 provides an overview of data collection in relation to the timeline of the pandemic in the UK. We report here data from Timepoint 1 (T1: commenced April 2020) and Timepoint 2 (T2: commenced July 2020) during which online surveys were completed and hair samples collected for the measurement of hairE. Online surveys were implemented through JISC Online Survey (https://www.onlinesurveys.ac.uk/). These collected demographic information (e.g., age, gender) and self-report measures of anxiety (7-item Generalized Anxiety Disorder Scale, GAD-7, T1 α = 0·92, T2 α = 0·91), depression (Patient Health Questionnaire, PHQ-9, T1 α = 0·88, T2 α = 0·87) and stress (4-item Perceived Stress Scale, T1 α = 0·76, T2 α = 0·75) (Cohen, 1988, Spitzer et al., 2006, Kroenke et al., 2010). Previous diagnosis of mental health disorders was measured by a single item ‘do you have a history of anxiety, depression or any other mental health issue for which you have received treatment in the past’ (yes/no/prefer not to say).Fig. 1 Trajectory of the COVID-19 pandemic in the UK and relationship with assessments in the current study.
Fig. 1
Collection of hair samples for the measurement of hairE followed standard methods (Staufenbiel et al., 2015). In brief, participants were provided with a step-by-step guide on the study website. This included text and video guidance on how to cut a hair sample of no less than 1 cm and approximate width of a pencil (3 mm) from the vertex posterior of the head. After taking the sample, participants were instructed to wrap the sample in kitchen foil and clearly label it with their study identifier, denote the root end (i.e., the end closest to the scalp) of the sample, and date the sample was taken. These instructions were repeated at T2 and participants were asked to return their samples using a freepost address. As the majority of participants who returned hair samples provided samples of 3 cm or longer (T1: n = 849, 95%; T2: n = 857, 96%) we analysed the first 3 cm of each sample to indicate hairE in the preceding 3 months (Staufenbiel et al., 2015) i.e., the T1 sample captured hairE in the 3 months prior to sample collection at T1, and the T2 sample captured hairE in the 3 months prior to sample collection at T2.
2.4 HairE measurement
Hair samples were prepared for the cortisone assay following standard methods described by Gao et al. (2013). Specifically, a minimum of 7·5 mg of hair was obtained from each hair sample ≥ 3 cm. Hair samples shorter than 3 cm were not assayed. The hair washing and cortisone extraction procedures were based on the protocol previously described in Stalder et al. (2012). In brief, hair strands were washed by shaking them in 2·5 mL isopropanol for 3 min at room temperature and then dried under a fume hood for at least 12 h. 7·5 mg of whole non-pulverised hair was then carefully weighed out and transferred into a 2 mL tube (Eppendorf, Hamburg, Germany). 50 μL internal standard and 1·8 mL methanol were added and the hair was incubated for 18 h at room temperature for cortisone extraction. Samples were spun in a centrifuge at 10,000 rpm for 2 min and the clear supernatant was transferred into a new 2 mL tube. The alcohol was evaporated at 65 ◦C under a constant stream of nitrogen for approximately 20 min until the samples were completely dried. The dry residue was resuspended using 250 μL distilled water, 200 μL of which was used for liquid chromatography tandem mass spectrometry (LC-MC) analysis. More details of the protocol are provided by Gao et al. (2013).
2.5 Sample size
The cohort was recruited to track the psychological and physical effects of the COVID-19 pandemic on the UK population. As such, we did not place an upper limit on participant numbers to enable us to obtain as precise estimates of population values and associations as possible, and to permit subgroup analysis where applicable. Nonetheless, power calculations indicate that to detect a small effect size change (dz=0·2) on cortisone levels from T1 to T2 in a two-tailed t-test with 90% power and alpha set at 0·05, a minimum of 265 participants would be needed to provide hair samples at both timepoints.
2.6 Statistical analysis
Summary statistics were used to describe characteristics of participants, hairE levels, and levels of stress, anxiety and depression at T1 and T2. Comparisons between hairE at T1 and T2 were conducted using a paired-samples t-test with log-transformed hairE values. Comparisons between mean hairE values at T1 and T2 with pre-pandemic population mean data were also conducted. These latter means were derived from a cohort of 10,814 individuals with a mean age of 47 years (SD 24·21) and 65% of whom were female (manuscript in preparation). As the raw data from this cohort were not available, independent t-tests were conducted. While t-tests are robust to deviations from normality when sample sizes are large, results of these specific analyses should be interpreted with caution (Lumley et al., 2002).
Four hairE outcomes were considered in the analyses: hairE values at T1, hairE values at T2, mean of T1 and T2 hairE values, and change in hairE from T1 to T2. Multivariable linear regressions were used to examine the associations between the hairE outcomes and history of mental health difficulties as well as stress, anxiety and depression reported at T1. Histograms showed that the distributions of the absolute hairE values at T1 and T2, and mean hairE across T1 and T2 were not normally distributed. Thus, log-transformed scores for these three outcomes were used in all multivariable linear regressions. For the history of mental health difficulties, relationships with all four hairE outcomes were considered. Analyses with stress, anxiety, and depression took into account temporal differences between when the measures of mood were captured and the periods covered by the hair samples (i.e., T1 hair samples captured hairE in the 3 months prior to the T1 mood measures and T2 hair samples captured hairE in the 3 months prior to the T2 mood measures). Thus, analyses examining the relationship between stress, anxiety and depression reported during the pandemic and hairE involved stress, anxiety and depression at T1 predicting (i) hairE at T2 and (ii) change in hairE between T1 and T2 only. For all analyses, anxiety and depression were determined by dichotomising scores on the GAD-7 and PHQ-9 (Spitzer et al., 2006, Kroenke et al., 2010) according to established cut-offs for high intensity psychological support in the National Health Service (National Collaborating Centre for Mental Health, 2019). Age and gender were controlled for in all regression analyses.
2.7 Additional analysis
Although we instructed participants to take the first hair sample in April and the second hair sample in July, some of the samples were dated later than suggested. Further, the time intervals between each sample from participants varied. To account for the influence of these time variations on our findings, we calculated an additional variable to represent these temporal factors. For T1 and T2 the first and last dates of sample collection covered a period of 97 and 91 days respectively. Accordingly, at both timepoints, an individual who provided a sample on the first day was allocated a score of 0; while samples collected on the last day were allocated a score of 97 for T1 and 91 for T2. The time interval between the two hair samples was then calculated as the number of calendar days between the two samples (median=92, range: 48–166). This was then included as a covariate in all regression models as an additional analysis to examine the influence of the differing time intervals between samples.
All analyses were performed using STATA (version 16) and GraphPad Prism (version 9.1.2).
Statistical significance was defined as p < 0·05.
3 Results
3.1 Cohort characteristics
Hair samples at both time points were received from 980 (32%) participants of our original cohort who completed the T1 survey (N = 3097). Of these, n = 89 (9%) were excluded for a variety of reasons including the samples being labelled with an incorrect study ID, the hair sample being insufficient/missing, or the root end of the sample being unclear (see Supplementary Appendix Fig. 1). The remaining 891 (91% of n = 980) pairs of samples were assayed. Of these, 64 participants (7% of n = 895) were excluded due to their samples being less than 3 cm in length (n = 56); participants collecting their T2 hair sample before the date requested (n = 4); participants providing both hair samples within 5 calendar days (n = 2); and hairE being undetectable in the samples (n = 2). Thus, the final cohort included in all analyses were 827 participants (84% of n = 980) who provided two hair samples. N = 788 participants (95% of n = 827) provided responses to the question regarding previous mental health difficulties. Thus, analyses pertaining to the relationship between hairE and previous mental health difficulties were restricted to this subsample.
Comparisons between participants who provided two hair samples (n = 827) and those who did not (n = 2268) indicated that older participants (t = −11·43, p < ·001, Cohen’s d=−0·46), female participants (Χ2 =91·96, p < ·001), White British participants (Χ2 =16·21, p < ·001), and those with lower baseline depression (t = 8·93, p < ·001, Cohen’s d=0·36), anxiety (t = 7·02, p < ·001, Cohen’s d=0·29), and stress (t = 7·03, p < ·001, Cohen’s d=0·29) were more likely to return two hair samples (see Table 1).Table 1 Characteristics of participants who provided two hair samples and those who did not.
Table 1 Participants who provided two samples Participants who provided < 2 samples
n (%) n (%)
N 827 (27%) 2268 (73%)
Gender*
Male 77 (9%) 399 (18%)
Female 749 (91%) 1866 (82%)
Prefer not to say 0 3 (0·1%)
Age (mean, SD)* 49.61 (15·32) 42.78 (14·46)
Age groups (years)
18–24 60 (7%) 302 (13%)
25–34 102 (12%) 525 (19%)
35–44 138 (17%) 499 (22%)
45–54 168 (20%) 522 (23%)
55–64 213 (26%) 357 (16%)
65–74 119 (14%) 138 (6%)
≥ 75 26 (3%) 23 (1%)
Ethnicity*
White British 776 (94%) 2017 (89%)
Black, Asian, and Minority Ethnic 50 (6%) 246 (11%)
Prefer not to say 0 5 (0·2%)
Education
No qualifications 9 (1%) 24 (1%)
GSCE/CSE/O-levels or equivalent 49 (6%) 203 (9%)
Post-16 vocational course 30 (4%) 71 (3%)
A-levels or equivalent 83 (10%) 318 (14%)
Undergraduate degree 377 (46%) 929 (41%)
Postgraduate degree 276 (33%) 699 (31%)
Prefer not to say 2 (0·2%) 24 (1%)
Previous diagnosis of mental health disorders
Yes 297 (36%) 222 (10%)
No 491 (59%) 359 (16%)
Indices of psychological well-being
T1 Depression (median, IQR)* 5 (2–8) 7 (3–12)
T1 Depression cases (mean ≥10)a 170 (20·6%) 806 (36%)
T1 Depression non-cases (mean<10)a 656 (79.4%) 1462 (64%)
T1 Anxiety (median, IQR)* 4 (1–8) 6 (2–11)
T1 Anxiety cases (mean≥8)a 208 (25%) 844 (37%)
T1 Anxiety non-cases (mean<8)a 618 (75%) 1424 (62.8%)
T1 Stress (mean, SD)* 5·79 (3·12) 6.72 (3·30)
*significant differences between participants who provided at least two hair samples and who did not.
a A ‘case’ is defined as the PHQ-9 score greater or equal to 10 for depression, or the GAD-7 score greater or equal to 8 for anxiety, at which level someone would qualify for high intensity psychological support in the English National Health Service.
3.2 HairE during the COVID-19 pandemic and comparisons with pre-pandemic data
Medians (IQRs) of cortisone values in the study participants at T1 and T2 and for pre-pandemic population data are presented in Table 2 and Fig. 2. A Paired-sample t-test showed that hairE levels (log-transformed values) were significantly higher over the 3-month period captured at T2 than the 3 month period captured at T1 (t = −8.42, p < ·001, Cohen’s d=−0·51), with mean values 23% higher at T2. Independent t-tests were used to compare the mean values of hairE in our cohort at T1 and T2 with pre-pandemic population data. Results showed no significant differences in mean hairE values at T1 compared with pre-pandemic data (t = 1·64, p = ·10, Cohen’s d=0·06). However, mean hairE levels at T2 were significantly higher than pre-pandemic data (t = 5·44, p < ·001, Cohen’s d=0·19).Table 2 HairE at T1 and T2 compared with pre-pandemic norms.
Table 2 Pre-pandemic population data for hairE (N = 10,814) T1 hairE (N = 827) T2 hairE (N = 827)
Sample 1 Sample 2
Median (IQR) 8·55 (5·23–14·18) 8·20 (5·52–12·22) 9·68 (6·18–15·44)
Fig. 2 Mean values of hairE at T1 and T2 and pre-pandemic population data. The boxes represent 25% quartiles to 75% quartiles. The solid lines inside the box represent medians.
Fig. 2
3.3 Relationship between hairE and history of mental health difficulties
Multivariable linear regressions ( Table 3) examined the relationship between hairE outcomes and having a history of mental health difficulties, after controlling for age and gender. For hairE at T1, being female (p < ·001) was significantly associated with lower hairE values, and having a history of mental health difficulties (p < ·001) was significantly associated with higher hairE values. Similar associations were evident for hairE at T2 (female: p = ·002; history of mental health difficulties: p = ·045). With regard to mean hairE values across T1 and T2, the results showed that having a history of mental health difficulties (p < ·001) was associated with significantly elevated average levels of hairE of T1 and T2 combined. It was, however, unrelated to the changes in hairE between T1 and T2 (p = ·19). Greater change in hairE between T1 and T2 was, however, associated with being female (p = ·019) and older (p = ·041). Our additional analysis of the influence of differing time intervals between samples showed that after adding the number of days between the two hair samples as a covariate, these results were largely unaffected (see Appendix Table 1).Table 3 Multivariable regression models showing the associations between previous diagnosis of mental health disorders and hairEs controlling for age and gender.
Table 3 HairEa HairEa HairEa HairEa
T1 T2 Average of T1 and T2 values Changes from T1 to T2
β, B (95% CI), p β, B (95% CI), p β, B (95% CI), p β, B (95% CI), p
Age (per 10 years increase) 0·004, 0·002 (−0·03, 0·04),·91 0·05, 0·02 (−0·01, 0·06),·19 0·03, 0·01 (−0·02, 0·04),·37 0·08, 0·50 (0·08, 0·91),·019 *
Female (Y/N) -0·14, − 0·38 (−0·56, −0·19), < ·001 * ** -0·11, − 0·29 (−0·47, −0·10),·002 * * -0·14, − 0·32 (−0·48, −0·15), < .001 * ** 0·07, 2.30 (0·10, 4.50),·041 *
Previous diagnosis of mental health disorder (Y/N) 0·11, 0·18 (0·07, 0·28),·001 * * 0·07, 0·11 (0·003, 0·22),·045 * 0·10, 0·14 (0·04, 0·23),·005 * * -0·05, − 0·88 (−2·20, 0·44),·19
Intercept 2·35 (2·10, 2·60), < ·001 * * 2·38 (2·13, 2·64), < ·001 * * 2·40 (2·18–2·60), < ·001 * *- 1·66 (−4·71, 1·39),·29
Adjusted R2= 0·03, n = 788 Adjusted R2= 0·01, n = 788 Adjusted R2= 0·02, n = 788 Adjusted R2= 0·01, n = 788
* ** p < ·001, * * p < ·01, * p < ·05
a A log transformation was applied to the dependent variable.
3.4 Relationship between hairE and depression, anxiety and stress reported during the pandemic
Multivariable linear regressions examined prospective associations between stress, anxiety and depression reported at T1 and (i) hairE at T2 and (ii) change in hairE between T1 and T2. These analyses (see Table 4) revealed that, after controlling for age and gender, meeting the criteria for high intensity support for anxiety and depression at T1 was not associated with hairE values at T2 or the change in hairE between T1 and T2 (see models 1–4 in Table 4). However, stress at T1 was significantly associated with hairE at T2 (model 5, p = ·017) as well as the change in hairE between T1 and T2 (model 6, p = ·032). The additional analysis of the influence of differing time intervals between samples showed that, after adding the number of days between the two hair samples as a covariate, these results were largely unaffected (see Appendix, Table 2).Table 4 Multivariable regression models showing demographic and psychological predictors of hairE outcomes.
Table 4 T2 hairE hairE change from T1 to T2
β, B (95% CI), p β, B (95% CI), p
Model 1 Model 2
Age (per 10 years increase) 0·06, 0·03 (−0·002, 0·07),0.07 0·10, 0·60 (0·16, 1·03),·007 * *
Female (Y/N) -0·11, − 0·29 (−0·47, −0·12),·001 * * 0·05, 1·76 (−0·45, 3·97),·12
Depression casesa at T1 (Y/N) 0·07, 0·13 (−0·01, 0·26),·06 0·01, 0·27 (−1·37, 1·92),·75
Intercept 2·37 (2·12, 2·62), < ·001 * * -2·15 (−5·26, 0·96),·18
Adjusted R2= 0·02, n = 826 Adjusted R2= 0·01, n = 826
Model 3 Model 4
Age (per 10 years increase) 0·05, 0·03 (−0·01, 0·06),·14 0·10, 0·60 (0·15, 1.02),·009 * *
Female (Y/N) -0·11, 0·29 (−0·46, −0·11),·002 * * 0·05, 1·78 (−0·43, 3·98),·12
Anxiety cases at T1a (Y/N) 0·02, 0·04 (−0·09, 0·16),·55 0·005, 0·11 (−1·44, 1·65),·89
Intercept 2·40 (2·15, 2·66), < ·001 * * -2·09 (−5·24, 1·07),·19
Adjusted R2= 0·02, n = 826 Adjusted R2= 0·01, n = 826
Model 5 Model 6
Age (per 10 years increase) 0·07, 0·03 (−0·001, 0·07),·054 0·11, 0·69 (0·26, 1·13),·002 * *
Female (Y/N) -0·11, − 0·29 (−0·47, −0·12),·001 * * 0·05, 1·66 (−0·55, 3·86),·14
T1 stress (per unit) 0·09, 0·02 (0·00, 0·04),·017 * 0·08, 0·23 (0·02, 0·44),·032 *
Intercept 2·26 (1·98, 2·54), < ·001 * * -3·84 (−7·27, −0·40),·03 *
Adjusted R2= 0·02, n = 826 Adjusted R2= 0·01, n = 826
* ** p < ·001, * * p < ·01, * p < ·05
a A ‘case’ is defined as the PHQ-9 depression score greater or equal to 10, or the GAD-7 anxiety score greater or equal to 8, at which level someone would qualify for high intensity psychological support in the National Health Service.
4 Discussion
The present study examined whether the HPA axis, as measured by hairE, was altered in a convenience sample of UK citizens during the COVID-19 pandemic and the extent to which hairE was related to an individual’s previous experience of mental health difficulties, as well as their experiences of stress, anxiety and depression during the pandemic. The findings indicate that in the 3-month period immediately before the UK’s first national lockdown (captured by hairE at T1), hairE levels in our cohort were not significantly different to pre-pandemic population levels. However, hairE at T2, which captured a 3-month period of both local and national restrictions in the UK, was significantly greater than hairE at T1 and also pre-pandemic population levels. Indeed, the mean hairE at T2 was 23% greater than mean levels observed at T1. Furthermore, we observed that individuals with a history of mental health difficulties had the greatest hairE (for three out of four hairE outcomes) and that higher stress scores at the start of the pandemic also predicted higher levels of hairE by T2 and a greater increase in hairE between T1 and T2.
Several issues are worthy of further discussion. First, the present work replicates and extends early findings on the experiences of health care workers during the pandemic which has shown an increase in cortisol levels when compared with pre-pandemic levels and positive associations between stress and cortisol in these populations (Ibar et al., 2021, Rajcani et al., 2021, Marcil et al., 2022). Here we show evidence that these psychobiological perturbations were also evident amongst a general population sample. Furthermore, our study, and that by Racjani et al. (2021) and Marcil et al. (2022), report comparable changes in cortisol/cortisone during the pandemic: with Rajcani et al., reporting an increase of 22% on pre-pandemic levels in health care professionals, Marcil et al. reporting an increase of 27%, while our data show a 23% increase in hair cortisone.
Second, the positive associations between hairE outcomes and having a history of mental health difficulties in this study, may provide a plausible mechanism for the observation (Yang et al., 2020, De Hert et al., 2021) that people with a history of psychiatric disorders appear to be at greater risk of COVID-19 infection, hospitalisation, and/or mortality. Cortisol is one of the primary products of the HPA axis and alterations in this hormone are a well-established consequence of psychological distress (Ibar et al., 2021, O'Connor et al., 2021). Furthermore, cortisol has immunomodulatory properties which enable distress-induced changes in cortisol to influence a range of health outcomes (Adam et al., 2017). Although we measured cortisone, rather than cortisol, the two hormones are structurally very similar, are highly correlated, and both can be measured in hair (Stalder et al., 2013, Staufenbiel et al., 2015). Cortisone is considered an acceptable surrogate for the measurement of free cortisol (i.e., the unbound biological active levels of the hormone) (Raul et al., 2004, Staufenbiel et al., 2015) and less prone to external contamination from cortisol-containing products (Raul et al., 2004, Wang et al., 2019, Feeney et al., 2020). Thus, it is plausible that the elevations in hairE observed in this cohort, offer insight into the mechanisms underlying the increased risk of COVID-19 infection and poorer outcomes in people with a history of psychiatric illness. (Yang et al., 2020).
A third, and related observation concerns the fact that the associations between hairE and psychological well-being were not restricted to those with a history of mental health difficulties. We observed that stress experienced early in the pandemic at T1 was also related to later increases in hairE. This may suggest that, as found with previous viral challenge studies, people experiencing greater psychological stress may be at greater risk from SARS-CoV-2 infection as well symptomatic illness, as a result of increases in cortisol and the consequent dysregulation of the immune system. However, the present study did not specifically examine relationships with SARS-CoV-2 infections. Future studies of serologically verified infections would further shed light on these relationships. The absence of comparable relationships with anxiety and depression at T1 and subsequent hairE may simply reflect the smaller effect sizes associated with psychopathology and hairE (Staufenbiel et al., 2013). It should be acknowledged, however, that the proportion of variance in hairE accounted for by both stress and history of mental health difficulties was small to modest. This is consistent with that reported in previous work. For example, Rajcani et al. (2021) reported small effect of stress (effect size <0.001) on nurses during COVID-19.(Rajcani et al., 2021), and Ibar et al. (2021) reported small association between stress and hair cortisol in health workers (r = 0.14) (Ibar et al., 2021). This indicates that the HPA axis is likely to be just one of the pathways by which stress and mental health alter the risk of SARS-CoV-2 infection. Other pathways such as through neurotransmitters (e.g., adrenaline and noradrenaline), other hormones involved in the HPA axis (e.g., corticotrophin-releasing hormone), and lifestyle factors (e.g., sleep, diet, physical activity) may also be involved in this relationship (DuPre et al., 2021, Peters et al., 2021). Future work may, therefore, want to consider a broader range of mechanisms through which stress and mental health influence the risks of SARS-CoV-2 infection.
Some limitations of this work should also be acknowledged. These include that our assessment of previous mental health difficulties was based on self-report and not verified through clinical records. The current findings are derived from an opportunistic self-selected cohort. Individuals who provided us with two hair samples suitable for analysis were more likely to be female, older but also less stressed, anxious and depressed than the remainder of participants in the original cohort. Although this has implications for the generalisability of our findings to the original cohort, it does also suggests that the magnitude of the change in hairE during the COVID-19 pandemic is likely to be an under-estimate as the most distressed individuals did not participate in this aspect of the research. There are likely several reasons for the predominance of female participants in our study. Firstly, one of the eligibility criteria was that participants had to be able to provide a sample of hair at least 1 cm long. This might have prevented men with no or shorter hair from participating. Secondly, a substantial portion of our recruitment came through promoting this study in NHS settings, with 39% of our cohort identified as healthcare workers. According to NHS figures, 76.7% of the 1.3 million members of NHS staff are women (NHS, 2021), therefore it is likely that a higher proportion of women were made aware of our study and ultimately chose to participate. Thirdly, typical of previous online research studies concerning mental health, women were overrepresented in our sample (Crisp and Griffiths, 2014). Although we controlled for age and gender in the analyses, other potential confounders of hairE such as BMI, smoking status, hair washing frequency, and use of hair products were not accounted for in this study (Stalder et al., 2013, Staufenbiel et al., 2015). Furthermore, the effect sizes of the associations between having a history of mental health difficulties, stress and increases in hairE were small. This suggests that there may be other factors contributing to elevated hairE levels which we did not measure in the present study e.g., socioeconomic status, lifestyle behaviours such as physical activity and sleep. Such parameters have been shown to be related to both cortisol (Cohen et al., 2006, Jia et al., 2022) and COVID-19 outcomes (Caroppo et al., 2021, DuPre et al., 2021).
5 Conclusion
This prospective cohort study has shown a 23% increase in hairE in a large general population sample, during a 3-month period early in the course of the COVID-19 pandemic, with the greatest levels seen in people with a history of mental health difficulties as well as those reporting elevated stress early in the pandemic. In view of the role of cortisol in regulating the immune system, this evidence of chronic increases in hairE may explain some of the increased risk of SARS CoV-2 infection and poorer clinical outcomes observed in people with a history of mental health difficulties.
Ethical approval
All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Ethical approval was granted from the University of Nottingham Faculty of Medicine & Health Sciences Research Ethics Committee (FMHS 506–2003) and the Health Research Authority (HRA, approval number 20/HRA/1858).
Funding details
K.A. acknowledges financial support from the National Institute for Health Research (10.13039/100006662 NIHR ) School for Primary Care Research. T.C. acknowledges financial support from the Department of Health via the NIHR Specialist Biomedical Research Centre for Mental Health award to the South London and Maudsley NHS Foundation Trust and the Institute of Psychiatry at King’s College London. C.C. acknowledges support from the NIHR Nottingham Biomedical Research Centre. The views expressed are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health and Social Care. No other funding supported the work described in this manuscript.
Role of sponsor
The study sponsor did not play a role in the study design, collection; analysis, and interpretation of data; in the writing of the report; or in the decision to submit the paper for publication.
Statement for disclosure of sample, conditions, measures, and exclusions
We have reported all measures, conditions, data exclusions, and determination of sample size for the study described in this manuscript. We confirm that this statement is accurate.
CRediT authorship contribution statement
Jia – conceptualization; data curation; formal analysis; investigation; methodology; writing original draft; writing review and editing. Ayling – conceptualization; data curation; formal analysis; investigation; methodology; writing original draft; writing review and editing. Coupland – conceptualization; data curation; formal analysis; investigation; methodology; writing review and editing. Chalder – conceptualization; data curation; formal analysis; investigation; methodology; writing review and editing. Massey – conceptualization; writing review and editing. Natar – conceptualization; methodology; writing review and editing. Broadbent – conceptualization; writing review and editing. Gasteiger – conceptualization; writing review and editing. Gao – data curation; methodology; writing review and editing. Kirschbaum – methodology; supervision; writing review and editing. Vedhara – conceptualization; data curation; formal analysis; investigation; methodology; writing original draft; writing review and editing.
Conflict of interest
AM is the director of Cortigenix (www.cortigenix.com). Cortigenix provided guidance on remote self-collection of hair samples by participants.
Appendix A Supplementary material
Supplementary material
Supplementary material
.
Analytic code availability
There is not analytic code associated with this study.
Appendix A Supplementary data associated with this article can be found in the online version at doi:10.1016/j.psyneuen.2022.105992.
==== Refs
References
Adam E.K. Quinn M.E. Tavernier R. McQuillan M.T. Dahlke K.A. Gilbert K.E. Diurnal cortisol slopes and mental and physical health outcomes: a systematic review and meta-analysis Psychoneuroendocrinology 83 2017 25 41 28578301
Caroppo E. Mazza M. Sannella A. Marano G. Avallone C. Claro A.E. Janiri D. Moccia L. Janiri L. Sani G. Will nothing be the same again?: changes in lifestyle during covid-19 pandemic and consequences on mental health Int J. Environ. Res Public Health 18 2021 16
Cohen S. Perceived Stress In A Probability Sample Of The United States. The Social Psychology Of Health 1988 Sage Publications, Inc Thousand Oaks, CA, US 31 67
Cohen S. Social status and susceptibility to respiratory infections Ann. N. Y Acad. Sci. 896 1999 246 253 10681901
Cohen S. Keynote Presentation at the Eight International Congress of Behavioral Medicine: the Pittsburgh common cold studies: psychosocial predictors of susceptibility to respiratory infectious illness Int. J. Behav. Med. 12 3 2005 123 131 16083315
Cohen S. Tyrrell D.A. Smith A.P. Psychological stress and susceptibility to the common cold N. Engl. J. Med. 325 9 1991 606 612 1713648
Cohen S. Doyle W.J. Skoner D.P. Rabin B.S. Gwaltney J.M. Jr. Social ties and susceptibility to the common cold Jama 277 24 1997 1940 1944 9200634
Cohen S. Frank E. Doyle W.J. Skoner D.P. Rabin B.S. Gwaltney J.M. Jr. Types of stressors that increase susceptibility to the common cold in healthy adults Health Psychol. 17 3 1998 214 223 9619470
Cohen S. Doyle W.J. Turner R.B. Alper C.M. Skoner D.P. Emotional style and susceptibility to the common cold Psychosom. Med 65 4 2003 652 657 12883117
Cohen S. Doyle W.J. Turner R.B. Alper C.M. Skoner D.P. Childhood socioeconomic status and host resistance to infectious illness in adulthood Psychosom. Med 66 4 2004 553 558 15272102
Cohen S. Doyle W.J. Baum A. Socioeconomic status is associated with stress hormones Psychosom. Med 68 3 2006 414 420 16738073
Cohen S. Janicki-Deverts D. Doyle W.J. Miller G.E. Frank E. Rabin B.S. Turner R.B. Chronic stress, glucocorticoid receptor resistance, inflammation, and disease risk Proc. Natl. Acad. Sci. USA 109 16 2012 5995 5999 22474371
Cohen S. Janicki-Deverts D. Turner R.B. Doyle W.J. Does hugging provide stress-buffering social support? A study of susceptibility to upper respiratory infection and illness Psychol. Sci. 26 2 2015 135 147 25526910
Crisp D.A. Griffiths K.M. Participating in online mental health interventions: who is most likely to sign up and why? Depress Res Treat. 2014 2014 790457
De Hert M. Mazereel V. Stroobants M. De Picker L. Van Assche K. Detraux J. COVID-19-related mortality risk in people with severe mental illness: a systematic and critical review Front Psychiatry 12 2021 798554
DuPre N.C. Karimi S. Zhang C.H. Blair L. Gupta A. Alharbi L.M.A. Alluhibi M. Mitra R. McKinney W.P. Little B. County-level demographic, social, economic, and lifestyle correlates of COVID-19 infection and death trajectories during the first wave of the pandemic in the United States Sci. Total Environ. 786 2021 147495
Feeney J.C. O'Halloran A.M. Kenny R.A. The association between hair cortisol, hair cortisone, and cognitive function in a population-based cohort of older adults: results from the irish longitudinal study on ageing J. Gerontol. A Biol. Sci. Med Sci. 75 2 2020 257 265 30412218
Gao W. Stalder T. Foley P. Rauh M. Deng H. Kirschbaum C. Quantitative analysis of steroid hormones in human hair using a column-switching LC-APCI-MS/MS assay J. Chromatogr. B Anal. Technol. Biomed. Life Sci. 928 2013 1 8
Horby P. Lim W.S. Emberson J.R. Mafham M. Bell J.L. Linsell L. Staplin N. Brightling C. Ustianowski A. Elmahi E. Prudon B. Green C. Felton T. Chadwick D. Rege K. Fegan C. Chappell L.C. Faust S.N. Jaki T. Jeffery K. Montgomery A. Rowan K. Juszczak E. Baillie J.K. Haynes R. Landray M.J. Dexamethasone in Hospitalized Patients with Covid-19 N. Engl. J. Med. 384 8 2021 693 704 32678530
Hsiao F.H. Jow G.M. Lai Y.M. Chen Y.T. Wang K.C. Ng S.M. Ho R.T. Chan C.L. Yang T.T. The long-term effects of psychotherapy added to pharmacotherapy on morning to evening diurnal cortisol patterns in outpatients with major depression Psychother. Psychosom. 80 3 2011 166 172 21389753
Ibar C. Fortuna F. Gonzalez D. Jamardo J. Jacobsen D. Pugliese L. Giraudo L. Ceres V. Mendoza C. Repetto E.M. Reboredo G. Iglesias S. Azzara S. Berg G. Zopatti D. Fabre B. Evaluation of stress, burnout and hair cortisol levels in health workers at a University Hospital during COVID-19 pandemic Psychoneuroendocrinology 128 2021 105213
Jia R. Ayling K. Chalder T. Massey A. Broadbent E. Coupland C. Vedhara K. Mental health in the UK during the COVID-19 pandemic: cross-sectional analyses from a community cohort study BMJ Open 10 9 2020 e040620
Jia R. Carlisle S. Vedhara K. The association of lifestyle and mood with long-term levels of cortisol: a systematic review Cogent Psychol. 9 1 2022 2036487
Kroenke K. Spitzer R.L. Williams J.B. Löwe B. The patient health questionnaire somatic, anxiety, and depressive symptom scales: a systematic review Gen. Hosp. Psychiatry 32 4 2010 345 359 20633738
Kwong A.S.F. Pearson R.M. Adams M.J. Northstone K. Tilling K. Smith D. Fawns-Ritchie C. Bould H. Warne N. Zammit S. Gunnell D.J. Moran P.A. Micali N. Reichenberg A. Hickman M. Rai D. Haworth S. Campbell A. Altschul D. Flaig R. McIntosh A.M. Lawlor D.A. Porteous D. Timpson N.J. "Mental health before and during the COVID-19 pandemic in two longitudinal UK population cohorts." Br. J. Psychiatry 2020 1 10
Luchetti M. Lee J.H. Aschwanden D. Sesker A. Strickhouser J.E. Terracciano A. Sutin A.R. The trajectory of loneliness in response to COVID-19 Am. Psychol. 75 7 2020 897 908 32567879
Lumley T. Diehr P. Emerson S. Chen L. The importance of the normality assumption in large public health data sets Annu. Rev. Public Health 23 1 2002 151 169 11910059
Marcil M.-J. Cyr S. Marin M.-F. Rosa C. Tardif J.-C. Guay S. Guertin M.-C. Genest C. Forest J. Lavoie P. Labrosse M. Vadeboncoeur A. Selcer S. Ducharme S. Brouillette J. Hair cortisol change at COVID-19 pandemic onset predicts burnout among health personnel Psychoneuroendocrinology 138 2022 105645
Marcus B. Weigelt O. Hergert J. Gurt J. Gelléri P. The use of snowball sampling for multi source organizational research: Some cause for concern Pers. Psychol. 70 2017 635 673
National Collaborating Centre for Mental Health (2019). The Improving Access to Psychological Therapies Manual - Appendices and helpful resources.
NHS. (2021). "NHS celebrates the vital role hundreds of thousands of women have played in the pandemic. Available from 〈https://www.england.nhs.uk/2021/03/nhs-celebrates-the-vital-role-hundreds-of-thousands-of-women-have-played-in-the-pandemic/〉 [Accessed 26/10/2022]."
O'Connor D.B. Thayer J.F. Vedhara K. Stress and health: a review of psychobiological processes Annu. Rev. Psychol. 72 1 2021 null
Pantell M. Rehkopf D. Jutte D. Syme S.L. Balmes J. Adler N. Social isolation: a predictor of mortality comparable to traditional clinical risk factors Am. J. Public Health 103 11 2013 2056 2062 24028260
Peters E.M.J. Schedlowski M. Watzl C. Gimsa U. To stress or not to stress: Brain-behavior-immune interaction may weaken or promote the immune response to SARS-CoV-2 Neurobiol. Stress 14 2021 100296
Rajcani J. Vytykacova S. Solarikova P. Brezina I. Stress and hair cortisol concentrations in nurses during the first wave of the COVID-19 pandemic Psychoneuroendocrinology 129 2021 105245
Raul J.S. Cirimele V. Ludes B. Kintz P. Detection of physiological concentrations of cortisol and cortisone in human hair Clin. Biochem 37 12 2004 1105 1111 15589817
Söndergaard H.P. Theorell T. A longitudinal study of hormonal reactions accompanying life events in recently resettled refugees Psychother. Psychosom. 72 1 2003 49 58 12466638
Spitzer R.L. Kroenke K. Williams J.B. Löwe B. A brief measure for assessing generalized anxiety disorder: the GAD-7 Arch. Intern Med 166 10 2006 1092 1097 16717171
Stalder T. Steudte S. Miller R. Skoluda N. Dettenborn L. Kirschbaum C. Intraindividual stability of hair cortisol concentrations Psychoneuroendocrinology 37 5 2012 602 610 21917384
Stalder T. Kirschbaum C. Alexander N. Bornstein S.R. Gao W. Miller R. Stark S. Bosch J.A. Fischer J.E. Cortisol in hair and the metabolic syndrome J. Clin. Endocrinol. Metab. 98 6 2013 2573 2580 23585660
Staufenbiel S.M. Penninx B.W. Spijker A.T. Elzinga B.M. van Rossum E.F. Hair cortisol, stress exposure, and mental health in humans: a systematic review Psychoneuroendocrinology 38 8 2013 1220 1235 23253896
Staufenbiel S.M. Penninx B.W. de Rijke Y.B. van den Akker E.L. van Rossum E.F. Determinants of hair cortisol and hair cortisone concentrations in adults Psychoneuroendocrinology 60 2015 182 194 26176863
Sterne J.A.C. Murthy S. Diaz J.V. Slutsky A.S. Villar J. Angus D.C. Annane D. Azevedo L.C.P. Berwanger O. Cavalcanti A.B. Dequin P.F. Du B. Emberson J. Fisher D. Giraudeau B. Gordon A.C. Granholm A. Green C. Haynes R. Heming N. Higgins J.P.T. Horby P. Jüni P. Landray M.J. Le Gouge A. Leclerc M. Lim W.S. Machado F.R. McArthur C. Meziani F. Møller M.H. Perner A. Petersen M.W. Savovic J. Tomazini B. Veiga V.C. Webb S. Marshall J.C. Association between administration of systemic corticosteroids and mortality among critically Ill patients with COVID-19: a meta-analysis Jama 324 13 2020 1330 1341 32876694
Tan T. Khoo B. Mills E.G. Phylactou M. Patel B. Eng P.C. Thurston L. Muzi B. Meeran K. Prevost A.T. Comninos A.N. Abbara A. Dhillo W.S. Association between high serum total cortisol concentrations and mortality from COVID-19 Lancet Diabetes Endocrinol. 8 8 2020 659 660 32563278
Theoharides T.C. Conti P. Dexamethasone for COVID-19? Not so fast J. Biol. Regul. Homeost. Agents 34 3 2020 1241 1243 32551464
Torales J. O’Higgins M. Castaldelli-Maia J.M. Ventriglio A. The outbreak of COVID-19 coronavirus and its impact on global mental health Int. J. Soc. Psychiatry 66 4 2020 317 320 32233719
Vedhara K. Cox N.K. Wilcock G.K. Perks P. Hunt M. Anderson S. Lightman S.L. Shanks N.M. Chronic stress in elderly carers of dementia patients and antibody response to influenza vaccination Lancet 353 9153 1999 627 631 10030328
Vedhara K. Miles J.N. Wetherell M.A. Dawe K. Searle A. Tallon D. Cullum N. Day A. Dayan C. Drake N. Price P. Tarlton J. Weinman J. Campbell R. Coping style and depression influence the healing of diabetic foot ulcers: observational and mechanistic evidence Diabetologia 53 8 2010 1590 1598 20411235
Wang X. Busch J.R. Banner J. Linnet K. Johansen S.S. Hair testing for cortisol by UPLC-MS/MS in a family: External cross-contamination from use of cortisol cream Forensic Sci. Int 305 2019 109968
Yang H. Chen W. Hu Y. Chen Y. Zeng Y. Sun Y. Ying Z. He J. Qu Y. Lu D. Fang F. Valdimarsdóttir U.A. Song H. Pre-pandemic psychiatric disorders and risk of COVID-19: a UK Biobank cohort analysis Lancet Healthy Longev. 1 2 2020 e69 e79 33521769
| 36495625 | PMC9705007 | NO-CC CODE | 2022-12-07 23:16:31 | no | Psychoneuroendocrinology. 2023 Feb 29; 148:105992 | utf-8 | Psychoneuroendocrinology | 2,022 | 10.1016/j.psyneuen.2022.105992 | oa_other |
==== Front
Can J Cardiol
Can J Cardiol
The Canadian Journal of Cardiology
0828-282X
1916-7075
Published by Elsevier Inc. on behalf of the Canadian Cardiovascular Society.
S0828-282X(22)01048-0
10.1016/j.cjca.2022.11.011
Review
Management of MIS-C (Multi-system Inflammatory Syndrome in Children): Decision-making regarding a new condition in the absence of clinical trial data
Harahsheh Ashraf S. MD, FACC, FAAP 1
Portman Michael A. 2
Khoury Michael 3
Elias Matthew D. 4
Lee Simon 5
Lin Justin 6
McCrindle Brian W. MD, MPH 6∗
1 Division of Cardiology, Department of Pediatrics, Children's National Hospital, George Washington University School of Medicine and Health Sciences, Washington, DC
2 Seattle Children's Research Institute, Seattle, WA, USA
3 Department of Pediatrics, Stollery Children's Hospital, University of Alberta, Edmonton, AB, Canada
4 Division of Cardiology, The Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania, USA
5 The Heart Center at Nationwide Children's Hospital, Columbus, OH, USA
6 Labatt Family Heart Centre, The Hospital for Sick Children, Department of Pediatrics, University of Toronto, Toronto, ON, Canada
∗ Correspondance: Brian W McCrindle MD, MPH, Labatt Family Heart Centre, The Hospital for Sick Children, Department of Pediatrics, University of Toronto, Toronto, ON, Canada,
29 11 2022
29 11 2022
21 10 2022
22 11 2022
24 11 2022
© 2022 Published by Elsevier Inc. on behalf of the Canadian Cardiovascular Society.
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Multisystem inflammatory syndrome in children (MIS-C) is a new illness that evolved during the COVID-19 pandemic with initial reports of severe disease including use of extracorporeal membrane oxygenation and death. Institutions rapidly assembled task forces to develop treatment algorithms. At the national/international levels, collaboratives and associations assembled consensus writing groups to draft guidelines. These guidelines and algorithms were initially based on expert opinion and small case series. Some groups utilized the Delphi approach, and the resultant guidelines often mimicked those for other conditions that resembled MIS-C, like Kawasaki disease (KD). For instance, intravenous immunoglobulin (IVIG), a known effective treatment in KD, was recommended for MIS-C. Early in the pandemic many favored IVIG over steroids as first line therapy. As evidence evolved so did some guidelines which now endorse the dual use of IVIG plus steroids as first line therapy. In contrast, withholding immunotherapy became an option for some MIS-C patients with mild symptoms. Here, we review guidelines and discuss the evidence informing early recommendations, how this has evolved, the role and limitations of expert opinion and observational data, and the importance of leveraging existing research infrastructures, such as the intensive care unit collaborative (Overcoming COVID-19 surveillance registry), and the International KD Registry. Finally, we discuss strategies to rapidly develop, deploy and adapt clinical trials evaluating the treatment of such rare conditions in children, which may include alternatives to conventional clinical trial design. The emergence of MIS-C during the COVID-19 pandemic has highlighted unmet needs regarding research of a new condition.
Graphical abstract
==== Body
pmcIntroduction:
Early experience during the 2019 Coronavirus disease (COVID-19) pandemic prompted the prediction that children would be largely spared from the severe disease noted in adults.1 However, several weeks after peak incidences of COVID-19 in Europe, global reports appeared describing a post-infectious inflammatory syndrome in children following exposure to severe acute respiratory syndrome coronavirus 2 (SARS-COV-2). Children with the syndrome exhibited clinical features overlapping with Kawasaki disease (KD) in addition to a shock-like presentation, often requiring intensive care management. The syndrome has had various names but for the purpose of this review, we will refer to the United States (US) CDC term, multisystem inflammatory syndrome in children (MIS-C).2, 3, 4, 5, 6 MIS-C also has multiple case definitions with recent call for development of international consensus on diagnostic criteria.7 Initial reports characterized MIS-C as a systemic vasculitis with a propensity for myocardial involvement.8 Emergence of new SARS-COV-2 variants as well as the impact of diagnostic and treatment algorithms have modified the clinical presentation. Thus, treatment algorithms developed from various professional societies will require adaptation. We present the readers with a review outlining the basis for these algorithms and their evolution in response to a changing disease landscape. Finally, we discuss how we can improve decision-making during the initial experience with a new condition, including a review of how clinical trials were expedited, and a discussion of the role of observational data and expert opinion.
Basis for management decisions early in the pandemic.
Faced with the rapid progression of an evolving post-COVID-19 severe illness along with reported cases of mortality and severe morbidity including the need for extracorporeal membrane oxygenation (ECMO)5, centers across the globe quickly formed task forces to develop institutional clinical practice guidelines and algorithms to help front line providers diagnose and promptly treat children presenting with signs and symptoms suggestive of MIS-C.9 , 10 Within months, treatments were leading to excellent outcomes.11
Initially, in the absence of evidence supporting best treatment practices, treatment decisions were largely influenced by the management of KD, given the significant overlap noted in presenting signs and symptoms.12 Both conditions are characterized by severe inflammatory disease, triggered by an environmental agent, which in the case of MIS-C is SARS-CoV-2.13 They may both have a genetic predisposition, albeit with different ethnic/race groups disproportionately affected in the two processes, and both affect the heart.13 , 14 They may present with similar clinical and laboratory features like fever, conjunctivitis, skin rash, and elevated inflammatory markers.13 Since diagnostic criteria in the United States allow for treatment in the presence of only 24 hours of fever, some treatment algorithms supported the early initiation of intravenous immunoglobulin (IVIG) in the presence of other supporting symptoms.10 KD signs and symptoms can arise at different timepoints throughout the illness and, thus, the diagnosis may be delayed.15 Thus, considering the ∼35 years-experience with IVIG for the treatment of KD, including the known effectiveness in reducing the incidence of coronary artery aneurysms from 25% to 4%15, institutional task forces and society guidelines largely supported prompt IVIG therapy as the first line treatment for MIS-C.16, 17, 18, 19 (Tables 1 and 2 )Table 1 Evolution of the American College of Rheumatology MIS-C clinical management guidelines with regard to immune modulation therapy
MIS-C was first described on 04/26/2020 American College of Rheumatology Clinical Guidance-Version 112 American College of Rheumatology Clinical Guidance-Version 216 American College of Rheumatology Clinical Guidance-Version 317
Date of online publication October 3, 2020 February 15, 2021 February 3, 2022
Primary Immunomodulation therapy IVIG and/or glucocorticoids
If contraindicated then Anakinra (IV or SC) IVIG IVIG and low-moderate dose glucocorticoids
Adjunct Immunomodulation therapy Glucocorticoids
If contraindicated then Anakinra (IV or SC) Low-moderate dose Glucocorticoids See above
Indication for adjunct Immunomodulation therapy N/A Shock and/or organ threatening disease
Rescue Immunomodulation therapy Anakinra (IV or SC) or in patients with contraindications to these treatments Low-moderate dose steroids (if not already given)
High dose glucocorticoids
Anakinra (IV or SQ), in patients with MIS-C and features of MAS, or in patients with contraindications to long-term use of glucocorticoids
Second dose of IVIG is not recommended High dose glucocorticoids
Anakinra (preferred anti-cytokine therapy)
Infliximab (except in patients with features of MAS)
Second dose of IVIG is not recommended
Indication for rescue Immunomodulation therapy Refractory to IVIG and/or glucocorticoids Refractory to IVIG and/or low-moderate glucocorticoids
“persistent fevers and/or ongoing and significant end-organ involvement” Persistent fevers and/or ongoing and significant end‐organ involvement
Special comment Immunomodulatory treatment may be withheld in some patients with mild symptoms
Patients may require a 2–3‐week, or even longer, taper of immunomodulatory medications, guided by serial laboratory and cardiac evaluations Immunomodulatory treatment may be withheld in some patients with mild symptoms
Patients may require a 2–3‐week, or even longer, taper of immunomodulatory medications, guided by serial laboratory and cardiac evaluations Immunomodulatory treatment may be withheld in some patients with mild symptoms
Patients may require a 2–3‐week, or even longer, taper of immunomodulatory medications, guided by serial laboratory and cardiac evaluations
Abbreviations: IV: Intravenous, IVIG: Intravenous immunoglobulin, MAS: macrophage activation syndrome, SC: subcutaneous
Table 2 Comparison of current published Society MIS-C Clinical Management guidelines with regard to immune modulation therapy
MIS-C was first described on 04/26/2020 PIMS-TS National Consensus Management Study Group18 Guidance from the Rheumatology Study Group of the Italian Society of Pediatrics19 American College of Rheumatology Clinical Guidance-Version 317 Practice Recommendations in Switzerland29
Rescue Immunomodulation therapy Second dose of IVIG
High dose glucocorticoids
Biological therapy should be considered as a third-line option in children who do not respond to IVIG and glucocorticoids (for those recruited in RECOVERY trial tocilizumab or standard of care,
If not recruited in RECOVERY trial the agent of choice for KD phenotype is infliximab and for non-specific presentation phenotype, the choice is left to clinician to choose from the following agents tocilizumab, anakinra, and infliximab Second dose IVIG
Anakinra High dose glucocorticoids
Anakinra (preferred anti-cytokine therapy)
Infliximab (except in patients with features of MAS)
Second dose of IVIG is not recommended Anakinra
Other biologics (Tocilizumab, Infliximab)
2nd dose IVIG
Indication for rescue Immunomodulation therapy Indication for Second dose of IVIG: Not responded or partially responded to the first dose
Indication for High dose glucocorticoids: Unwell 24 hours after infusion of intravenous immunoglobulin, particularly if they have ongoing pyrexia Persistent disease activity 48 h after first-line treatment Persistent fevers and/or ongoing and significant end‐organ involvement No clinical improvement 24 -36 hours after IVIG with persistent fever and/or inflammation
Special comment All children who meet the criteria for the RECOVERY trial should be invited to participate in the first stage of randomization for the trial
Management is divided according to phenotype: KD phenotype versus non-specific presentation phenotype
For the for the non-specific presentation phenotype, indications for therapy includes: evidence of CAA; meeting the criteria for toxic shock syndrome; evidence of progressive disease, and extended duration of fever (>5 days). In other words those not meeting above criteria can be observed.
No guidance was provided on tapering immunomodulatory medications Although different doses of steroids was suggested depending on severity and cardiac involvement, no guidance was provided on tapering immunomodulatory medications Immunomodulatory treatment may be withhold in some patients with mild symptoms
Patients may require a 2–3‐week, or even longer, taper of immunomodulatory medications, guided by serial laboratory and cardiac evaluations The guidelines suggest considering immunomodulation therapy (IVIG, prednisolone) in patients who presented with undefined inflammatory phenotype (not shock or KD phenotype)
In other words, immunomodulatory treatment may be withheld in some patients
Slow wean of steroids, taper over 2–6 weeks depending on the clinical course and considering the clinical and biochemical (such as CRP, D-Dimer, and ferritin levels) response
Abbreviations: CAA: coronary artery abnormality, HLH: hemophagocytic lymphohistiocytosis, IVIG: Intravenous immunoglobulin, KD: Kawasaki disease, PIMS-TS: Pediatric inflammatory multisystem syndrome temporally associated with COVID-19
To further highlight how early management decisions were based more on expert opinion and clinician experience with other diseases, one should consider how recommendations for steroid use in the management of children with MIS-C evolved. MIS-C displays similarities to other inflammatory diseases, including rheumatological conditions such as macrophage activation syndrome (MAS).20 Cytopenias, coagulopathy, and elevated ferritin and interleukins are shared features between MIS-C and MAS, albeit to a different degree.20 Accordingly, some centers, prioritized the use of steroids alone or in-combination with IVIG.21 , 22 In contrast, early in the pandemic some argued against the use of steroids in the management of MIS-C, evoking caution for steroid use in adults with active COVID-19 infection and lack of support for routine use in myocarditis, a common complication of MIS-C.8 , 23, 24, 25, 26 For some centers, anakinra, an interleukin-1 receptor antagonist, was the preferred agent, and was typically used as an adjunct or second-line therapy rather than corticosteroids.8
Current management algorithms and how they evolved during the pandemic.
Now within the third year of the pandemic, clinicians have more familiarity with the presentation and clinical course of MIS-C. The importance of expeditious initiation of immunomodulatory treatment remains a cornerstone in all the currently published guidelines for treatment. As previously noted in Table 2, most published recommendations include combination therapy of IVIG and steroids as the front-line therapy for MIS-C. Nonetheless, there has been much more practice variation regarding the use of steroids, with one early survey finding that only 14% of respondents used steroids for all patients.27 The benefits of steroids were related to their potential effectiveness in treating cytokine storm syndrome, and their added utility for patients with a shock-like presentation. However, there were initial concerns regarding immunosuppression in a potentially septic and bacteremic/viremic patient, as well as the overall concerns regarding adverse effects, including hyperglycemia, hypertension, agitation, hospital-acquired infection, and hip osteonecrosis.28 , 29 The presence and severity of adverse effects also increases with increased duration and dose, necessitating careful consideration of the risk versus benefits of steroid therapy.30 , 31 Other contributing factors were the lack of consensus on the specific indication, dosage, or type of steroids to use, institutional preference and comfort level with using other biologics, and practitioner reluctance to use steroids in the setting of KD translating to reluctance for MIS-C. One of the major turning points for steroid use came about with the publication of a non-randomized cohort study by Son et al. demonstrating superior cardiovascular outcomes at hospital discharge in a group treated with IVIG plus steroids compared to IVIG alone.32 They also found that adjunctive therapy was utilized less frequently in the combination group compared to the single regimen group, suggesting a lower need for escalation of therapy. These findings have been noted in other studies22 , including some studies suggesting that the use of steroids as monotherapy may be adequate treatment.33 While current data are compelling, the observational nature of these studies continues to result in practice variation and, at times, conflicting results and overall low level of evidence.34 , 35 Another consideration is the variation in the use of a tapering course of steroids. (Table 2) The recommendations for a taper remained relatively broad and non-specific (ranging from 2 to 6 weeks, depending on clinical response), consistent with the low level of supporting evidence.17, 18, 19 , 29
To address the diagnostic and treatment issues some professional societies convened multidisciplinary task forces to provide guidance on the management of MIS-C based primarily on expert consensus. For instance, the American College of Rheumatology (ACR) task force was composed of 9 pediatric rheumatologists, 2 adult rheumatologists, 2 pediatric cardiologists, 2 pediatric infectious disease specialists, and 1 pediatric critical care physician.12 , 16 , 17 The ACR task force developed consensus using a modified Delphi process based on the principle that forecasts (or decisions) from a structured group of individuals are more accurate than those from unstructured groups. The ACR used 2 rounds of anonymous voting and 2 webinars. The committee applied a 9-point Likert scale to determine each statement’s appropriateness (median scores of 1-3 as inappropriate, 4-6 as uncertain, and 7-9 as appropriate).Consensus was then rated as low, moderate, or high based on vote dispersion along the numeric scale. Approved ACR guidance statements required moderate or high levels of consensus. The ACR statement was intended as a “living document”, which would be modified in response to emerging data.12 , 16 , 17
While recommendations for first-line immunomodulatory therapy are approaching a relative consensus, choices for adjunct and/or escalation therapy remain highly variable. Some centers utilized a tiered treatment algorithm based on disease severity at presentation, although there is no consensus for classification of severity, and each center would develop their own criteria for severity and what warranted adjunctive therapy.36 Published recommendations continue to note use of a wide variety of medications and leave these decisions to the discretion of the treating provider (Table 2). Like KD, data to support a preferred treatment for refractory disease are lacking. In contrast to KD, a second dose of IVIG for MIS-C is not recommended by some (Table 2).37
Evidence based medicine to support the currently used therapies and impact of the management on clinical outcomes including cardiac outcomes (short term) given no clinical trials
As discussed earlier, upon the initial reports of MIS-C in early 2020, there was a lack of evidence and experience in treating this novel condition. Management was based on overlapping clinical features between MIS-C and other inflammatory syndromes, such as KD, and fortunately, patients often responded well to these established treatments.38 , 39 The following will summarize the current evidence, or lack of, since these initial reports with a focus on the three tenets of MIS-C management: treatment of shock, immunomodulatory therapies, and thromboprophylaxis.
Shock
Shock is a common presentation among MIS-C patients, with signs of cardiogenic, distributive, or hypovolemic shock. Regardless of the specific cause, patients should be treated with fluid resuscitation. Given the significant risk of left ventricular dysfunction in MIS-C, clinicians should be aware of potential dysfunction upon administering fluid and reassess frequently. There are no studies comparing vasoactive agents for patients with fluid-refractory shock, and guidance is based on existing protocols for pediatric shock. For pediatric septic shock, the Surviving Sepsis guidelines40 suggest using either epinephrine or norepinephrine rather than dopamine. Two studies have demonstrated lower risk of mortality41 and less organ dysfunction42 among those who received epinephrine instead of dopamine. With the concern of myocardial dysfunction though in MIS-C, the beta-agonist actions of epinephrine may be more favorable than the increase in systemic vascular resistance from norepinephrine.
Immunomodulatory therapy
Primary therapy
There have been no randomized clinical trials comparing immunomodulatory therapies. An early nonrandomized series from France compared 18 patients who received IVIG alone to 22 patients who received IVIG and intravenous methylprednisolone (0.8 mg/kg/day for 5 days). Dual therapy was associated with a faster recovery time for left ventricular systolic function (2.9 vs. 5.4 days), along with decreased length of stay (LOS) in the intensive care unit (ICU) (3.4 versus 5.3 days).43 In a subsequent retrospective study with propensity-matched analysis using the national surveillance system in France, 32 children received IVIG and methylprednisolone and 64 received IVIG alone.22 The methylprednisolone dose was 0.8-1 mg/kg every 12 hours (maximum 30 mg for 12 hours) for 5 days, although four patients received a bolus of 15-30 mg/kg/day for 3 days. Patients who received dual therapy had a more severe presentation, yet they had a more favorable outcome. Treatment with IVIG and methylprednisolone was associated with lower risk of treatment failure compared to IVIG alone (9% vs. 38%) and decreased likelihood of requiring intensification of therapy (9% vs. 31%), vasoactive agents (6% vs 23%), or acute left ventricular dysfunction (17% vs. 35%), along with shorter duration of stay in the ICU (4 vs. 6 days).
The larger US-based Overcoming COVID-19 surveillance registry performed a propensity score-matched comparison of 103 patients who received IVIG and glucocorticoids versus 103 who received IVIG alone.32 Methylprednisolone was the most common glucocorticoid received, primarily 2 mg/kg/day with fewer receiving pulse doses of 10-30 mg/kg/day. Dual therapy was associated with a lower risk of cardiovascular dysfunction on or after day 2, defined as a composite of left ventricular dysfunction or shock resulting in use of vasopressors (17% vs. 31%). Left ventricular dysfunction was lower among those receiving dual therapy (8% vs. 17%), as was the likelihood of receiving vasoactive agents (13% vs. 24%) or second-line therapy (34% vs. 70%).
An analysis by the Best Available Treatment Study (BATS) Consortium, published simultaneously with the Overcoming COVID-19 study, provided what seemed to be conflicting results.35 In an international observational cohort study of patients with suspected or confirmed MIS-C, 246 patients received primary treatment with IVIG alone, 208 IVIG and glucocorticoids, and 99 glucocorticoids alone. The primary outcomes were 1) a composite of inotropic support or mechanical ventilation by day 2 or later or death and 2) reduction in disease severity by day 2. There were no significant differences in outcomes among the treatment groups. A subgroup analysis among patients who met World Health Organization (WHO) criteria for MIS-C and received glucocorticoids alone had a lower risk of requiring respiratory support by day 2 or later or death compared to IVIG alone (OR 0.3 (0.1-0.85)). However, there are several potential limitations. The study included many patients with suspected MIS-C who did not meet diagnostic criteria, and only 12% of patients had left ventricular systolic dysfunction, less than reported in other series, suggestive of milder disease in this cohort.
A more recent single-center retrospective study reviewed the potential for steroids as monotherapy, including standard therapy (2 mg/kg/day methylprednisolone) or pulse-dosing.33 Propensity score analysis compared differing treatments for 179 patients with MIS-C (68 with steroids alone, 111 with IVIG and steroids). Steroid monotherapy was associated with similar rates of treatment failure but shorter steroid course duration (5 vs. 10 days) and shorter hospital LOS (5 vs. 6 days).
Escalation of therapy
Patients with MIS-C may decompensate quickly, and there should be a low threshold to escalate therapy. There are no clinical trials comparing treatment for refractory disease, although the most common reported therapies include high-dose glucocorticoids, anakinra, and infliximab, with a few earlier studies reporting use of tocilizumab.32 At least for one guidelines, two doses of IVIG are not recommended given the risk of volume overload and hemolytic anemia.17
In a small single-center study including 33 patients, 22 received ICU care. While all 22 received IVIG, 12 received second-line therapy with infliximab, all demonstrating improvement with no adverse reactions.44 A subsequent larger single-center retrospective study including 72 children with MIS-C, 20 received IVIG alone and 52 received IVIG and infliximab (10 mg/kg).45 Although infliximab was used as primary therapy in this instance rather than intensification, dual therapy was associated with less additional therapy (31% vs. 65%), shorter duration of ICU stays (1.8 vs. 3.3 days), decreased development of left ventricular systolic dysfunction (4% vs 20%), and faster decrease in C-reactive protein levels.
Thromboprophylaxis
Pediatric patients with both acute COVID-19 and MIS-C are presumed to have increased risk for thrombosis. In a multi-center, retrospective study of 138 MIS-C patients, 9 (6.5%) developed a thrombotic event: stroke (n=1), intracardiac thrombosis (n=1), and deep venous thrombosis (n=7), all in the setting of central venous catheters.46 One patient with a thrombotic event died while two others receiving anticoagulation had major bleeding events. Across the entire study of MIS-C and acute COVID-19 patients, age ≥12 years old and a D-dimer five times greater than the upper limit of normal were associated with thrombotic events.
Antiplatelet treatment with low-dose aspirin (ldASA) is generally recommended for all patients, based on a similar recommendation for KD, as well as the likelihood of platelet activation and evidence of microvascular thrombotic events in adults with acute COVID-19.47, 48, 49 It is less clear when and how aggressive thromboprophylaxis should be applied. While pediatric patients with acute COVID-19 appear to have a much lower risk of thrombotic complications compared to adults, patients with MIS-C, as well as older patients (age >12 years), a history of cancer, or the presence of a central venous catheter, were at the highest risk in one multi-center study.46 Left ventricular dysfunction is also felt to be a risk factor for cardiac thrombosis, with guidelines suggesting a need for anticoagulation if there is at least moderate left ventricular dysfunction.17,50
However, current recommendations for additional thromboprophylaxis in addition to ldASA are based on expert consensus, and include consideration of pre-COVID-19 risk factors, presence of cardiovascular abnormalities (severe ventricular dysfunction, large coronary aneurysms, etc.), markedly elevated D-dimer (>5 times the upper limit of normal), and risk factors for hospital-associated venous thrombosis and thromboembolism (VTE).51
Facilitating clinical decision-making and building evidence with a new condition
As discussed earlier, both the COVID-19 pandemic and the subsequent first descriptions of MIS-C arrived suddenly and dramatically.52 The first reported cases of MIS-C seemed particularly severe. This was presumably due to delays in presentation, diagnostic uncertainties, and a lack of realization early on of the need for prompt immunomodulatory therapy, together with the fact that the health care system was reeling from the impact of severe acute COVID-19 cases in adults.5 The assumption that children were largely spared the consequences of severe COVID-19 complications was quickly dashed, once the association with a prior COVID-19 infection was determined.13 , 53
How can we inform decision-making during the initial experience with a new condition?
For the first cases of MIS-C, care was appropriately supportive and empiric specific to the complications that were observed in individual patients. Given that this appeared to be a new condition, three important strategies were deployed that were aimed at informing clinical care and discovery. First, sharing of information was wide-spread, initially by peer-reviewed journals pivoting to prioritize rapid review and publication, often through open access to case reports and series.54 In addition, increased utilization of pre-print services expedited dissemination, albeit at the expense of peer-review and the occasional retraction of incorrect or misleading information.55 Second, multi-center data collections were quickly organized. These took the form of public health surveillance and reporting as well as the formation or leveraging of existing networks of clinicians, investigators and institutions.56 These efforts, together with the huge shift to virtual meetings, also facilitated greater and more rapid communication and sharing, as well as the critical examination of pooled data. For example, the International KD Registry (IKDR) held frequent open virtual webinars to review and discuss recent literature, and to provide a forum for open sharing and discussion regarding participants’ experience with clinical challenges related to MIS-C and KD, in addition to pivoting data collection to include MIS-C and acute KD patients.57 Of particular importance for surveillance and scientific inquiry, expert opinion and the evaluation of pooled data were used to define and refine initial case definitions for this new condition.5 , 8 , 58
The third strategy was to look for homologous conditions. It was immediately apparent that a large proportion of presumed MIS-C patients also met diagnostic criteria for KD.59 This homology led to the use of IVIG as initial therapy, the evidence-based standard of care for initial therapy of KD, followed by steroids and other inflammatory therapies in the event of insufficient response.10 The strategy appeared to be effective. However, without knowing the exact underlying pathophysiology of MIS-C, the development and use of alternative or adjunctive specific therapies, and the creation of treatment algorithms, is challenging. In addition, once one therapy quickly becomes the standard of care it becomes difficult to study what may be more specific and effective alternatives. This appears to be the case regarding whether IVIG or steroids or both together would be the more effective initial therapy for MIS-C.60
The strengths and challenges of clinical trials
Clinical decision-making should be based on critical appraisal and synthesis of the best available research evidence. However, clinician and institutional expertise and experience together with patient preferences for treatments and outcomes also should be incorporated to personalize the decision.61 This concept was first developed in the early 1990’s by David Sackett and colleagues, and has stood the test of time to become the ideal approach to clinical practice61, albeit with calls for refinements and expansion of scope.62 There is a hierarchy of evidence, with systematic reviews and meta-analyses being at the top, which ideally should be based on the next level of evidence being randomized controlled clinical trials. Clinical trials produce high quality evidence for two main reasons. First, random allocation to treatments, if successful and the sample size is large enough, reduces bias by balancing both known and unknown confounders equally between the groups to be compared. Hence, they provide the best evidence as to whether the intervention causally impacted the outcome. Second, the study design is inherently prospective in nature, allowing for standardization of study procedures and measurements and strategies to mitigate loss to follow-up. However, clinical trials are resource-intensive, have restrictive inclusion and exclusion criteria that often limit generalizability, only allow for the study of the randomized interventions, and are usually powered to detect the effect on a specific outcome. They are also prone to give equivocal answers, usually focused on average effect and not balancing different benefits and risks or specifying patient and clinical characteristics that mediate the effect of the interventions. In addition, real-world generalizability of the results depend on the degree of deviation between the clinical circumstance/patient being treated and the study design/participant population. Clinical trials are also particularly challenging for the study of rare conditions and rare or time-related outcomes (especially long-term outcomes), which has been true for both KD and MIS-C and for the outcome of coronary artery involvement and cardiovascular events.63, 64, 65, 66 The need for head-to-head or adjunctive trials necessitates a large participant population to generate necessary power for the study design, providing an added challenge for pediatric clinical trials of uncommon disease processes.
For ethical reasons, the comparator to a new intervention must include the minimal current standard of care. Lacking equipoise, placebo-controlled trials, particularly those in children, can be ethically challenged. In some cases, the new intervention must be provided as an adjunctive therapy to the standard of care to meet clinical equipoise requirements. This strategy makes it difficult to determine if a new intervention is superior to the standard of care alone. With the first cases of MIS-C, the rapid adoption of the empiric use of IVIG as the standard of care made it difficult to study the use of steroids alone as first line therapy.60 This rapid acceptance of a standard of care without strong evidence has further impeded the study of newer biologic agents, such as infliximab, etanercept and anakinra, which are currently reserved for patients not responsive to the current standard of care therapy or they are used as an adjunct therapy.
Expediting clinical trials during a pandemic
Given that randomized clinical trials provide the highest quality of research evidence, how might they be rapidly conceived, implemented and adapted to provide timely answers in new and perhaps rapidly evolving clinical scenarios? First, they must take advantage of existing infrastructure and expertise. Relevant consortia and networks of investigators might be engaged and empowered. They could be augmented by external experts, particularly if a multi-specialty and inter-disciplinary approach is needed. Some of these may be methodologic experts, particularly those with expertise in novel trial approaches/designs and statistical methods. Merging of networks may be desired. Relevant existing patient registries may pivot to prioritize the new patient population for data collection. There are precedents for nesting clinical trials within existing registries to provide efficiency in recruitment, data collection and study management at lower cost.67
Regarding study costs, funding agencies need to rapidly adapt and prioritize research into the new condition, which is what happened with the onset of the COVID-19 pandemic. However, investigators need to be prepared to respond rapidly to these prioritized requests for proposals, and funding agencies need to be prepared to provide expedited review and funding decisions. Both would have to ensure that the best science be put forward, evaluated, and funded, and that this is not compromised by the compressed timelines. Clinical trials must negotiate numerous regulatory hurdles that might need to be carefully expedited, including review and approvals from governmental regulatory bodies such as Health Canada and the US Food and Drug Administration, negotiation of multi-institutional contracts and data-sharing agreements, and obtaining institutional ethics review board approvals. The entire research governance structure might need to go into overdrive, together with strategies and oversight to minimize and mitigate risk.
Given the rarity of MIS-C, conventional and timely randomized clinical trial designs would not likely be feasible due to the small participant numbers limiting statistical power to reliably detect relative treatment effects. This persistent problem has plagued pediatric clinical trials, has hampered acquisition of the necessary evidence required to support regulatory approvals for pediatric drug use, and has made it hard to support strong recommendation statements in clinical practice guidelines.68 , 69
Given the importance of quickly developing high quality and reliable evidence on which to base clinical care for a new condition, a clear strategy is needed. Rather than relying on funding agencies to solely determine which investigator-initiated trials are performed, a clearinghouse is needed for coordination across agencies. Such a strategy would provide for rapid deployment of overlapping and complementary trials (perhaps shorter and smaller in scope, perhaps powered for reasonably justified surrogate endpoints), avoidance of competition for enrollment and facilitation of sharing of both expertise and resources. Therapies identified to be ineffective or potentially harmful should be abandoned as quickly as possible. Nimble governance would be essential to quickly and efficiently respond to findings and changes in the condition, such as with the emergence of new variants of the SARS-CoV-2 virus or the additional impact of vaccination.
Alternatives to conventional clinical trials might have a particular utility for evaluating new therapies for rare diseases. These may include pragmatic trials, use of cluster randomization, stepped wedge designs, factorial designs and Bayesian methods.70 , 71 Another strategy would be to utilize adaptative trials designs72, those that adapt or are modified based on new information that emerges as the trial progresses. That information most commonly derives from different types of pre-specified interim analysis strategies. Non-comparative analyses may lead to adjustments in sample size informed by improved estimates of variance or of baseline risk of the outcome. The interim analyses could be performed comparatively to assess for rigorous pre-specified criteria for stopping a trial early for either efficacy, futility or safety (group sequential design). If comparative interim analyses show futility in the overall group but efficacy in a pre-specified subpopulation then further enrollment may be restricted for that subpopulation, known as adaptive enrichment. Comparative interim analyses may also be used to plan modifications to treatment arms, such as dose-escalation, or adding or deleting certain treatment arms, often in comparison to a common control or comparator arm. Adaptations can be applied to treatment arm allocation based on comparative interim analyses, such as consecutive allocation aimed to balance baseline covariates between treatment arms or basing consecutive allocation on accumulating outcome data that may favor allocation to the more beneficial treatment (play the winner designs). All of these adaptation strategies must pre-planned and strictly monitored to maintain the overall integrity and rigor of the trial. In addition, since those adaptations described all entail interim analyses, a detailed statistical strategy must be in place to compensate or adjust for an increasing probability of a Type 1 error. These adaptation techniques could be applied in the setting of rapidly evolving clinical scenarios and emergence of new treatment strategies, such as was seen with the COVID-19 pandemic and MIS-C.
What is the role of observational data?
Observational data can be very useful to determine the clinical disease spectrum within a new condition. Such data can identify variations in clinical presentation and serve to refine case definitions, which were somewhat diverse for MIS-C after the onset of the pandemic. Observational datasets can also define the natural and modified history of a condition, which is critical information for determining patient prognosis and prediction of health care system resource requirements. Prospective cohort studies with defined inclusion and exclusion criteria and standardized and adjudicated data collection provide the optimum observational data. Patient reported outcomes should be formally elicited, and patient advocates actively engaged. For optimal estimates of incidence and risk, population-based studies with a defined denominator and complete reporting are needed. This limits the utility of passive surveillance systems (incomplete reporting) and administrative data (lack of granularity and standardization).
The primary limitation of using observational data to determine or compare outcomes of interventions is that the allocation to intervention type is non-random and, hence, subject to bias, since patients may get one intervention over another due to factors that may also influence the outcome. A number of statistical methods can be used to make the comparison fairer, but they only partially adjust for differences, and can only adjust for factors that were measured, in contrast to randomization which ideally creates equal groups balanced by unmeasured factors as well. Introduction of these biases represents the primary reason why studies using observational data to make non-randomized comparisons often yield differing results. A recent example for MIS-C were 3 multi-institutional observational studies regarding initial immunomodulatory treatment with IVIG alone versus IVIG with steroids. The study by Son et al. included 518 patients and noted, as mentioned earlier, a relative benefit of IVG with steroids on left ventricle dysfunction at and after 2 days.32 The comparison incorporated propensity score matching and covariate adjustment, and a further analysis using inverse probability weighting, both yielding similar results. The study by McArdle et al. included 614 patients and did not note a relative benefit of either steroids alone or in combination with IVIG over IVIG alone on a composite outcome of ventilatory or inotropic support or death.35 Likewise, these investigators used a similar statistical adjustment approach. The reasons for the disparity in the results for these two observational studies were explored by the combined teams of investigators. They noted differences in disease severity and cardiovascular involvement between these two studies, which also used differing case definitions.73 A further multi-institutional study by Ouldali et al. included 181 patients and noted a relative benefit of IVIG with steroids over IVIG alone on treatment failure, defined as persistent or recurrent fever, as well as other clinically relevant outcomes.22 These investigators used propensity score matching for statistical adjustment, with further analyses incorporating covariates and using inverse probability weighting. Given the potential for residual bias for each of these 3 studies, it is difficult to know where the true answer lies.
How do we make recommendations?
Recommendations for management strategies must be based on accurate and reliable evidence of effectiveness and safety. However, early in the experience with a new condition this evidence may be lacking, indirect, incomplete or at risk of being inaccurate. Expert opinion can be used as a starting point, but it must be used cautiously. The search for relevant experts should be broad and diverse, and all perspectives should be represented. The questions to be addressed should be carefully specified, and the starting point should be from a critical appraisal of what is already known as published in the literature where possible. Methods for reaching consensus should be fair, inclusive and transparent, such as with the Delphi approach and others.74
Eventually, research evidence will emerge and either complement. refine or replace expert opinion alone. Many studies, including observational studies, will make concluding statements that highlight the need for further research, but some will include statements regarding the clinical relevance of the findings and then proceed to make a recommendation. These recommendations should be viewed with caution and only in light of critical appraisal of the specific study.
Cumulative evidence and experience can be synthesized into higher level collections of recommendations that may have broader applicability. This is the process for development of guidelines, for which there are guidelines for creating guidelines (www.agreetrust.org). Pivotal to the process is an underlying pre-specified and detailed review and critical appraisal of the published research evidence, aimed at addressing key questions that are defined according to the patient, population or problem at hand, the intervention and its comparators, and the outcomes. The review may be performed by experts independent of those drafting the recommendation statements. An evidence review should culminate in the production of evidence tables, which ultimately guide decisions as to the wording and grading of recommendation statements which is in alignment with the overall quantity and quality of the appraised evidence. These tables should be published with the guideline statement and kept alive and updated with new evidence periodically. Recommendation statements relevant to MIS-C are characterized by lower classes, reflecting the magnitude and uncertainties in the size of the treatment effect or association, which are accompanied by lower levels of evidence, reflecting uncertainties about the precision of that treatment effect or association.12 , 16, 17, 18, 19 , 29
As may be seen in rare diseases, such as MIS-C, the evidence is often lacking or of suboptimal quality, yet guidance is needed.75 To that end, some advocated for the development of rapid guidelines but caution was raised in the adaptation of rapid guidelines as they often lack clarity around reason for development, how the quality of evidence was assessed, and how management of conflicts was handled.76 Gaps in recommendations can lead to inconsistencies and uncertainties in clinical practice which can impact patient outcomes. It is necessary to provisionally fill these gaps, and we are often left with expert opinion. However, great caution is needed. Expert opinion is often based on anecdotal practice which has not been systematically reviewed and appraised. It may be skewed by outlier experiences, firmly held opinions and outside influences, and it may be biased by local circumstances and practices. If expert opinion is to be used to fill recommendation gaps, it must be used with the provisions previously described.
The conclusion of any guideline development project should not only include an accompanying plan for knowledge translation, but also a detailed accounting of the evidence gaps and the research strategies to address them. This can be done by identifying gaps and challenges during the evidence review and incorporating findings into evidence tables that are updated. Methodologists should be included in this activity to inform research approaches. The results should be specified priori as an output in addition to the guideline itself. Guidelines are often developed by organizations which fund research, and the results should be used to inform funding priorities. Finally, the research strategy needs to have its own knowledge translation and implementation plan.
Penultimately, recommendations and guidelines are really based on the average treatment effects as derived from research evidence. Optimally, decisions regarding management need to move from this approach toward management that is precision-based.77
Conclusion
While results have been encouraging, the optimal treatment of children with MIS-C remains to be determined. Current therapies are based on expert opinion, similarities to other pediatric conditions like KD and multiple observational studies. Herein, we have provided guidance on developing future clinical trials for such a rare condition to help inform optimal treatment strategies.
Uncited reference
50..
Acknowledgments: The authors are grateful for Angela J. Doty, MD for her editorial assistance and for everyone treating and researching MIS-C along with our patients and their families.
Conflict of interest: None
Fundingand financial interest: Dr. Harahsheh is supported by a Sub-agreement from the Johns Hopkins University with funds provided by Grant No. R61HD105591 from the Eunice Kennedy Shriver National Institute of Child Health & Human Development and the Office of the Director, National Institute of Health (OD). Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the Eunice Kennedy Shriver National Institute of Child Health & Human Development, the Office of the Director, National Institute of Health (OD), the National Institute of Health, the NIBIB, the NHLBI, or the Johns Hopkins University.
Non-financial interest: Dr. Harahsheh serves as a scientific advisory board member of OP2 DRUGS (“OP2”). This advisory position has no relevant disclosures for this manuscript.
Ethics approval: Not applicable
Availability of data and material: N/A.
Author contributions: All authors were involved in drafting the article or revising it critically for important intellectual content. All authors approved the final version to be published.
==== Refs
References:
1 Adhikari S.P. Meng S. Wu Y.J. Epidemiology, causes, clinical manifestation and diagnosis, prevention and control of coronavirus disease (COVID-19) during the early outbreak period: a scoping review Infect Dis Poverty. Mar 17 9 1 2020 29 10.1186/s40249-020-00646-x
2 DeBiasi R.L. Song X. Delaney M. Severe Coronavirus Disease-2019 in Children and Young Adults in the Washington, DC, Metropolitan Region J Pediatr. Aug 223 2020 199 203 e1 10.1016/j.jpeds.2020.05.007
3 Lauer R.M. Lee J. Clarke W.R. Predicting adult cholesterol levels from measurements in childhood and adolescence: the Muscatine Study Bull N Y Acad Med. Dec 65 10 1989 1127 1142 ; discussion 1154-60
4 de Ferranti S.D. Steinberger J. Ameduri R. Cardiovascular Risk Reduction in High-Risk Pediatric Patients: A Scientific Statement From the American Heart Association Circulation. Mar 26 139 13 2019 e603 e634 10.1161/CIR.0000000000000618
5 Whittaker E. Bamford A. Kenny J. Clinical Characteristics of 58 Children With a Pediatric Inflammatory Multisystem Syndrome Temporally Associated With SARS-CoV-2 JAMA 324 3 Jul 21 2020 259 269 10.1001/jama.2020.10369 32511692
6 Harahsheh A.S. Dahdah N. Newburger J.W. Missed or delayed diagnosis of Kawasaki disease during the 2019 novel coronavirus disease (COVID-19) pandemic J Pediatr. Jul 222 2020 261 262 10.1016/j.jpeds.2020.04.052
7 Molloy E.J. Nakra N. Gale C. Dimitriades V.R. Lakshminrusimha S. Multisystem inflammatory syndrome in children (MIS-C) and neonates (MIS-N) associated with COVID-19: optimizing definition and management Pediatr Res Sep 1 2022 10.1038/s41390-022-02263-w
8 Harahsheh A.S. Krishnan A. DeBiasi R.L. Cardiac echocardiogram findings of severe acute respiratory syndrome coronavirus-2-associated multi-system inflammatory syndrome in children Cardiol Young Aug 5 2021 1 9 10.1017/S1047951121003024
9 Tritt A. Abda I. Dahdah N. Review of MIS-C Clinical Protocols and Diagnostic Pathways: Towards a Consensus Algorithm CJC Pediatric and Congenital Heart Disease 1 2 2022:Apr 86 93
10 DeBiasi R.L. Harahsheh A.S. Srinivasalu H. Multisystem Inflammatory Syndrome of Children: Sub-phenotypes, Risk Factors, Biomarkers, Cytokine Profiles and Viral Sequencing J Pediatr Jun 7 2021 10.1016/j.jpeds.2021.06.002
11 Feldstein L.R. Tenforde M.W. Friedman K.G. Characteristics and Outcomes of US Children and Adolescents With Multisystem Inflammatory Syndrome in Children (MIS-C) Compared With Severe Acute COVID-19 JAMA. Mar 16 325 11 2021 1074 1087 10.1001/jama.2021.2091
12 Henderson L.A. Canna S.W. Friedman K.G. American College of Rheumatology Clinical Guidance for Multisystem Inflammatory Syndrome in Children Associated With SARS-CoV-2 and Hyperinflammation in Pediatric COVID-19: Version 1 Arthritis Rheumatol. Nov 72 11 2020 1791 1805 10.1002/art.41454
13 Loke Y.H. Berul C.I. Harahsheh A.S. Multisystem inflammatory syndrome in children: Is there a linkage to Kawasaki disease? Trends Cardiovasc Med 30 7 Oct 2020 389 396 10.1016/j.tcm.2020.07.004 32702413
14 Hejazi O.I. Loke Y.H. Harahsheh A.S. Short-term Cardiovascular Complications of Multi-system Inflammatory Syndrome in Children (MIS-C) in Adolescents and Children Curr Pediatr Rep Oct 22 2021 1 11 10.1007/s40124-021-00258-5 33457108
15 McCrindle B.W. Rowley A.H. Newburger J.W. Diagnosis, Treatment, and Long-Term Management of Kawasaki Disease: A Scientific Statement for Health Professionals From the American Heart Association Circulation 135 17 Apr 25 2017 e927 e999 10.1161/CIR.0000000000000484 28356445
16 Henderson L.A. Canna S.W. Friedman K.G. American College of Rheumatology Clinical Guidance for Multisystem Inflammatory Syndrome in Children Associated With SARS-CoV-2 and Hyperinflammation in Pediatric COVID-19: Version 2 Arthritis Rheumatol. Apr 73 4 2021 e13 e29 10.1002/art.41616
17 Henderson L.A. Canna S.W. Friedman K.G. American College of Rheumatology Clinical Guidance for Multisystem Inflammatory Syndrome in Children Associated With SARS-CoV-2 and Hyperinflammation in Pediatric COVID-19: Version 3 Arthritis Rheumatol. Apr 74 4 2022 e1 e20 10.1002/art.42062
18 Harwood R. Allin B. Jones C.E. A national consensus management pathway for paediatric inflammatory multisystem syndrome temporally associated with COVID-19 (PIMS-TS): results of a national Delphi process Lancet Child Adolesc Health. Feb 5 2 2021 133 141 10.1016/S2352-4642(20)30304-7
19 Cattalini M. Taddio A. Bracaglia C. Childhood multisystem inflammatory syndrome associated with COVID-19 (MIS-C): a diagnostic and treatment guidance from the Rheumatology Study Group of the Italian Society of Pediatrics Ital J Pediatr. Feb 8 47 1 2021 24 10.1186/s13052-021-00980-2
20 Lee P.Y. Day-Lewis M. Henderson L.A. Distinct clinical and immunological features of SARS-CoV-2-induced multisystem inflammatory syndrome in children J Clin Invest. Nov 2 130 11 2020 5942 5950 10.1172/JCI141113
21 Harthan A.A. Nadiger M. McGarvey J.S. Early combination therapy with immunoglobulin and steroids is associated with shorter ICU length of stay in Multisystem Inflammatory Syndrome in Children (MIS-C) associated with COVID-19: A retrospective cohort analysis from 28 U.S. Hospitals Pharmacotherapy 42 7 Jul 2022 529 539 10.1002/phar.2709 35661394
22 Ouldali N. Toubiana J. Antona D. Association of Intravenous Immunoglobulins Plus Methylprednisolone vs Immunoglobulins Alone With Course of Fever in Multisystem Inflammatory Syndrome in Children JAMA. Mar 2 325 9 2021 855 864 10.1001/jama.2021.0694
23 Li Y. Yu Y. Chen S. Liao Y. Du J. Corticosteroids and Intravenous Immunoglobulin in Pediatric Myocarditis: A Meta-Analysis Front Pediatr 7 2019 342 10.3389/fped.2019.00342 31475124
24 Caldeira D. Lopes L.R. Vaz-Carneiro A. Costa J. Cochrane Corner: Corticosteroids for viral myocarditis Rev Port Cardiol. Jan 34 1 2015 65 67 10.1016/j.repc.2014.08.006
25 Julia P. Young H.H. Buckberg G.D. Kofsky E.R. Bugyi H.I. Studies of myocardial protection in the immature heart. II: Evidence for importance of amino acid metabolism in tolerance to ischemia Journal of Thoracic and Cardiovascular Surgery 100 1990 888 895 2246911
26 Belhadjer Z. Meot M. Bajolle F. Acute Heart Failure in Multisystem Inflammatory Syndrome in Children in the Context of Global SARS-CoV-2 Pandemic Circulation 142 5 Aug 4 2020 429 436 10.1161/CIRCULATIONAHA.120.048360 32418446
27 Elias M.D. McCrindle B.W. Larios G. Management of Multisystem Inflammatory Syndrome in Children Associated With COVID-19: A Survey From the International Kawasaki Disease Registry CJC Open. Nov 2 6 2020 632 640 10.1016/j.cjco.2020.09.004
28 Weusten B.L. Jacobs J.W. Bijlsma J.W. Corticosteroid pulse therapy in active rheumatoid arthritis Semin Arthritis Rheum. Dec 23 3 1993 183 192 10.1016/s0049-0172(05)80039-3
29 Schlapbach L.J. Andre M.C. Grazioli S. Best Practice Recommendations for the Diagnosis and Management of Children With Pediatric Inflammatory Multisystem Syndrome Temporally Associated With SARS-CoV-2 (PIMS-TS; Multisystem Inflammatory Syndrome in Children, MIS-C) in Switzerland Front Pediatr 9 2021 667507 10.3389/fped.2021.667507
30 Min K.H. Rhee C.K. Jung J.Y. Suh M.W. Characteristics of adverse effects when using high dose short term steroid regimen Korean J Audiol. Sep 16 2 2012 65 70 10.7874/kja.2012.16.2.65
31 Schacke H. Docke W.D. Asadullah K. Mechanisms involved in the side effects of glucocorticoids Pharmacol Ther. Oct 96 1 2002 23 43 10.1016/s0163-7258(02)00297-8
32 Son M.B.F. Murray N. Friedman K. Multisystem Inflammatory Syndrome in Children - Initial Therapy and Outcomes N Engl J Med 385 1 Jul 1 2021 23 34 10.1056/NEJMoa2102605 34133855
33 Villacis-Nunez D.S. Jones K. Jabbar A. Short-term Outcomes of Corticosteroid Monotherapy in Multisystem Inflammatory Syndrome in Children JAMA Pediatr 176 6 Jun 1 2022 576 584 10.1001/jamapediatrics.2022.0292 35344042
34 Harris SI, Balaban RS, Barrett L, Mandel LJ. Mitochondrial respiratory capacity and Na+ and K+-dependent adenosine triphosphatase-mediated ion transport in the intact renal cell. Journal of Biological Chemistry J Biol Chem. 1981;256(20, October 25):10319-10328.
35 McArdle A.J. Vito O. Patel H. Treatment of Multisystem Inflammatory Syndrome in Children N Engl J Med 385 1 Jul 1 2021 11 22 10.1056/NEJMoa2102968 34133854
36 Jonat B. Gorelik M. Boneparth A. Multisystem Inflammatory Syndrome in Children Associated With Coronavirus Disease 2019 in a Children's Hospital in New York City: Patient Characteristics and an Institutional Protocol for Evaluation, Management, and Follow-Up Pediatr Crit Care Med. Mar 22 3 1 2021 e178 e191 10.1097/PCC.0000000000002598
37 Hering JP, Schroder T, Singer D, Hellige G. Influence of pH management on hemodynamics and metabolism in moderate hypothermia. Journal of Thoracic and Cardiovascular Surgery. 1992;104(5, November 1992):1388-1395.
38 Riphagen S. Gomez X. Gonzalez-Martinez C. Wilkinson N. Theocharis P. Hyperinflammatory shock in children during COVID-19 pandemic Lancet 395 10237 May 23 2020 1607 1608 10.1016/S0140-6736(20)31094-1 32386565
39 Verdoni L. Mazza A. Gervasoni A. An outbreak of severe Kawasaki-like disease at the Italian epicentre of the SARS-CoV-2 epidemic: an observational cohort study Lancet 395 10239 Jun 6 2020 1771 1778 10.1016/S0140-6736(20)31103-X 32410760
40 Weiss S.L. Peters M.J. Alhazzani W. Surviving Sepsis Campaign International Guidelines for the Management of Septic Shock and Sepsis-Associated Organ Dysfunction in Children Pediatr Crit Care Med. Feb 21 2 2020 e52 e106 10.1097/PCC.0000000000002198
41 Ventura A.M. Shieh H.H. Bousso A. Double-Blind Prospective Randomized Controlled Trial of Dopamine Versus Epinephrine as First-Line Vasoactive Drugs in Pediatric Septic Shock Crit Care Med. Nov 43 11 2015 2292 2302 10.1097/CCM.0000000000001260
42 Ramaswamy K.N. Singhi S. Jayashree M. Bansal A. Nallasamy K. Double-Blind Randomized Clinical Trial Comparing Dopamine and Epinephrine in Pediatric Fluid-Refractory Hypotensive Septic Shock Pediatr Crit Care Med. Nov 17 11 2016 e502 e512 10.1097/PCC.0000000000000954
43 Belhadjer Z. Auriau J. Meot M. Addition of Corticosteroids to Immunoglobulins Is Associated With Recovery of Cardiac Function in Multi-Inflammatory Syndrome in Children Circulation 142 23 Dec 8 2020 2282 2284 10.1161/CIRCULATIONAHA.120.050147 33112651
44 Abdel-Haq N. Asmar B.I. Deza Leon M.P. SARS-CoV-2-associated multisystem inflammatory syndrome in children: clinical manifestations and the role of infliximab treatment Eur J Pediatr Jan 16 2021 10.1007/s00431-021-03935-1
45 Cole LD, Osborne CM, Silveira LJ, et al. IVIG Compared to IVIG Plus Infliximab in Multisystem Inflammatory Syndrome in Children. Pediatrics. Sep 22 2021;doi:10.1542/peds.2021-052702
46 Whitworth H. Sartain S.E. Kumar R. Rate of thrombosis in children and adolescents hospitalized with COVID-19 or MIS-C Blood 138 2 Jul 15 2021 190 198 10.1182/blood.2020010218 33895804
47 Ankola A.A. Bradford V.R. Newburger J.W. Coagulation profiles and viscoelastic testing in multisystem inflammatory syndrome in children Pediatr Blood Cancer 68 12 Dec 2021 e29355 10.1002/pbc.29355
48 Sharathkumar A.A. Faustino E.V.S. Takemoto C.M. How we approach thrombosis risk in children with COVID-19 infection and MIS-C Pediatr Blood Cancer 68 7 Jul 2021 e29049 10.1002/pbc.29049
49 Chow J.H. Khanna A.K. Kethireddy S. Aspirin Use Is Associated With Decreased Mechanical Ventilation, Intensive Care Unit Admission, and In-Hospital Mortality in Hospitalized Patients With Coronavirus Disease 2019 Anesth Analg 132 4 Apr 1 2021 930 941 10.1213/ANE.0000000000005292 33093359
50 Giglia T.M. Massicotte M.P. Tweddell J.S. Prevention and treatment of thrombosis in pediatric and congenital heart disease: a scientific statement from the American Heart Association Circulation 128 24 Dec 17 2013 2622 2703 10.1161/01.cir.0000436140.77832.7a 24226806
51 Goldenberg N.A. Sochet A. Albisetti M. Consensus-based clinical recommendations and research priorities for anticoagulant thromboprophylaxis in children hospitalized for COVID-19-related illness J Thromb Haemost. Nov 18 11 2020 3099 3105 10.1111/jth.15073
52 Boekholdt S.M. Hovingh G.K. Mora S. Very low levels of atherogenic lipoproteins and the risk for cardiovascular events: a meta-analysis of statin trials J Am Coll Cardiol 64 5 Aug 5 2014 485 494 10.1016/j.jacc.2014.02.615 25082583
53 Harahsheh A.S. Dahdah N. Newburger J.W. Reply. J Pediatr Jun 30 2020 10.1016/j.jpeds.2020.06.072
54 Downing S.E. Lee J.C. Taylor J.F.N. Cardiac function and metabolism during cholinergic stimulation in the newborn lamb American Journal of Physiology AM J Physiol 233 1977 H451 457
55 Ravinetto R. Caillet C. Zaman M.H. Preprints in times of COVID19: the time is ripe for agreeing on terminology and good practices BMC Med Ethics 22 1 Jul 28 2021 106 10.1186/s12910-021-00667-7 34320970
56 Feldstein L.R. Rose E.B. Horwitz S.M. Multisystem Inflammatory Syndrome in U.S. Children and Adolescents N Engl J Med Jun 29 2020 10.1056/NEJMoa2021680
57 Parrish M.D. Ayres N.A. Kendrick B.T. Fixler D.E. Maturational differences in the silated isovolumic rabbit heart American Journal of Physiology AM J Physiol 251 1986 H1143 H1148
58 Haney E.M. Huffman L.H. Bougatsos C. Freeman M. Steiner R.D. Nelson H.D. Screening and treatment for lipid disorders in children and adolescents: systematic evidence review for the US Preventive Services Task Force. Research Support, U.S. Gov't, P.H.S. Review Pediatrics. Jul 120 1 2007 e189 214 10.1542/peds.2006-1801
59 McCrindle B.W. Manlhiot C. SARS-CoV-2-Related Inflammatory Multisystem Syndrome in Children: Different or Shared Etiology and Pathophysiology as Kawasaki Disease? JAMA 324 3 Jul 21 2020 246 248 10.1001/jama.2020.10370 32511667
60 DeBiasi R.L. Immunotherapy for MIS-C - IVIG, Glucocorticoids, and Biologics N Engl J Med 385 1 Jul 1 2021 74 75 10.1056/NEJMe2108276 34133878
61 Sackett D.L. Rosenberg W.M. Gray J.A. Haynes R.B. Richardson W.S. Evidence based medicine: what it is and what it isn't BMJ 312 7023 Jan 13 1996 71 72 10.1136/bmj.312.7023.71 8555924
62 Sheridan D.J. Julian D.G. Achievements and Limitations of Evidence-Based Medicine J Am Coll Cardiol 68 2 Jul 12 2016 204 213 10.1016/j.jacc.2016.03.600 27386775
63 McCrindle B.W. Rowley A.H. Improving coronary artery outcomes for children with Kawasaki disease Lancet. Mar 16 393 10176 2019 1077 1078 10.1016/S0140-6736(18)33133-7
64 McCrindle B.W. Selamet Tierney E.S. Acute Treatment for Kawasaki Disease: Challenges for Current and Future Therapies J Pediatr. May 184 2017 7 10 10.1016/j.jpeds.2017.01.072
65 Tremoulet A.H. Jain S. Jaggi P. Infliximab for intensification of primary therapy for Kawasaki disease: a phase 3 randomised, double-blind, placebo-controlled trial Lancet 383 9930 May 17 2014 1731 1738 10.1016/S0140-6736(13)62298-9 24572997
66 Portman M.A. Dahdah N.S. Slee A. Etanercept With IVIg for Acute Kawasaki Disease: A Randomized Controlled Trial Pediatrics 143 6 Jun 2019 10.1542/peds.2018-3675
67 Wang J. Nong L. Wei Y. Qin S. Zhou Y. Tang Y. Association of interleukin-12 polymorphisms and serum IL-12p40 levels with osteosarcoma risk DNA Cell Biol 32 10 Oct 2013 605 610 10.1089/dna.2013.2098 23991654
68 Torok R.D. Li J.S. Kannankeril P.J. Recommendations to Enhance Pediatric Cardiovascular Drug Development: Report of a Multi-Stakeholder Think Tank J Am Heart Assoc. Feb 10 7 4 2018 10.1161/JAHA.117.007283
69 Harris K.C. Mackie A.S. Dallaire F. Unique Challenges of Randomised Controlled Trials in Pediatric Cardiology Can J Cardiol. Sep 37 9 2021 1394 1403 10.1016/j.cjca.2021.06.013
70 Psioda M.A. Xue X. A Bayesian Adaptive Two-Stage Design for Pediatric Clinical Trials J Biopharm Stat. Nov 1 30 6 2020 1091 1108 10.1080/10543406.2020.1821704
71 Wang Y. Travis J. Gajewski B. Bayesian adaptive design for pediatric clinical trials incorporating a community of prior beliefs BMC Med Res Methodol 22 1 Apr 21 2022 118 10.1186/s12874-022-01569-x 35448963
72 Ellis D. Thomas R.C. Direct measurement of the intracellular pH of mammalian cardiac muscle Journal of Physiology J Physiol 262 1976 755 771 13205
73 Melgar M. Seaby E.G. McArdle A.J. Treatment of Multisystem Inflammatory Syndrome in Children: Understanding Differences in Results of Comparative Effectiveness Studies ACR Open Rheumatol 4 9 Sep 2022 804 810 10.1002/acr2.11478 35759535
74 Jones J. Hunter D. Consensus methods for medical and health services research BMJ 311 7001 Aug 5 1995 376 380 10.1136/bmj.311.7001.376 7640549
75 Neumann I. Schunemann H.J. Guideline groups should make recommendations even if the evidence is considered insufficient CMAJ. Jan 192 2 13 2020 E23 E24 10.1503/cmaj.190144
76 Kowalski S.C. Morgan R.L. Falavigna M. Development of rapid guidelines: 1. Systematic survey of current practices and methods Health Res Policy Syst 16 1 Jul 13 2018 61 10.1186/s12961-018-0327-8 30005712
77 Tonelli M.R. Shirts B.H. Knowledge for Precision Medicine: Mechanistic Reasoning and Methodological Pluralism JAMA. Nov 7 318 17 2017 1649 1650 10.1001/jama.2017.11914
| 36455760 | PMC9705008 | NO-CC CODE | 2022-12-01 23:19:32 | no | Can J Cardiol. 2022 Nov 29; doi: 10.1016/j.cjca.2022.11.011 | utf-8 | Can J Cardiol | 2,022 | 10.1016/j.cjca.2022.11.011 | oa_other |
==== Front
Int J Disaster Risk Reduct
Int J Disaster Risk Reduct
International Journal of Disaster Risk Reduction
2212-4209
Elsevier Ltd.
S2212-4209(22)00688-4
10.1016/j.ijdrr.2022.103469
103469
Article
The public needs more: The informational and emotional support of public communication amidst the Covid-19 in China
Zhu Ruilin a
Hu Xuan b∗
a Management Science, Lancaster University Management School Lancaster United Kingdom LA1 4YX, UK
b School of Public Policy and Administration, Chongqing University, 174 Shazheng St, Chongqing, 440044, China
∗ Corresponding author.
29 11 2022
1 2023
29 11 2022
84 103469103469
8 7 2022
26 11 2022
26 11 2022
© 2022 Elsevier Ltd. All rights reserved.
2022
Elsevier Ltd
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Public communication is critical for responding to disasters. However, most research on public communication is largely focused on its informational support function, overlooking the emotional support that could equally offer. This study takes the lead to investigate their separate impacts. In particular, the variable public engagement, which is a function of the number of Shares, Likes, and Comments in a particular post, is introduced to benchmark the effect of public communication. Besides, considering the evolving nature of the crisis, their dynamic impacts across different COVID-19 pandemic stages are examined. Data from Dec 2019 to Jul 2020 were collected from 17 provincial government-owned social media (Weibo) accounts across COVID-19 in China with a Natural Language Processing-based method to compute the strengths of informational support and emotional support strength. An econometric model is then proposed to explore the impacts of two supports. The findings are twofold: the impact of emotional support on public engagement is empirically confirmed in the study, which is not in lockstep with the informational support; and their impacts on public communication are dynamic rather than static across stages throughout the crisis. We highlighted the importance of emotional support in public engagement by deriving its impact separately from informational support. The findings suggest incorporating both social supports to create stronger public communication tactics during crises.
Keywords
COVID-19
Emotional support
Government-owned social media
Social support theory
Public communication
==== Body
pmc1 Introduction
Government agencies have become unprecedentedly relied on social media for public communication [1], especially during emergent situations [2]. Its impelling advantage for this role is genuine and deeply ingrained communication amongst users given that social media entails the systemic capacity to the public information and warning approaches [3]. Comparatively, the information on social media is updated in a highly prompt manlner [4] and its flat networked structure transmits the information timely to a wider range of audiences [5]. Those affordances together ensure that social media can efficiently and effectively disseminate situational awareness information, which is of utmost importance in a crisis context [6,7]. Particularly, the restrictive measures involving self-isolation, quarantine, and city lockdown at the height of Covid-19 have largely reduced mobility, resulting in social media as a prominent channel for the government authorities to broadcast and disseminate information [8,9].
In addition to its information influences, there is an increasing understanding that social media can be leveraged to impose psychological impacts on social communities [10,11]. Specifically, social media can create a mutual aid environment, providing an outlet for social interaction and voicing fear, and offering a voluntary reciprocal exchange of resources and services for mutual benefit [12]. Nevertheless, most extant studies emphases on mutual aid amongst the public, overlooking the situation between the government (the sender) and the public (the receivers), where psychological influence is transposed through public communication as the public are seeking psychological support in addition to information through social media [13]. Indeed, the public communication posts released by government-owned social media play a prominent role not only in publishing pandemic information but also in providing emotional support (e.g., encouragement, and sympathy) during the COVID-19 pandemic [10]. As a result, while a multitude of studies have posited that emotional support can be conveyed through public communication [14,15], empirical evidence on how the government can exploit social media to transpose emotional support to the online community remains scant.
Furthermore, the devastating pandemic has gripped the globe for more than two years as of February 9, 2022, when Sweden first set to lift all coronavirus restrictions. Given this prolonged duration of the pandemic, the impacts of both informational and emotional support are prone to be rather complex than fixed. According to [16]; the impact of web service is affected by public expectations and possible disconfirmation between such expectations and web performance. Waves of Covid-19 compounded by rounds of lockdowns have altered the public's expectations for public communication with the government [8], giving rise to changes in their expectations for social media [17]. In such a sense, the impact of social media is prone to be dynamic throughout the emergency rather than static. Further, amidst the emergency, reasonable caution has been allocated to the “infodemic” or information overload [18]. It is identified as a pressing issue as it hinders the effective dissemination of information on social media, and may further impose negative impacts on an individual's mental well-being by triggering stress, frustration, dissatisfaction, and feeling of loss of control [19]. In addition, the interim public policies formulated in the first place during crises are not often fully supported by sufficient scientific evidence when being disseminated via social media [20], which may lead to confusion and/or uncertainty [21]. Likely, the information provided through social media may not always stay in line with the public's expectations [22,23]. To this end, instead of assuming social media's support are constantly positive, we must accept that the impact of information support and emotional support can be negative. It thus highlights the need of making the best use of social media between the government and the public for desirable outcomes.
Considering the above, we aim to examine two research questions in this study:(1) What is the empirical evidence for emotional support along with informational support on Weibo amidst COVID-19 in China?
(2) How did the impacts and dynamic changes of informational support and emotional support present during the emergency?
To answer these questions, we utilized Natural Language Processing (NLP) techniques to investigate and quantify the information support and emotional support from the government to the public by analyzing the social media data from 17 government-owned Weibo in China. We then deployed public engagement as the proxy for the effect of public communication and an econometric model is then proposed to explore the impacts of two supports. Then, we deployed COVID-19 as the context because it is an ongoing global event, which offers us long enough time and sufficient data to empirically collect and analyze data.
2 Literature review
2.1 Public communication
It has been well documented that public communication has a significant impact in emergent situations to limit the scope and mitigate the impacts of adverse events through disseminating timely and verified information to wide public audiences [24,25]. According to [26]; the role of public communication in crisis is three-fold, it improves situational informing, facilitates information exchange, and supports government reputation restoration.
The importance of early warning and risk information cannot be overstated during disasters. Public communication address this need by reaching the public timely and informing them of the nature, magnitude, and significance of the disaster, its associated risk, and possible coping strategies through the production of the public message [27]. This process seeks to alert individuals, provide protective action guidance and induce a public behavioral change in alleviating the threat [27,28]. Particularly during the Covid-19 pandemic, an absence of verified information has caused “panic buying” in many parts of the world [29,30]. Such behavior not only impairs the government's central effort in mobilizing the resource but also stressing to the psychological distress among the citizens [10].
Its next role is surrounding information exchange, where public feedback is sought on specific procedures and/or policies to further address public concern [31]. This process emphasizes developing communication strategies that respond to and anticipate the public's needs [27,[32], [33], [34]]. It is widely argued that public service needs to account for the demand sides [35,36]. An absence of communication between the government and the public, especially during disasters, may consequence in mismatching demand. As stated above, rumors or unverified information may lead to the public's psychological distress, distrust in the government, and non-compliance behaviors during the disaster response, which imposes significant challenges to the ability of society in coping with the disasters. The imbalance of demand-supply relationships is prone to further result in a waste of public resources and improving the effectiveness of public communication.
In addition, good public communication can also be exploited to restore public trust in government and minimalizes the potential reputation loss caused by rumors or misinformation [22]. When a crisis occurs, rumors may distort the truth, further reinforcing the public's distrust of the government authorities [37]. If it is not properly managed, such a miscommunication would erode public trust in the government [38]. Good public communication can not only facilitate the clarification of rumors, misunderstandings, and distorted facts but also signify the government's responsibility and accountability to manage disasters [39], which are drivers for the public's trust in the government. Further,Rosenberg [40]it is argued that the level of citizens' compliance with policy reflects their level of trust in government [40], whereas a low level of trust may result in non-compliance behaviors and even social chaos [41].
2.2 Social media based public communication
Compared to traditional medium (e.g., radio, television), online medium, such as social media, has quickly evolved into a new impetus for public communication during a crisis because they can greatly address the public's needs [32,42]. As an alternative to traditional media, social media is considered a reliable channel for situational informing, information exchange, and reputation restoration [10,12,43,44].
Its remarkable benefit in public communication is ingrained in its disperse networked structure. Such structure, as opposed to the conventional top-down hierarchical public communication paradigm, enables government information to communicate directly and promptly to a wider range of people [5]. This benefit has been empirically observed in scenarios such as the Haitian earthquake [45], Hurricane Sandy [46], and the Covid-19 pandemic [12,43,44]. Particularly during Covid-19 when strict quarantine and city lockdown measures are taken, social media has become an indispensable tool for public communication [12]. Such rapid information and warning delivery facilitate the public to prepare for the coming risk and adverse impact, which is foremost in the public's response to the crisis.
Second, with the increasing adoption of social media amongst government agencies, such as the U.S. Federal Emergency Management Agency [45,46], the National Health Commission of China [43], in promoting public communication during crises, the information posted is carefully scrutinized and validated [47], further enhancing the credibility. The timely and trustworthy information from the government in the online medium is most likely to exert positive impacts on bundles of multiple and heterogeneous aspirations, values, and perspectives between governments and the public [48,49].
In addition, online mediums have radically revolted public communication during crises by bringing forward two-way communications [50]. Indeed, one-way asymmetrical communication might be efficient in terms of speed [51], but two-way dialogical interaction is more effective in information exchange because it enables the public to voice their needs and concerns while also making it easier for government to collect of first-hand information (ibid). These information exchange activities will serve as a bottom-up channel to inform the government of better disaster situational awareness, which is critical for policy-forming and policy development amid a crisis [52]. Moreover, recent studies have implied that social media shift the role of the public in public communication from a passive recipient to an active information seeker [53] or even an information service co-providers [44].
Nevertheless, most of the existing studies on the possible influence of public communication during disasters are revolving around the information service provision, sporadic evidence also implies that beyond information, online public communication can also convey positive emotions, perception of support, and companionship [10].
2.3 Social support in social media based public communication during the crisis
Social support theory [14] provides a theoretical foundation to reveal the impact of both informational and emotional support through social media based public communication amidst an emergency Broadly, social support has been defined in the literature as the assistance and protection given to others, especially to individuals [54], shielding them from precarious events and adverse effects [55]. describe social support as a process of resource exchange between individuals, giving rise to the notion that social support is reciprocal in nature [56]. Indeed, social support is a complex concept and researchers have put forth various taxonomies [[56], [57], [58]] to categorize it, including the classic four-dimensional framework [58]; namely informational, emotional, instrumental, and appraisal support. Notwithstanding the diversity of taxonomies, more recent studies classified different social support constructs into two main types [[10], [59], [60]]. Specifically, informational support comes in the form of the transmission of information during a time of stress while emotional support in the form of provisioning caring, concern, empathy, love, and trust [59]).
In social media, social support theory is also a popular theoretical framework for understanding the use of impact of the online community on individuals [[61], [62], [63]]. One of the most noticeable practices is social support reinforces two-way interactions in the online community, as the public perceives supportive resources by collectively interacting with the posts through embedded functions, such as like, share, and comment [63]. This further motivates their engagement in resource exchange in social media, contributing to the decision-making process for fast-evolving situations [64]. Particularly, the influence of social support in social media can be understood through two influence mechanisms, namely emotional and informational support [15,65,66]. While the compelling advantages of social media in delivering timely information update to a wider audience is extensively acknowledged in literature [4] [5], recants studies have shifted the focus to social media's role in providing emotional support. For instance , it is posited that social media can create a mutual aid environment, providing an outlet for social interaction and voicing fear, and offering a voluntary reciprocal exchange of resources and services for mutual benefit [12]. Similar findings are evidenced in the [63] that online communities via social media can construct different social relationships, by which to exchange emotional support.
Particularly in times of crisis, the need for informational and emotional support is highlighted in the literature. For instance, it is repeatedly identified that fast information and warning delivery helps the public prepare for potential danger and negative effects, which is crucial in the public's response to the crisis [5,12,46]. On the other hand, recent COVID-19 pandemic related studies have stressed the importance to address the prevailing mental health issues among the general public and suggested that social support in social media may provide a potential solution to address when professional treatment is not readily available to the massive public [10,43,[67], [68], [69]].
In line with [10], we argued that facilitated by the interactive function of social media, public communication can exert a positive impact on the massive public through social support provision. Indeed, evidence from China during the COVID-19 pandemic demonstrated that social media postings did play a crucial role in establishing the truth (e.g., sharing pandemic information, refuting rumors, fostering public self-protect measures) [70] and boosting the social morale (e.g., encouraging messages, voices from top scientists and the COVID-19 fighters) [10]. Nevertheless, several research gaps remain unaddressed. First, while the impacts of both emotional and informational support have been well studied independently, they have not been compared. Second, much of the study treats the effects of social support as static, omitting to look at their dynamic effects at various stages. Last but not least, despite all the attempts, there is a dearth of empirical evidence on how government can leverage social support for better public communication.
3 Methodology
3.1 Research setting
To investigate the dynamic of informational support and emotional support strategies, we phased the timeliness of the pandemic into five stages according to the White Paper released by China's State Council Information Office [71] as depicted in Appendix 1. In addition, according to the National Health Commission, after nearly two months of no new local Covid-19 transmissions, Beijing reported 79 fresh cases since June 12, 2020, where the public fell into fear of the second wave in Beijing after the Xinfadi market outbreak [72]. Therefore, we further separate Stage 5 into Stage 5a and Stage 5b (Fig. 1 ).Fig. 1 Division of pandemic stage.
Fig. 1
3.2 Data
We took extra efforts in mitigating the data collection challenges in social science research [73] by collecting time-series Weibo data (including id, post time, post content, like, share, comment, etc.) and pandemic data that both cover all five pandemic stages. First, applying the inter-rater policy, each author had a preliminary search independently on the scope of Weibo data. We then reached the conclusion that 23 out of the 34 provincial administrative units in China so far have operated social media accounts (by Information Officer) at Weibo, the most influential social media in China with 550 million monthly active users [74], and 17 provincial administrative units (Table 1 ) released posts that covered all above five stages. For selected provincial administrative units’ Weibo accounts, Sina Weibo Application Processing Interface (API) is employed to collect relevant such as daily activity (e.g., daily information released frequency, the content of each post) and its corresponding feedback (like share, and comment) from the public. In addition, we used daily pandemic data from the health commission office as a proxy for the severity of the pandemic (provincial level). The pandemic data includes both newly and accumulated cases (confirmed, suspicious, cured, and dead) data.Table 1 The sample scope of this study.
Table 1Total Numbers Sample Scope * （17） Out-of-Sample Scope（17）
Provincial Administrative Unit Beijing, Tianjin, Shanghai, Chongqing, Henan, Hubei, Jiangsu, Jiangxi, Jilin, Heilongjiang, Shanxi, Shandong, Qinghai, Guangdong, Guizhou, Zhejiang, Xinjiang Hebeia, Hunana, Liaoning, Shaanxi, Anhui, Hainan, Fujian, Taiwan, Gansu, Yunnan, Sichuana, Tibet, Ningxia, Guangxi, Inner Mongolia, Hong Kong, Macau
Note: * The social media account selected for analysis in this study are Weibo accounts that are officially operated by the Information Office of the government of each provincial administrative unit respectively.
a Hebei, Hunan, and Sichuan do have official Weibo accounts, but neither of them posts information that covers the entire epidemic period, and therefore not included in this study.
3.3 Variables
The definition and description of all the variables are depicted in Appendix 3.
3.3.1 Dependent variables
The public engagement level in social media is set as the dependent variable. We deduced the measuring metric from [75,76] that daily public engagement is computed as a function of the feedback from the public (like share, and comment) and the characteristics of the e-government platform (number of followers, number of daily posts). In particular, the public engagement level Engagementi,t of all posts in province i at date t is computed as the equation below:Engagementi,t=Likesi,t+Sharesi,t+Commentsi,tFollowersi×Posti,t
Where Likesi,t, Sharesi,t, Commentsi,t, Posti,t denotes the total number of likes, shares, comments, and posts in province i at date t. Followersi represents the total number of followers of province i. It should be noted that consistent with previous work [[75], [76]] ( we treat Followersi as a time-independent variable, since it is assumed that changes in followers are relatively small compared to changes in other variables.
3.3.2 Independent variables
We considered daily public communication frequency, informational support, and emotional support strength as three independent variables for this study. Specifically, the daily public communication frequency is computed as the number of posts released from a particular social media account on a particular date. The variable is introduced to verify whether the frequency of public communication activity (e.g., too many, or too less) would influence public engagement.
To quantify the informational and support strength of the post, a supervised machine learning approach is adopted. The complex post content was first tokenized into simple unit tokens. A Part of Speech (POS) tagger is then applied to identify the parts of speech (e.g., noun, verb, adjective, adverb, preposition, and conjunction) of each token. For the informational support strength study, we only kept nouns and verbs, however for the emotional support analysis, we kept nouns, verbs, adjectives, and adverbs. The frequency of each token is then determined using the Frequency-Inverse Document Frequency (TF-IDF) vectorizer. To save effort, only tokens with a frequency of more than 50 times (7123 tokens) are chosen for the manual annotation.
In the annotation process, three domain experts are asked to annotate the score for emotional support and informational support independently using a 5 points Likert scale measurement (1: not at all, 2: slightly, 3: somewhat, 4: very, 5: extremely). The strength is computed as the mean score from the three experts. To ensure internal consistency, the three experts are asked to reach a consensus on the annotation standard before labeling. The internal consistency test achieved a Cronbach alpha value of 0.93, indicating excellent internal consistency. The strength for emotional support and informational support are respectively computed as the maximum emotional support and informational support strength of all tokens in a post.
Using a training dataset of 3104 posts, we then trained the model using Naive Bayes classification in “SnowNLP”, a popular NLP toolkit [77]. The whole set of test data (61,297 posts) is then applied to the model. It is noted that all the informational support strength and emotional support strength are normalized to [0,1], where 0 indicates no support and 1 is extremely strong support. Sample texts in the data set are depicted in appendix 2.
3.3.3 Control variables
We controlled for a set of other factors that could potentially influence the level of public engagement in social media amid a crisis including provincial characteristics and pandemic development tally. In addition to Gross domestic product (GDP), and population (POP), we also controlled for the number of 3A hospitals (Hospital), distance to Wuhan, adjacency with Wuhan, number of followers in the Weibo account, etc. It should be noted that we adopted the number of 3A hospitals as the control variable because it is argued that the number of 3A hospitals mirrors a province's medical capability. In addition, among all the pandemic tally, we adopted accumulative confirmed cases, accumulative cured cases, newly confirmed cases, and newly cured cases as control variables.
3.3.4 Model specification
We developed the research model by drawing on the Social Reciprocity Theory (SRT). SRT suggests that positive reciprocity occurs when an action committed by one individual that has a positive effect on someone else is returned with an action that has an approximately equal positive effect [78]. When social media is deemed as a social community, members consider exchange behaviors based on positive interactions, which impacts their intention to engage in the social community in the future [79]. In social media, there are two primary participants – the sender (post agencies) and the receiver (the public); reciprocity will be established by positive interactions when agencies provide social support through releasing information and the public engages in the information and provides feedback through functions such as like, share and comment [63]. It is thus understandable that the more social support offered, the higher the public engagement would be on social media. We conjectured accordingly that (1) social support provided by government agencies in social media has impacts on public engagement, and (2) the effect of social support strategies on public engagement may vary across different pandemic stages.
3.3.5 Analysis procedure
Our main interest lies in how social support strategies, particularly informational support and emotional support as proposed by [[15], [59], [63]], may influence public engagement in social media-based public communication. The analysis procedure is described as follows. First, the normality for all the variables is checked using the Q-Q plot in R. For those variables (engagement, emotional support, information support, accumulated confirmed cases, accumulated cured cases, newly confirmed cases, newly cured cases) that do not follow the normal distribution, a log-transformation is applied. Then the multicollinearity of all variables is checked using the variance inflation factor (VIF) value. The VIF values for all variables are less than 5, suggesting moderate multicollinearity problems among these variables are not likely to exist [80]. Besides, the descriptive statistic and the correlation matrix of all the variables are displayed in Appendix 4 and Appendix 5, respectively. After validating the assumption testing for the model, we conducted ordinary least squares (OLS) regression analysis on the whole dataset and at different stages, respectively.
4 Results
We analyzed the Weibo based social media data in regard to public communication amidst the pandemic from all accounts of 17 provincial governments in China and several illuminating findings arise.
4.1 Descriptive statistics results
In this section, the quantity of the post, the strength of both social support, and the corresponding public engagement (e.g., Likes, Shares, Comments) outcome across different stages of the pandemic will be presented.
The number of posts or government engagement on social media, according to [30], reflects a government's attentiveness toward the COVID-19 pandemic. The accumulated number of post across different stages are depicted in Fig. 2 . In general, two findings, in particular, are noteworthy. First of all, as the pandemic has progressed, the overall tendency for government activity shows an upward trend, indicating an intention on the part of the government to improve public communication, at least in terms of quantity, on social media. Particularly, there are more than three times as many posts in Stage 5B (16,411) than there are in Stage 2. (5389). Second, the overall tendency of the number of posts is fluctuating rather than constantly increase, suggesting the government's response (in terms of public communication activity) to the pandemic is dynamic rather than static. Particularly, the uncertainty in the COVID-19 pandemic and restrictive measures (e.g., social distancing, city lockdown) may have boosted the need for government to communicate with the public in Stage 3, while Stage 4 may have seen a decrease in the number of posts as the need for communication with the public decreases.Fig. 2 The accumulated public communication counts at different stages.
Fig. 2
Fig. 3 shows the mean and standard deviation of social support, including both informational and emotional support, as determined by the NLP-based content analysis. Surprisingly, compared to the aforementioned post quantity, the average social support strength encompassed in public communication is also fluctuating, however, demonstrated a different hump-shaped pattern from Stage 2 to Stage 5B. This implies that the responsibility of public communication may extend beyond simple posting to include more intricate social support provision functions, which confirms the necessity of the present study. Further, while the strength of both informational and emotional support peaks at mid-stages (Stage 3 and Stage 3, respectively), there are not identical. Particularly, the overall strength of emotional support is weaker than that of informational support, and it peaks later. All of these findings imply that public communication should not be conducted without strategies and that revisiting its outcomes is necessary in order to provide nuanced insights into its impacts.Fig. 3 The strength of informational and emotional support at different stages.
Fig. 3
In this study, we used public engagement level (measured by a function of shares, likes, and comments) (as depicted in Table 2 ) to proxy the outcome of public communication. Two findings are worth noting. First, compared to the average number of followers (272 thousand Appendix 4), the average daily Likes, Shares, and Comments start at a low level. In addition, the large standard deviation implies that the distribution of Likes, Shares, and Comments is highly dispersed. This indicates that the strategy for enhanced public engagement in social media is shy of systematic. Second, the distribution of likes, shares, and comments (Table 2) has a similar hump-shaped pattern to the strength of social support (Fig. 3), raising the possibility that public engagement and social support from the government are related. However, the relationship between social support and public engagement has not yet been established, necessitating additional research on the interactions between the public and the government in order to offer more insightful conclusions.Table 2 Descriptive statistics of likes, shares, and comments across stages.
Table 2Stages Likes Shares Comments
Mean S.D Mean S.D Mean S.D
STAGE-2 70.76 565.13 4.33 15.97 8.39 32.98
STAGE-3 371.47 2874.80 60.67 263.28 32.32 282.63
STAGE-4 89.70 724.76 20.14 104.09 12.54 51.95
STAGE-5A 49.20 275.35 13.46 65.86 9.86 45.61
STAGE-5B 57.86 2727.58 7.57 159.38 7.72 52.09
4.2 Regression analysis results
Fig. 4 summarizes the regression analysis result. Both social supports (emotional support and information support) are testified to positively and significantly with the public engagement level in the full model, however, demonstrate a subtle difference in the staged model. Regarding the effect of emotional support, the significant correlation is only identified in later stages (Stage 4, Stage 5A, and Stage 5B), but not in earlier stages (Stage 2, Stage 4). Specifically, the effect reaches a peak at Stage 4 (β=0.577) when phased success in controlling Covid-19 is witnessed, however, drops at stage 5A (β=−0.268), when a new wave of the Pandemic strikes, and then bounces back at Stage 5B, when the pandemic is properly handled (β=0.794). Regarding the effect of informational support, in contrast to the effect of emotional support, the significant correlation is only witnessed in earlier stages (Stage 2, Stage 3, Stage 4), not in the later stages (Stage 5A, Stage 5B). Particularly, the trend of the effect is on the decline from Stage 2 (0.709) when mitigation and Containment of COVID-19 are observed to Stage 4 (0.327) when phased success in controlling Covid-19 is witnessed. It is also intriguing to point out that the effect of both support is not entirely positive. For instance, the effect of emotional support at Stage 5A is significant and negative (β=−0.268). This implies that the impact mechanism of social support is more complex than previously thought.Fig. 4 Dynamic impacts of informational support and emotional support.
Fig. 4
Regarding the effect of public communication frequency, the negative effect is confirmed in both the full model (β=−0.008). Particularly the effect of public communication frequency on public engagement is consistently negative and significant in all staged models, except for stage 5B. This is in line with the concern over information overload or “infodemic” that has been widely seen on social media amidst the pandemic [18,21], which could eventually stifle communication between the government and the public. Surprisingly, the number of followers is found a minor predictor of public engagement. In all staged models except Stage 3 (Daily tally dropped to a single digit), the effect of the follower on public engagement is insignificant.
In terms of control variables, all provincial strength characteristics (e.g., GDP, EGDI, followers, 3A hospitals, etc.) are found significantly related to public engagement in the full model, however, the effect is comparably small compared to the effect of emotional support, informational support, or public communication frequency. This gives rise to the thought that provincial strength may not be directly linked to the public communication between the government and the public, reinforcing the notion that social support is complex and dynamic. Regarding the control variables for the daily pandemic, it is identified that while the daily pandemic tally is generally significantly related to public engagement in the full model, but not necessarily significant in the staged model. This means that public engagement may not be strongly linked to the development of the pandemic. Besides, the adjacency of the province to Wuhan is found an insignificant predictor of public engagement.
5 Discussion
Our primary goal is to analyze the impact mechanisms of public communication. The findings confirm the significant correlation between both social support and public engagement, suggesting that the social support theory can be a well-founded framework to explain the impact of public communication. Our findings also suggest that the effect of public communication (both emotional support and informational support) are dynamically evolving rather than static during a crisis. This means that public communication is less likely to be a “one-size-fits-all” government-oriented process, but rather should be handled with strategic adjustments.
5.1 Impacts of emotional support cannot be underestimated
While most of the extant studies emphasized the importance of public communication on information dissemination and exchange [6,7], drawing upon social support theory, we argued that the power of public communication should go beyond informational support, but encompass emotional support that is equally if not more important. Specifically, the value of social support theory in dissecting the effects of public communication has been demonstrated by the different impact patterns of the two dimensions of support. First, the two aspects of social support have distinctive coefficients and levels of significance. Specifically, the emotional support coefficient has shown fluctuation, but the informational support coefficient is monotonously declining. This implies that the impact mechanism of emotional support and informational support on public engagement may be inconsistent. In light of this, the social support theory offers a theoretical framework to treat emotional support as a separate dimension of information support, and in turn, make it easier to comprehend how public communication exerts positive psychological or emotional influence on the public [15,81].
Second, as discussed earlier, the outcome of public communication is prone to be stage-based because the demand for informational and emotional support is likely to be stage-depend across different stages of a crisis. Further, we can identify the distinct demand pattern for public communication across stages with the aid of our incorporation of social support theory. Based on the significant level (Fig. 4), the demand for informational support is substantial in the early stages (Stage 2, Stage 3, Stage 4) whereas the demand for emotional support may be lagged (Stage 4, Stage 5A, Stage 5b) but cannot be overlooked. Information-focused research can therefore undervalue or underestimate the necessity of offering emotional support through public communication. In sum, it is evidenced that emotional support differs from informational support regarding support strength, and stage-based variations. Our incorporation of social support theory provides a more comprehensive understanding of both the influence of public communication and the dynamic demand of the public across different stages. Underestimating the influence of emotional support would otherwise result in an incomplete perception of the impact of public communication, and further, restrain the rationality and effectiveness of public communication strategies.
5.2 The stage-based pattern of social support influence
Our findings show that public communication outcome (as benchmarked by public engagement in this study) is closely related to the strength of staged-dependent social support rather than being anticipated by a province's strength (e.g., GDP, Pop, EDGI). This means, in order to achieve substantial rather than symbolic public communication, government agencies may need to play a more active role in communication by tailoring their communication strategies to the public's staged-dependent demand in crisis [30], rather than treating it solely as the government-led process of information dissemination. Indeed, since the Government Performance and Results Act of 1993, outcome-based performance evaluations are made formal for measuring the service provision by the governments [82]. Particularly for public communication, numerous scholars are investigating the metrics to quantify the performance from public feedback such as public satisfaction [83], public engagement [75,76], etc.
Despite the differences, these works are all built upon an underlying assumption that there is a major causality between public communication and the positivity of public activities: the better the public communication becomes, the more positive the public's activities would be. This is justifiable in a static situation where the public's demand for the communication service remains almost the same. However, in a real-life setting, which is dynamic and fast-evolving, such as Covid-19, the rapid change of public demand in public communication may alter their evaluation because social support may fail to satisfy the changing demand. Given that the expectations of informational demands and emotional demands from the public are evolving across stages [84,85], leading to the possible disconfirmation between the demands of the public and support provision from the government. For instance, in the early stages (Stage 2 and Stage 3) when the pandemic unsettled the public, the informational support provided through Weibo posts greatly addressed the public's concern for situational awareness, resulting in a high significance of the correlation between informational support and public engagement. When the pandemic was taken under control (Stage 4, Stage 5A, Stage 5B), the demand for informational support dropped, resulting in a low significance level. In contrast, the full model, which treats the entire stages as a whole (Table 3 ), failed to identify such insights at the granular level. As a result, the impact of public communication on public engagement amid a crisis should be better understood and further assessed by identifying the pandemic stages and evaluating them correspondingly.Table 3 Ordinary least squares regression results.
Table 3Dependent variable: log (Engagement + 1)
OLS panel linear
Full Model Stage 2 Stage 3 Stage 4 Stage 5A Stage 5B
log (Emo_Sup + 1) 0.499*** −0.043 0.230 0.577*** −0.268** 0.794*
（0.079） (0.263) (0.181) (0.086) (0.120) (0.426)
log (Info_Sup + 1) 0.376*** 0.709*** 0.356* 0.327*** 0.021 −0.709
(0.076) (0.222) (0.183) (0.097) (0.096) (0.483)
Freq −0.008*** −0.014*** −0.015*** −0.005*** −0.003* −0.002
(0.001) (0.004) (0.003) (0.002) (0.002) (0.005)
Followers −0.0004*** −0.0005 −0.001*** −0.0003 0.0005 0.0002
(0.0001) (0.001) (0.0003) (0.0003) (0.001) (0.001)
Adjacency 0.027 0.14 0.116 −0.052 −0.262 −0.066
(0.020) (0.291) (0.152) (0.115) (0.209) (0.257)
Distance −0.0001***,. −0.0003 −0.0001 0.00003 −0.0001 −0.0002
(0.000) (0.000) (0.000) (0.000) (0.000) (0.000)
GDP −0.00001*** −0.00002 −0.00002*** −0.00001** −0.00001 −0.00001*
(0.00000) (0.00001) (0.00001) (0.00000) (0.00001) (0.00001)
Pop 0.0001*** 0.0002 0.0002*** 0.00004 0.0001 0.0002*
(0.00001) (0.00002) (0.0001) (0.00005) (0.0001) (0.0001)
EGDI 0.018*** 0.035 0.046*** 0.010 0.002 0.017
(0.001) (0.022) (0.009) (0.007) (0.013) (0.014)
Hospital −0.003** −0.017 −0.013* 0.003 0.008 −0.008
(0.001) (0.017) (0.007) (0.005) (0.009) (0.011)
log (Conf_Acu + 1) 0.064*** 0.158*** −0.087 −0.079 0.388* 0.389
(0.020) (0.049) (0.130) (0.113) (0.223) (0.992)
log (Cure_Acu + 1) −0.141*** −0.226*** −0.039 0.021 −0.584*** −0.488
(0.017) (0.056) (0.088) (0.126) (0.217) (0.982)
log (Conf_delta +1) −0.007 −0.082* 0.088*** 0.078*** 0.070*** −0.022
(0.012) (0.042) (0.029) (0.014) (0.027) (0.108)
log (Cure_delta +1) 0.018** 0.074 0.021 −0.008 0.047** 0.012
(0.008) (0.053) (0.023) (0.013) (0.021) (0.078)
Constant −0.091 −0.627 −1.089 0.064 1.595 0.216
(0.094) (1.380) (0.683) (0.55) (1.001) (1.180)
Observations 2939 535 445 696 1049 214
R2 0.314 0.154 0.173 0.191 0.045 0.063
Adjusted R2 0.311 0.131 0.146 0.174 0.032 −0.003
Residual Std. Error 0.468 (df = 2924)
F Statistic 95.669*** (df = 14; 2924) 94.296*** 89.860*** 160.346*** 48.612*** 13.443
Note: *p < 0.1; **p < 0.05; ***p < 0.01.
5.3 Information fatigue inhibits the impact of social support
Infodemic, a term used frequently in relation to social media during COVID-19, refers to the experience of information fatigue brought on by exposure to excessive amounts of information [18]. In this study, we introduced daily public communication frequency as an independent variable to see if such phenomena might be present in public communication. The negative and significant association between public communication frequency and public engagement confirms that too much public information may exert a negative effect on public communication outcomes. In line with other works [9,18], this research confirms that the overwhelming volume of posts could dampen the intended social support. According to the full model in Table 3, the daily public communication frequency is reported to have a significant and negative (β=−0.008,p<0.05) impact on public engagement. This means excessive government communication may not only fail to bring about good outcomes but also cause information fatigue among the public. This phenomenon is echoed by the observation in [86] that excessive use of public communication via social media may backfire as it can cause information overload or over-thinking amongst individuals, negating public motivation to positively engage in public communication or even crisis response.
Additionally, it is also necessary to provide social support that caters to the demand of the public at different stages to prevent the detrimental effects of infodemic on public communication [16]. Indeed, the public's demand for both emotional support and information could vary along with the development of the pandemic. If the type or the amount of social support encompassed in public communication does not adjust accordingly, the excessive provision of social support may overwhelm the public's demand, leading to the feeling of exhaustion and lower levels of engagement level [87]. For instance, the expectation for informational support may drop when the pandemic was progressively brought under control and the situation grew less worrisome and unclear. The effect may diminish as seen from Stage 2 to Stage 5A if the amount of information support remains constant (Table 3). In other words, the change in stage-based need for emotional and informational support may have also resulted in an abundance of social support that creates information fatigue and further reduces the impact of information support.
5.4 Implications
Theoretically, drawing upon the social support theory, we proposed a prototype attempt to comprehend the impact of social media based public communication amid a crisis. Regarding public communication amid a crisis, the incorporation of social support theory provides nuanced insights into how emotional support encompassed in public communication can exert a positive influence on the public. Besides, we additionally introduced the variable daily public communication frequency to conceptualize the commonly related “infodemic” phenomena in COVID-19-related literature and unravel its potential impact. Regarding social support theory, this study enhances its context by expanding it to public communication, urging further research to better understand the interaction between the public and the government, which would, in turn, support the theory's development and empirical examination.
Practically, the findings of this study can be extended to developing better public communication strategies amid a crisis. Through identifying the distinctive impact patterns of the two supports, we highlight that while informational support is crucial in the earlier stages of a crisis, emotional support could be of great help in the later stages of the crisis to comfort the emotions of the public. This means that when developing tactics for public communication amid a crisis, the pivot role of emotional support cannot be overlooked. Second, our analysis demonstrates that the influence of social support on public communication in times of crisis varies depending on the stage. Governments must, therefore, adapt and tailor their public communication strategy as the crisis develops rather than creating a strategy that is “one-size-fits-all”. Finally, we empirically validated that the “infodemic” could also occur in public communication amid a crisis. We, therefore, argued that improving the quantity of public communication would not help promote the public communication outcome. Rather, public communication that accounts for the public's demand and tailors to the evolving of the crisis is more effective.
6 Conclusion
The present studies investigate the impact of public communication amid crisis through the theoretical lens of social support theory. Particularly, we utilized public engagement as a proxy for the outcome of public communication amid the crisis and the dynamic impacts of two facets of social support are examined. Using the 17 Chinese provincial government-owned social media (Weibo) accounts, the separate impact of emotional support and informational support on public engagement is examined Based on the findings, this study recommends that government organizations take into account emotional support as a strategy for public communication, dynamically adapt their strategies to the people's demands as the crisis develops, and exercise prudence when it comes to the “infodemic” phenomenon. Despite being undertaken in a COVID-19 pandemic context, it is argued that all these findings are focused on public communication strategies, which can extend beyond the scope of the pandemic to general crises. The results of this study are preliminary overall, but they can be used as a starting point and encourage more research into the role that emotional support plays in public communication.
The present study is not without limitations. We adopted the public engagement metrics from existing literature [75] to benchmark the outcome of public communication, which is a function of share, like, and comment. However, it should be also noted that the three facets (like, share, and comment) of engagement may reflect different attitudes, which might be a consequence of different influences. For instance, a “like” in the communication post may imply that the influence of the post on the citizens is positive, but does not necessarily enact citizens to actively disaster preparation. On the other hand, a “comment” in the post reflects citizens' active involvement in the response, however, it may be motivated by negative rather than positive affections. Thus, it is necessary to separately investigate the mechanisms of how social support encompassed in public communication can lead to three engagement activities and how three engagement activities connect to the public's better physical and mental readiness for disasters.
Two potential changes to the research methodology should be mentioned. First, it takes a lot of time to annotate the work done by the domain experts for the supervised training model that we used to extract both social support strengths. Future research on creating a more practical strategy to address timeliness during a crisis is acknowledged as being necessary. Second, we include daily public communication frequency as an independent variable in the regression model to confirm the impact of social support overload. However, the post frequency may have a moderation effect on emotional support and informational support, which needs to be addressed in our future research.
Regarding the data, this study used 17 Chinese provincial government-owned social media as the study cope and covered the period from the initial outbreak (Dec 2019) to the successful control of Covid-19 (Jul 2020) in China. However, However, Covid-19 is a global crisis rather than a regional incidence, and a dataset covering a longer duration would be favorable.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Appendix 1 Phased China's COVID-19 action timeline (SCIO, 2020)
Stage Description Duration
Stage 1 Initial Response to Covid-19 December 27, 2019–January 19, 2020
Stage 2 Mitigation and Containment of Covid-19 January 20, 2020–February 20, 2020
Stage 3 The daily tally dropped to single digit February 21, 2020–March 17, 2020
Stage 4 Phased success in controlling Covid-19 March 18, 2020–April 28, 2020
Stage 5 Ongoing prevention and control April 29, 2020 onwards
Appendix 2 Sample texts in the data set
ID Sample Content Sample Content (in English) Emotional Support Strength Information Support Strength
1 截至1月29日24时，国家卫生健康委收到31个省（自治区、直辖市）和新疆生产建设兵团累计报告确诊病例7711例，现有重症病例 … As of 24:00 on January 29, the National Health Commission has received a total of 7711 confirmed cases from 31 provinces (autonomous regions and municipalities) and the Xinjiang Production and Construction Corps: severe cases … 0.12 0.89
2 … 0-6岁儿童日常如何做好新型冠状病毒的预防？外出时可采取哪些预防措施？当孩子的照护者出现可疑症状时有哪些建议？孩子生病时又该如何应对？来看中国疾控中心的一图解读。详见↓ #上海战疫##上海加油# 0-6岁儿童如何预防新型冠状病毒？一图解读 … How do children aged 0–6 prevent the new coronavirus? What precautions can be taken when going out? What advice do you have when your child's caregiver has suspicious symptoms? What should I do when my child is sick? Take a look at a picture interpretation of the China Centers for Disease Control and Prevention. For details, see ↓ #Fight！Shanghai# #Coming Shanghai# … 0.47 0.71
3 … 近日，湖南疫情防控一线再传好消息。截至2月6日16时，湖南已有75例新型冠状病毒感染的肺炎患者治愈出院。走出隔离医院，他们会说什么？ … Recently, good news has spread on the front line of Hunan epidemic prevention and control. As of 16:00 on February 6, 75 cases of pneumonia patients infected by the new coronavirus in Hunan have been cured and discharged. What would they say when they walked out of the isolation hospital? 0.74 0.74
4 【为奋战在“战疫”一线的白衣天使而歌】抗疫歌曲⟪托起生命的风采⟫致敬逆行者，呼唤众志成城！加油中国！加油武汉！ [Eulogy for the angels in white who are fighting on the front line of the epidemic] The anti-epidemic song “The Demeanor of Life” pays tribute to retrogrades and calls for unity! Come on China! Come on Wuhan! 0.91 0.09
Appendix 3 Definition and Description of Variables
Variable Description
Dependent Variable
Engagement Computed as a function of (share, like, comment), denoted the daily engagement level
Independent Variable
Emo_Sup Emotional Support Strength
Info_Sup Information Support Strength
Freq Daily public communication frequency
Control for Provincial Characteristics
Followers Number of followers (in thousands)
Adjacency The adjacency of the province to the pandemic center
Distance The travel distance between the province and the pandemic center (in km)
GDP GDP of the province (in billion yuan)
Pop The population of the province (in thousands)
EGDI The e-government development index of the province developed by the National School of Administration
Hospital No of 3A hospitals in the province an indicator for benchmarking the medical care level
Control for Pandemic Development
Conf_Accu Daily accumulative confirmed cases of the province
Cure_ Accu Daily accumulative Cured cases of the province
Conf_Newly Daily newly confirmed cases of the province
Cure_Newly Daily newly Cured cases of the province
Appendix 4 Descriptive Statistics
Variable Mean S.D. Min Max
log (Engagement + 1) (1) 0.37 0.56 0 5.38
log (Emo_Sup + 1) (2) 0.18 0.14 0 0.69
log (Info_Sup + 1) (3) 0.28 0.16 0 0.69
Freq (4) 16.99 12.83 1 68
Followers (5) 274.27 255.48 10.31 933.2
Distance (6) 1130.7 793.67 0 3268
GDP (7) 36542.45 28963.69 2966 107,671
Pop(8) 4565.05 2849.83 608 11,521
EGDI (9) 64.71 12.73 41.35 94.88
Hospital (10) 40.22 19.94 9 102
log (Conf_Acu + 1) (11) 5.89 1.94 0 11.13
log (Cure_Acu + 1) 5.45 2.26 0 11.07
log (Conf_delta +1) 0.66 1.3 0 9.61
log (Cure_delta +1) (14) 0.82 1.39 0 8.01
Note: To avoid multi-collinearity and skewness, engagement, emotional support, information support, accumulated confirmed cases, accumulated cured cases, newly confirmed cases, and newly cured cases are log-transformed.
Appendix 5 Correlation Matrix of the variables
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14)
log (Engagement + 1) (1) 1
log (Emo_Sup + 1) (2) 0.27 1
log (Info_Sup + 1) (3) 0.34 0.52 1
Freq (4) −0.15 −0.02 0.02 1
Followers (5) −0.16 0.03 0.13 0.25 1
Distance (6) −0.03 −0.12 −0.11 0.06 0.07 1
GDP (7) −0.13 0.1 0.08 −0.19 0.2 −0.47 1
Pop(8) −0.02 0.13 0.06 −0.3 −0.12 −0.43 0.87 1
EGDI (9) 0.02 0.15 0.31 0 0.47 −0.45 0.67 0.5 1
Hospital (10) −0.02 0.06 0.13 −0.25 0.1 −0.35 0.77 0.81 0.68 1
log (Conf_Acu + 1) (11) −0.14 0.15 0.1 −0.06 0.06 −0.53 0.41 0.44 0.52 0.44 1
log (Cure_Acu + 1) (12) −0.26 0.01 −0.12 −0.09 0.03 −0.45 0.33 0.36 0.41 0.36 0.92 1
log (Conf_delta +1) (13) 0.23 0.3 0.47 0.07 0.14 −0.15 0.16 0.13 0.25 0.21 0.14 −0.18 1
log (Cure_delta +1) (14) 0.09 0.4 0.36 0.04 0.12 −0.24 0.2 0.2 0.27 0.26 0.41 0.26 0.47 1
Data availability
Data will be made available on request.
Acknowledgement
This study is supported by the 10.13039/501100001809 National Natural Science Foundation of China (71904020), 10.13039/501100002369 Chongqing University (2019GGXY04), the 10.13039/501100012226 Fundamental Research Funds for the Central Universities (2021CDJSKJC03).
==== Refs
References
1 Reuter C. Kaufhold M.A. Fifteen years of social media in emergencies: a retrospective review and future directions for crisis informatics J. Contingencies Crisis Manag. 26 1 2018 41 57
2 Palen L. Vieweg S. Liu S.B. Hughes A.L. Crisis in a networked world: Features of computer-mediated communication in the April 16, 2007, Virginia Tech event Soc. Sci. Comput. Rev. 27 4 2009 467 480
3 Tagliacozzo S. Magni M. Government to Citizens (G2C) communication and use of social media in the post-disaster reconstruction phase Environ. Hazards 17 1 2018 1 20
4 Al-Saggaf Y. Simmons P. Social media in Saudi Arabia: Exploring its use during two natural disasters Technol. Forecast. Soc. Change 95 2015 3 15
5 Dabner N. ‘Breaking Ground’in the use of social media: A case study of a university earthquake response to inform educational design with Facebook Internet High Educ. 15 1 2012 69 78
6 O'Brien M. Moore K. McNicholas F. Social media spread during Covid-19: the pros and cons of likes and shares Ir. Med. J. 113 4 2020 52 32268046
7 Zhang C. Fan C. Yao W. Hu X. Mostafavi A. Social media for intelligent public information and warning in disasters: An interdisciplinary review Int. J. Inf. Manag. 49 2019 190 207
8 Chen E. Lerman K. Ferrara E. Tracking social media discourse about the covid-19 pandemic: Development of a public coronavirus twitter data set JMIR Public Health and Surveillance 6 2 2020 e19273
9 Islam A.N. Laato S. Talukder S. Sutinen E. Misinformation sharing and social media fatigue during COVID-19: An affordance and cognitive load perspective Technol. Forecast. Soc. Change 159 2020 120201
10 Hu X. Song Y. Zhu R. He S. Zhou B. Li X. Bao H. Shen S. Liu B. Understanding the impact of emotional support on mental health resilience of the community in the social media in Covid-19 pandemic J. Affect. Disord. 308 2022 360 368 35460730
11 Hu X. Zhang J. Shen S. Exploring the Pathway from Seeking to Sharing Social Support in E-Learning: an Investigation Based on the Norm of Reciprocity and Expectation Confirmation Theory 2022 Current Psychology Online
12 Marzouki Y. Aldossari F.S. Veltri G.A. Understanding the buffering effect of social media use on anxiety during the COVID-19 pandemic lockdown Humanities and Social Sciences Communications 8 1 2021 1 10
13 Jin Y. Making sense sensibly in crisis communication: How publics' crisis appraisals influence their negative emotions, coping strategy preferences, and crisis response acceptance Commun. Res. 37 4 2010 522 552
14 Langford C.P.H. Bowsher J. Maloney J.P. Lillis P.P. Social support: a conceptual analysis J. Adv. Nurs. 25 1 1997 95 100 9004016
15 Liang T.-P. Ho Y.-T. Li Y.-W. Turban E. What drives social commerce: The role of social support and relationship quality Int. J. Electron. Commer. 16 2 2011 69 90
16 McKinney V. Yoon K. Zahedi F.M. The measurement of web-customer satisfaction: An expectation and disconfirmation approach Inf. Syst. Res. 13 3 2002 296 315
17 Huang J. Liu F. Teng Z. Chen J. Zhao J. Wang X. Wu R. Care for the psychological status of frontline medical staff fighting against Coronavirus Disease 2019 (COVID-19) Clin. Infect. Dis. 71 12 2020 3268 3269 32246142
18 Zarocostas J. How to fight an infodemic Lancet 395 10225 2020 676 32113495
19 Bucher E. Fieseler C. Suphan A. The stress potential of social media in the workplace Inf. Commun. Soc. 16 10 2013 1639 1667
20 Moore A. MacKenzie M.K. Policy Making during Crises: How Diversity and Disagreement Can Help Manage the Politics of Expert Advice 2020 371 Bmj
21 Erku D.A. Belachew S.A. Abrha S. Sinnollareddy M. Thomas J. Steadman K.J. Tesfaye W.H. When fear and misinformation go viral: Pharmacists' role in deterring medication misinformation during the'infodemic'surrounding COVID-19 Res. Soc. Adm. Pharm. 17 1 2021 1954 1963
22 Coombs W.T. Protecting organization reputations during a crisis: The development and application of situational crisis communication theory Corp. Reput. Rev. 10 3 2007 163 176
23 Coombs W.T. Crisis communication and its allied fields The Handbook of Crisis Communication 2010 54 64
24 Seeger M.W. Best practices in crisis communication: An expert panel process J. Appl. Commun. Res. 34 3 2006 232 244
25 Wirtz J.G. Zimbres T.M. A systematic analysis of research applying ‘principles of dialogic communication’to organizational websites, blogs, and social media: Implications for theory and practice J. Publ. Relat. Res. 30 1–2 2018 5 34
26 Perry R.W. Nigg J.M. Emergency management strategies for communicating hazard information Publ. Adm. Rev. 45 1985 72 77
27 Reynolds B. Seeger W. M. Crisis and emergency risk communication as an integrative model J. Health Commun. 10 1 2005 43 55 15764443
28 Sutton J. Spiro E.S. Johnson B. Fitzhugh S. Gibson B. Butts C.T. Warning tweets: Serial transmission of messages during the warning phase of a disaster event Inf. Commun. Soc. 17 6 2014 765 787
29 Torero M. Without Food, There Can Be No Exit from the Pandemic 2020 Nature Publishing Group London, UK
30 Zhu R. Song Y. He S. Hu X. Hu W. Liu B. Toward Dialogue through a Holistic Measuring Framework-The Impact of Social Media on Risk Communication in the COVID-19 2021 Information Technology & People
31 Covello V.T. Risk communication: An emerging area of health communication research Annals of the International Communication Association 15 1 1992 359 373
32 Lachlan K.A. Spence P.R. Lin X. Najarian K. Del Greco M. Social media and crisis management: CERC, search strategies, and Twitter content Comput. Hum. Behav. 54 2016 647 652
33 Reynolds B. Galdo J.H. Sokler L. Freimuth V.S. Crisis and emergency risk communication Center for Disease Control and Prevention 2014
34 Veil S. Reynolds B. Sellnow T.L. Seeger M.W. CERC as a theoretical framework for research and practice Health Promot. Pract. 9 4_suppl 2008 26S 34S 18936257
35 Andrews R. Entwistle T. Four faces of public service efficiency: What, how, when and for whom to produce Publ. Manag. Rev. 15 2 2013 246 264
36 Andrews R. Van de Walle S. New public management and citizens' perceptions of local service efficiency, responsiveness, equity and effectiveness Publ. Manag. Rev. 15 5 2013 762 783
37 Cheng Y. Yu J. Shen Y. Huang B. Coproducing responses to COVID-19 with community-based organizations: lessons from Zhejiang Province, China Publ. Adm. Rev. 80 5 2020 866 873
38 Schmelz K. Enforcement may crowd out voluntary support for COVID-19 policies, especially where trust in government is weak and in a liberal society Proc. Natl. Acad. Sci. USA 118 1 2021
39 Goldfinch S. Taplin R. Gauld R. Trust in government increased during the Covid-19 pandemic in Australia and New Zealand Aust. J. Publ. Adm. 80 1 2021 3 11
40 Rosenberg D. Ethnic perspective in e-government use and trust in government: A test of social inequality approaches New Media Soc. 23 6 2021 1660 1680
41 Bargain O. Aminjonov U. Trust and compliance to public health policies in times of COVID-19 J. Publ. Econ. 192 2020 104316
42 Ma L. Zhan M. Effects of attributed responsibility and response strategies on organizational reputation: A meta-analysis of situational crisis communication theory research J. Publ. Relat. Res. 28 2 2016 102 119
43 Liao Q. Yuan J. Dong M. Yang L. Fielding R. Lam W.W.T. Public engagement and government responsiveness in the communications About COVID-19 during the early epidemic stage in China: Infodemiology study on social media data J. Med. Internet Res. 22 5 2020 e18796
44 Mori E. Barabaschi B. Cantoni F. Virtuani R. Local governments' communication through Facebook. Evidences from COVID-19 pandemic in Italy J. Publ. Aff. 21 4 2021 e2551
45 Yates D. Paquette S. Emergency knowledge management and social media technologies: A case study of the 2010 Haitian earthquake Int. J. Inf. Manag. 31 1 2011 6 13
46 Hughes A.L. St Denis L.A. Palen L. Anderson K.M. Online public communications by police & fire services during the 2012 Hurricane Sandy Proceedings of the SIGCHI Conference on Human Factors in Computing Systems 2014 1505 1514
47 Li J. Xu Q. Cuomo R. Purushothaman V. Mackey T. Data mining and content Analysis of the Chinese social media platform Weibo during the early COVID-19 outbreak: retrospective observational Infoveillance study JMIR Public Health Surveill 6 2 2020 e18700
48 Li H. Communication for coproduction: Increasing information credibility to fight the coronavirus Am. Rev. Publ. Adm. 50 6–7 2020 692 697
49 Wu Y. Xiao H. Yang F. Government Information Disclosure and Citizen Coproduction during COVID-19 in China 2021 (Governance)
50 Panagiotopoulos P. Barnett J. Bigdeli A.Z. Sams S. Social media in emergency management: Twitter as a tool for communicating risks to the public Technol. Forecast. Soc. Change 111 2016 86 96
51 Bertot J.C. Jaeger P.T. Hansen D. The impact of polices on government social media usage: Issues, challenges, and recommendations Govern. Inf. Q. 29 1 2012 30 40
52 Brombal D. Moriggi A. Marcomini A. Evaluating public participation in Chinese EIA. An integrated public participation Index and its application to the case of the new Beijing airport Environ. Impact Assess. Rev. 62 2017 49 60
53 Oh H.J. Lauckner C. Boehmer J. Fewins-Bliss R. Li K. Facebooking for health: An examination into the solicitation and effects of health-related social support on social networking sites Comput. Hum. Behav. 29 5 2013 2072 2080
54 Wortman C.B. Dunkel-Schetter C. Baum A. Singer J.E. Conceptual and Methodological Issues in the Study of Social Support 1987 Lawrence Erlbaum Associates Hillsdale, NJ
55 Shumaker S.A. Brownell A. Toward a theory of social support: Closing conceptual gaps J. Soc. Issues 40 4 1984 11 36
56 Cohen S.E. Syme S. Social Support and Health 1985 Academic Press
57 Coulson N.S. Buchanan H. Aubeeluck A. Social support in cyberspace: a content analysis of communication within a Huntington's disease online support group Patient Educ. Counsel. 68 2 2007 173 178
58 House J.S. Work Stress and Social Support 1983 Addison-Wesley Series on Occupational Stress
59 Kort‐Butler L.A. Social Support Theory 2017 The Encyclopedia of Juvenile Delinquency and Justice 1 4
60 Yan L. Tan Y. Feeling blue? Go online: An empirical study of social support among patients Inf. Syst. Res. 25 4 2014 690 709
61 Reblin M. Uchino B.N. Social and emotional support and its implication for health Curr. Opin. Psychiatr. 21 2 2008 201
62 Stephens J.B. Berner M. Learning from your neighbor: The value of public participation evaluation for public policy dispute resolution Journal of Public Deliberation 7 1 2011
63 Yan L. Pedraza‐Martinez A. Social media for disaster management: Operational value of the social conversation Prod. Oper. Manag. 28 10 2019 2514 2532
64 Lovari A. Bowen S.A. Social media in disaster communication: A case study of strategies, barriers, and ethical implications J. Publ. Aff. 20 1 2019 9
65 Carr C.T. Wohn D.Y. Hayes R.A. As social support: Relational closeness, automaticity, and interpreting social support from paralinguistic digital affordances in social media Comput. Hum. Behav. 62 2016 385 393
66 Rains S.A. Keating D.M. The social dimension of blogging about health: Health blogging, social support, and well-being Commun. Monogr. 78 4 2011 511 534
67 Cohen S. McKay G. Social support, stress and the buffering hypothesis: A theoretical analysis Handbook of Psychology and Health ume IV 2020 Routledge 253 267
68 Li Y. Peng J. Does social support matter? The mediating links with coping strategy and anxiety among Chinese college students in a cross-sectional study of COVID-19 pandemic BMC Publ. Health 21 1 2021 1 10
69 Yan L. Pedraza‐Martinez A.J. Social media for disaster management: Operational value of the social conversation Prod. Oper. Manag. 28 10 2019 2514 2532
70 Shen Y. Cheng Y. Yu J. From recovery resilience to transformative resilience: How digital platforms reshape public service provision during and post COVID-19 Publ. Manag. Rev. 2022 1 24
71 SCIO Fighting COVID-19: China in Action 2020 China's State Council Information Office Beijing
72 Tan W. Niu P. Zhao X. Pan Y. Zhang Y. Chen L. Zhao L. Wang Y. Wang D. Han J. Reemergent cases of COVID-19-Xinfadi wholesales market, Beijing municipality, China, June 11, 2020 China CDC Weekly 2 27 2020 502 504 34594688
73 Starr M.K. Van Wassenhove L.N. Introduction to the special issue on humanitarian operations and crisis management Prod. Oper. Manag. 23 6 2014 925 937
74 WeiboCorporation Weibo Reports First Quarter 2020 Unaudited Financial Results 2020 Weibo Corporation
75 Agostino D. Arnaboldi M. A measurement framework for assessing the contribution of social media to public engagement: An empirical analysis on Facebook Publ. Manag. Rev. 18 9 2016 1289 1307
76 Bonsón E. Ratkai M. A set of metrics to assess stakeholder engagement and social legitimacy on a corporate Facebook page Online Inf. Rev. 37 5 2013 787 803
77 Wang R. isnowfy/snownlp - Python library for processing Chinese text Github: Github 2020
78 Falk A. Fischbacher U. A theory of reciprocity Game. Econ. Behav. 54 2 2006 293 315
79 Wu J.-J. Khan H.A. Chien S.-H. Lee Y.-P. Impact of emotional support, informational support, and norms of reciprocity on trust toward the medical aesthetic community: The moderating effect of core self-evaluations Interactive Journal of Medical Research 8 1 2019 e11750
80 Shrestha N. Detecting multicollinearity in regression analysis Am. J. Appl. Math. Stat. 8 2 2020 39 42
81 Ko H.-C. Wang L.-L. Xu Y.-T. Understanding the different types of social support offered by audience to A-list diary-like and informative bloggers Cyberpsychol., Behav. Soc. Netw. 16 3 2013 194 199 23363225
82 Heinrich C.J. Outcomes-based performance management in the public sector: implications for government accountability and effectiveness Publ. Adm. Rev. 62 6 2002 712 725
83 Ho A.T.K. Cho W. Government communication effectiveness and satisfaction with police performance: A large-scale survey study Publ. Adm. Rev. 77 2 2017 228 239
84 Ozamiz-Etxebarria N. Idoiaga Mondragon N. Dosil Santamaría M. Picaza Gorrotxategi M. Psychological symptoms during the two stages of lockdown in response to the COVID-19 outbreak: an investigation in a sample of citizens in Northern Spain Front. Psychol. 11 2020 1491 32625157
85 Zhao Y. Cheng S. Yu X. Xu H. Chinese public's attention to the COVID-19 epidemic on social media: observational descriptive study J. Med. Internet Res. 22 5 2020 e18825
86 Farooq A. Laato S. Islam A.N. Impact of online information on self-isolation intention during the COVID-19 pandemic: cross-sectional study J. Med. Internet Res. 22 5 2020 e19128
87 Maier C. Laumer S. Eckhardt A. Weitzel T. Giving too much social support: social overload on social networking sites Eur. J. Inf. Syst. 24 5 2015 447 464
| 36465702 | PMC9705009 | NO-CC CODE | 2022-12-01 23:20:29 | no | Int J Disaster Risk Reduct. 2023 Jan 29; 84:103469 | utf-8 | Int J Disaster Risk Reduct | 2,022 | 10.1016/j.ijdrr.2022.103469 | oa_other |
==== Front
J Affect Disord
J Affect Disord
Journal of Affective Disorders
0165-0327
1573-2517
Elsevier B.V.
S0165-0327(22)01342-8
10.1016/j.jad.2022.11.082
Research Paper
Impacts of mental health in the sleep pattern of healthcare professionals during the COVID-19 pandemic in Brazil
dos Santos Alves Maria Gustavo a⁎
de Oliveira Serpa Alexandre Luiz b
de Medeiros Chaves Ferreira Clarice c
de Andrade Vitor Douglas d
Ferreira Alessandra Rodrigues Hansen a
de Souza Costa Danielle a
Diaz Alexandre Paim e
da Silva Antônio Geraldo f
de Miranda Débora Marques a
Nicolato Rodrigo a
Malloy-Diniz Leandro Fernandes a
a Faculdade de Medicina, Universidade Federal de Minas Gerais. Professor Alfredo Balena Avenue, 190, 30.130-100 Belo Horizonte, Minas Gerais, Brazil
b Universidade Presbiteriana Mackenzie, Maria Antonia Street, 164, 01.222-010 São Paulo, Brazil
c Psychology Department, FUMEC University, Cobre Street, 200, Cruzeiro, 30.310-190 Belo Horizonte, Minas Gerais, Brazil
d Faculdade de Filosofia e Ciências Humanas, Universidade Federal de Minas Gerais, Professor Alfredo Balena Avenue, 190, 30.130-100 Belo Horizonte, Minas Gerais, Brazil
e Department of Psychiatry and Behavioral Sciences, The University of Texas, 7000 Fannin Street, 77030 Houston, TX, USA
f Associação Brasileira de Psiquiatria, Buenos Aires Street, 48, 3rd floor, 20.070-022 Rio de Janeiro, Brazil
⁎ Corresponding author at: Center of Technology in Molecular Medicine, Professor Alfredo Balena Avenue, 190, Belo Horizonte, Minas Gerais, Brazil.
29 11 2022
29 11 2022
27 4 2022
21 11 2022
24 11 2022
© 2022 Elsevier B.V. All rights reserved.
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Background
After >2 years of the Coronavirus Disease-19 (COVID-19) pandemic, it is well established how sleep symptoms are rising, especially among healthcare workers (HCW). The aim of this study is to evaluate what features are associated with sleep disturbances in the HCW population.
Methods
Cross-sectional and longitudinal analysis of social and clinical variables associated with sleep problems and insomnia incidence in HCW in a large, national-level cohort. The measurement of sleep problems was assessed by self-report using Jenkins Sleep Scale (JSS). A multivariate analysis was used in the cross-sectional design and generalized linear models were used in the longitudinal design.
Results
10,467 HCW were analyzed in the cross-sectional analysis, 3313 participants were analyzed in the three timepoints of the study. Sex, previously diagnosed mental illness and frontline work with COVID-19 were associated with higher scores in JSS in the univariate analysis. In the multivariate analysis, only previous diagnosis of mental illness was related with sleep difficulties, especially previously diagnosed insomnia. The longitudinal analysis concluded that previous diagnosis of mental illnesses was associated with higher levels of insomnia development (OR = 11.62). The self-reported disorders found to be major risk factors were addiction (OR = 7.69), generalized anxiety disorder (OR = 3.67), social anxiety (OR = 2.21) and bipolar disorder (OR = 2.21).
Limitations
Attrition bias.
Conclusions
Previous diagnosis of mental illness was strongly related to insomnia development in HCW during the COVID-19 pandemic. Strategies that focus on this population are advised.
Keywords
Sleep
Insomnia
Mental illness
Health professionals
COVID-19
Abbreviations
ADHD, attention-deficit hyperactivity disorder
ANOVA, Analysis of Variance
BSI, Brief Symptoms Inventory
CFI, comparative fit index
COVID-19, Coronavirus Disease-19
GAD, generalized anxiety disorder
GSI, Global Severity Index
HCW, healthcare workers
JSS, Jenkins Sleep Scale
MERS-CoV, Middle-East Respiratory Syndrome Coronavirus
OR, odds ratio
PTSD, post-traumatic stress disorder
RMSEA, root mean square error of approximation
SARS-CoV-1, Severe Acute Respiratory Syndrome Coronavirus-1
SD, standard deviation
SRMR, standardized root mean square residual
TLI, Tucker-Lewis index
==== Body
pmc1 Introduction
Sleep disorders are common psychiatric conditions, associated with considerable health and economic burden. Insomnia, the most common of them, is associated with a significant loss of quality-adjusted life years, more than other diseases such as arthritis, depression, and hypertension (Olfson et al., 2018). It is considered that direct and indirect costs associated with insomnia exceed >1 billion dollars annually in the United States (Wickwire et al., 2016), presumably due to its high incidence: approximately 1 among 4 American individuals develop insomnia yearly, as reported by Olfson et al. (2018).
There is a well-established relationship between sleep disturbances and acute stressors, especially those related to work (Hansen et al., 2021; Jiang et al., 2021; Scovelle et al., 2021). Not only do daily stressors influence sleep quality, but the reverse causal relationship is also reported in the literature, with higher sleep quality leading to positive mood effects (Blaxton et al., 2017). This cyclic relationship may play a major role in the pathophysiology of sleep disturbances, as well as other psychological disorders.
Indeed, in a pathological context, sleep diseases seem to be commonly comorbid to other psychiatric conditions. Some researchers have already been trying to explore a possible relationship between insomnia and mental disorders, such as different types of anxiety, post-traumatic stress disorder (PSTD), mood disorders, attention-deficit/hyperactivity disorder (ADHD), and others (Gould et al., 2017; Mantua et al., 2018; Tong et al., 2018).
It is a relative consensus that the Coronavirus-19 disease (COVID-19) pandemic has provoked significant changes in the population's lifestyle, on a global scale. The completely unexpected stressors related to COVID-19 may be, therefore, a major contributor to both sleep disturbances and psychological distress. 2 years after the formal declaration of COVID-19 as a pandemic, data is becoming available about the short- and long-term effects of the Coronavirus-Disease 19 (COVID-19) in mental health (Chen et al., 2021). The stressors commonly related with the pandemic involve social isolation, fear of death, bereavement, and direct impact of COVID-19 infection in psychiatric symptoms (Jin et al., 2021; Joaquim et al., 2021; Pérez-Mengual et al., 2021; de Sousa Moreira et al., 2021; Rogers et al., 2020).
Due to the intimate relationship between acute stressors and sleep disturbances, it is reasonable to consider that a unique scenario such as the pandemic would affect both symptoms and diagnoses related to sleep. In previous viral outbreaks, such as the Severe Acute Respiratory Syndrome Coronavirus-1 (SARS-CoV-1, 2002) and the Middle-East Respiratory Syndrome Coronavirus (MERS-CoV, 2012), the risks of psychological distress were already described (Maunder et al., 2003; Hawryluck et al., 2004; Mok et al., 2005; Almutairi et al., 2018). Until date, it has been reported that insomnia diagnoses were rising during the COVID-19 pandemic, as observed in Greece and China (Voitsidis et al., 2020; Li et al., 2020).
Previous large meta-analyses have already pointed out that healthcare workers (HCW) were at higher risk of insomnia than the general population (Pappa et al., 2020; Cénat et al., 2021). Two large studies evaluated sleep symptoms in HCW in Turkey during the COVID-19 pandemic, finding a prevalence of over 50 % of poor sleep quality among HCW (Şahin et al., 2020; Yılmaz et al., 2021). In these studies, the main risk factors for sleep disturbances in Turkish HCW were sex, education, length of professional experience, working with inpatient care or in frontline against COVID-19, and being a nurse. However, there is a lack of information about how Brazilian HCW are affected and, among them, who is at an even higher risk of insomnia development. Therefore, despite knowing that the sleep of HCW was more affected during the pandemic, the risk features and the strategies that can be made to reduce this burden remains unclear.
The purpose of this study is to analyze the risk factors associated with poor sleep and insomnia development during the COVID-19 pandemic in a large cohort of Brazilian HCW, altogether with clinical and psychiatric covariates that may be related to a higher risk of insomnia development.
2 Participants and methods
This research is part of the project “Influência da COVID-19 na Saúde Mental da população brasileira e de seus profissionais de saúde” (Influence of COVID-19 on the Mental Health of the Brazilian population and its health professionals), which was approved by the National Research Ethics Commission in May 2020 (Registration Number: 30,823.620.6.0000.5149). It follows the principles of the Declaration of Helsinki (1989).
As already described in previous papers, the project has as one of its main goals to evaluate a) physical health status; b) COVID-19 diagnosis and contact history; c) perceptions and concerns related to COVID-19 pandemic; and d) precautionary measures against COVID-19 (Joaquim et al., 2021). In the case of this study, our focus was on investigating a and b, especially sleep patterns and its possible association with COVID-19 incidence and sociodemographic data.
2.1 Participants
All health professionals registered on the official platform of the Ministry of Health of Brazil received an invitation to participate and fulfill the digital research's questionnaire, which included questions about sex, education level, marital status, and work category, including if the patient worked or not as a frontline professional against COVID-19, psychiatric symptoms and COVID-19 related stressors. The contact was made using the e-mail address registered in the Ministry of Health of Brazil. The patients were advised that the participation was voluntary, and that no incentive would be provided to the participants. They were also informed that the survey would take about 25 min to be completed and that it was composed of 13 pages.
The patients who did not consent to participate in the research, those under 18 years old and those who self-reported neurological diseases (epilepsy, seizures, brain tumors, hydrocephalus, agenesis of the corpus callosum, etc.) with a declared impact on cognitive capacity were excluded from the study. Only patients who consented to participate in the research and those 18 years old or over were allowed to proceed with the questionnaire.
Those with incomplete questionnaires (defined as >20 % of missing data) were not included in the statistical analysis. For the purposes of this study, patients with recent hospitalization and thyroid diseases were also excluded due to possible impacts on sleep.
2.2 Methods
The questionnaire of this study was made using the SurveyMonkey platform (Survey-Monkey, Palo Alto, California, www.surveymonkey.com), which implements data security protocols in compliance with the rules established by the Health Insurance Portability and Accountability Act for the business plan, which was used in this study. The data was pre-processed to remove any information that allows the identification of the participants and stored in a private cloud folder created in OneDrive, which was available only to the registered members of the project.
The questionnaire had been tested by the researchers to evaluate its usability and technical functionality. In the first pages of the questionnaire, they answered an informed consent that stated the approximate length of time of the survey, who were the investigators responsible for the study and which data would be stored. The survey was not open and only the target population was given access to it. In any question, the patient could leave a blank answer and he/she could use a “back” button to revise a specific question before submitting the questionnaire. Each patient could answer the survey only once. Cookies were used to identify possible duplicate answers from the same patient. If any duplicate was found, the less complete one was deleted from the dataset.
To avoid attrition bias in the cross-sectional and longitudinal analyses of this study, only the patients that answered both the first and the second survey were included in the cross-sectional analysis (which, however, was performed using only the answers of the first survey).
2.3 Instruments
The questionnaire was composed by anonymous questions that asked about the covariates included in the study, including sex, gender, age, type of work, psychiatric background, educational degree, marital status and 2 specific instruments, the Jenkins Sleep Scale (JSS) (Jenkins et al., 1988) and the Brief Symptoms Inventory (BSI) (Derogatis, 1982).
In the cross-sectional analysis, the primary outcome was to evaluate whether the covariates included in the questionnaire influenced the sleep pattern of the patients. It was measured by the JSS, which varies from 0 to 6. JSS consists of 4 questions, each approaching how often the respondent felt different sleep symptoms, resulting in larger scores for increased sleep distress. It was validated for the Portuguese language in 2014 (Reis et al., 2014).
BSI was used both in the cross-sectional and longitudinal analyses. It is a 53-item self-report inventory designed to assess psychological distress and psychopathological symptoms across nine dimensions (Somatization, Obsessive-Compulsion, Interpersonal Sensitivity, Depression, Anxiety, Hostility, Phobic Anxiety, Paranoid Ideation, Psychoticism) and a Global Severity Index (GSI), which represents the sum of all the other items. Each dimension is quantified via a Likert scale from 0 to 4, ranging from ‘not at all’ to ‘extremely’. It is designed for a minimum age of 13 years, comprehending the target audience of this study, and its internal structure was verified for the Brazilian population during the COVID-19 pandemic (Serpa et al., 2021).
2.4 Statistical analysis
All quantitative data was tested for normality with the Shapiro-Francia test. In the descriptive analysis, normally distributed data were presented as mean (standard deviation) and non-normally distributed data were presented as median (25th percentile–75th percentile).
In the cross-sectional arm of the study, the covariates with only two answers were analyzed using one-tailed Student's t-test. The covariates with three or more answers were analyzed with Analysis of Variance (ANOVA). The analysis of correlation between quantitative covariates was made using linear regression. α-Value was fixed as 0.05 for the univariate analysis and 0.025 for the multivariate analysis. Pearson's R2 value was used to determine correlation in linear regression, with a threshold of 0.5.
The longitudinal analysis was performed using a mixed effects logistic regression. The predicted variable was the presence or absence of insomnia according to the JSS classification at the second data collection timepoint. Sex, whether has previous mental disorder condition, being a COVID-19 frontline health professional, occupation, and the interval between the two data collections for each individual were the predictors. Random effects were tested for occupation and the data collection interval. A second model was tested, where the previous mental disorder condition variable was changed by each kind of mental disorder. All the mental disorders were tested in the same model because some participants reported several conditions.
In the end, a cross-lagged panel model was designed to investigate the longitudinal effect between insomnia and global psychological distress across the timepoints. The model fit will be evaluated considering the recommendations of Kline (Kline, 2015), considering an acceptable model fit when comparative fit index (CFI) > 0.930, Tucker-Lewis index (TLI) > 0.900, root mean square error of approximation (RMSEA) and standardized root mean square residual (SRMR) < 0.08. If the model reached an acceptable fit, the standardized regression estimates were interpreted in terms of effect sizes. Standardized β equal or >0.10, 0.20 and 0.30 were considered small, typical and large, according to Gignac and Szodorai (2016). All the analyses were performed on R software (R Core Team, 2021, https://www.r-project.org/), using lmerTest (Kuznetsova et al., 2017), lme4 (Bates et al., 2015), emmeans (Lenth, 2022), lavaan (Rosseel, 2012), semPlot (Epskamp, 2022) and jtools (Long, 2020).
2.5 Timing of data collection
The data collection for the longitudinal analysis was performed in three different periods: the first one was from 05/09/20 to 06/07/20, the second from 11/23/20 to 01/29/21 and the third one from 05/11/21 to 08/12/21. Fig. 1 illustrates the periods of data collection (timepoints) accompanied by the daily number of deaths per million Brazilian inhabitants, a graph adapted from Our World in Data (Ritchie et al., 2020).Fig. 1 Periods of data collection (blue) and number of deaths due to Covid-19 per million of inhabitants in Brazil (green). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Fig. 1
2.6 Standardization of report
The report of the results in this study was developed following the STROBE checklist for cohort studies, the STROBE checklist for cross-sectional studies39, and the CHERRIE checklist for online questionnaires (von Elm et al., 2008; Eysenbach, 2004).
3 Results
3.1 Cross-sectional analysis
The first survey reached a total of 223,867 participants. Those who did not consent to participate in the research (n = 18,276) and those under 18 years old (n = 25,461) were excluded at the first moment. There was 73 % of attrition rate from the first to the second timepoints of data collection.
10,490 participants were included in the first two surveys and 23 met the exclusion criteria, totalizing 10,467 eligible individuals. Most of the sample was female (77 %), without previous diagnosis of mental illness (65 %) and did not work in the frontline against COVID-19 (56 %). The JSS score was found to be normally distributed in both periods, with a global mean of 2.06 (SD = 1.38) and <1 % of missing data in the first collection. Table 1 summarizes the analyses of JSS scores among the categorical covariates included. Fig. 2 illustrates the differences in JSS scores between patients with and without previous diagnosis of insomnia, the covariate with the largest distinction in JSS score means (p < 0.01).Table 1 Analysis of the differences in JSS mean scores among categorical covariates in the cross-sectional study.
Table 1Covariates n (%)a JSS score p-Value Missing data (%)
Sex 5.37
Female 8113 (77 %) 2.15 (1.38) Reference
Male 1792 (17 %) 1.66 (1.29) <0.01
Marital status 6.15
Single 4136 (39 %) 2.11 (1.36) Reference
Married/live together 4837 (46 %) 2.02 (1.38) 0.14
Divorced 850 (8 %) 2.02 (1.42) 0.15
Education 12.03
Graduation 7622 (72 %) 2.06 (1.38) Reference
Master's degree 1107 (10 %) 2.09 (1.33) 0.53
Doctorate degree 479 (4 %) 1.97 (1.32) 0.14
Previously diagnosed mental illness 5.03
Yes 3116 (29 %) 2.65 (1.35) Reference
No 6824 (65 %) 1.80 (1.30) <0.01
Previously diagnosed insomnia 5.03
Yes 784 (7 %) 3.42 (1.16) Reference
No 9156 (87 %) 1.95 (1.33) <0.01
Healthcare providers 66.62
Not related 657 (6 %) 1.92 (1.41) Reference
Related, but not directly associated with care 1075 (10 %) 2.20 (1.47) <0.01
Directly associated with care 1762 (16 %) 2.06 (1.35) 0.03
Covid-19 frontline healthcare professional 5.43
Yes 3969 (37 %) 2.17 (1.40) Reference
No 5930 (56 %) 1.98 (1.35) <0.01
Previous trauma 55.97
Unspecified 282 (2 %) 2.39 (1.50) Reference
Is not associated to Covid-19 pandemic 3776 (36 %) 2.22 (1.38) 0.046
Is indirectly associated to Covid-19 pandemic 221 (2 %) 2.48 (1.36) 0.45
Is directly associated to Covid-19 pandemic 330 (3 %) 2.40 (1.48) 0.97
a (%) refers to the percentage of respondents in the total sample of the cross-sectional arm (n = 10,467).
Fig. 2 Raincloud plot of JSS scores by previously diagnosed insomnia stated in timepoint 1.
JSS: Jenkins Sleep Scale.
Fig. 2
Neither age nor any of the BSI dimension scores were normally distributed. The median age of the sample was 37.24 (30.50–45.37) years old, with a missing data of 10 %. Ages below 18 years and above 100 years (1 % of total sample) were considered unreliable due to the target audience of the study and were, therefore, excluded from the analysis. The linear regression comparing age and JSS scores resulted in a Pearson's R2 value of 0.01 (p < 0.001). The results of correlation between BSI and JSS scores are summarized in Table 2 .Table 2 Description of BSI dimensions and their correlations with JSS scores in the cross-sectional arm.
Table 2Dimensions Median (25th percentile–75th percentile) R2 p-Value Missing data (%)
Somatization 0.29 (0.00–0.71) 0.22 <0.01 <1 %
Obsessive-Compulsion 0.67 (0.33–1.33) 0.25 <0.01 <1 %
Interpersonal Sensitivity 0.50 (0.00–1.00) 0.16 <0.01 <1 %
Depression 0.67 (0.33–1.33) 0.23 <0.01 <1 %
Anxiety 0.67 (0.33–1.33) 0.27 <0.01 <1 %
Hostility 0.60 (0.20–1.00) 0.19 <0.01 <1 %
Phobic Anxiety 0.60 (0.20–1.40) 0.14 <0.01 <1 %
Paranoid Ideation 0.40 (0.20–1.00) 0.13 <0.01 <1 %
Psychoticism 0.20 (0.00–0.80) 0.18 <0.01 <1 %
GSI 0.57 (0.26–1.08) 0.30 <0.01 <1 %
BSI: Brief Symptoms Inventory; GSI: Global Severity Index; JSS: Jenkins Sleep Scale; R2: Pearson's R value squared.
3.2 Longitudinal analysis
The covariates included in the longitudinal analysis were sex, previously diagnosed mental illness, direct association with healthcare and frontline work against COVID-19. Age was excluded in the multivariate analysis due to its low Pearson's R value.
As Fig. 1 implicitly shows, the intervals among data collection varied per each individual; therefore, these intervals also became a new covariate (ΔT), due to its possible influence on insomnia development.
The first model of the multivariate analysis compared the first and second periods of data collection. The results of the generalized linear model are shown in Table 3 and the Odds Ratios for all the covariates analyzed are illustrated in Fig. 3 . ΔT was found to have no influence on insomnia development, so it was excluded in further analyses. In addition, the random effects for both occupation and ΔT were not statistically significant.Table 3 Summary of mixed effects logistic regression for models 1 and 2.
Table 3Covariates Model 1 Model 2
OR 95 % CI p-Value OR 95 % CI p-Value
ΔT 1.00 [0.99, 1.01] 0.98 1.00 [0.99, 1.01] 0.86
Male 0.80 [0.53, 1.19] 0.27 0.75 [0.49, 1.13] 0.17
Non-health occupation 1.05 [0.72, 1.54] 0.79 1.07 [0.72, 1.58] 0.75
Health occupation 1.07 [0.77, 1.49] 0.69 1.08 [0.77, 1.53] 0.66
Covid19 frontline health professional 1.18 [0.91, 1.53] 0.21 1.20 [0.92, 1.57] 0.18
Mental disorder 11.61⁎⁎⁎ [8.83, 15.28] <0.001
Addiction 7.74⁎⁎ [1.68, 35.62] 0.01
Eat disorder 1.12 [0.63, 1.98] 0.69
Anxiety generalized 3.67⁎⁎⁎ [2.79, 4.81] <0.001
Anxiety social 2.22⁎⁎ [1.35, 3.66] <0.001
Bipolar disorder 2.23⁎ [1.16, 4.29] 0.02
Conduct disorder 3.02 [0.32, 28.66] 0.34
ADHD 2.37⁎ [1.15, 4.89] 0.02
Depression 3.37⁎⁎⁎ [2.56, 4.44] <0.001
Premenstrual MDD 1.35 [0.61, 2.99] 0.46
Postpartum depression 0.44 [0.16, 1.19] 0.11
Learning specific disorder 8.32 [0.84, 82.49] 0.07
Autism 278,605.62 [0.00, 7,37E+281] 0.97
Schizophrenia 0.09 [0.00, 4.22] 0.22
PTSD 1.62 [0.81, 3.22] 0.17
N 3671 3671
AIC 1907.68 1927.43
BIC 1951.13 2051.59
ADHD: attention-deficit/hyperactivity disorder; MDD: major depressive disorder; PTSD: post-traumatic stress disorder; AIC; BIC.
⁎⁎⁎ p < 0.001.
⁎⁎ p < 0.01.
⁎ p < 0.05.
Fig. 3 Risk factors found in the cross-sectional arm and their OR for insomnia development in the first model of the multivariate analysis of the longitudinal arm.
OR: odds ratio; empty dot: OR; continuous line: confidence interval (95 %).
Fig. 3
As only previously mental illnesses were associated with insomnia development (OR = 11.62 [8.83–15.29]), a second model was performed, stratifying the previous diagnoses collected in the first survey: Generalized Anxiety Disorder (GAD, n = 1869), depression (n = 1835), social anxiety (n = 310), eating disorder (n = 199), PTSD (n = 169), ADHD (n = 165), bipolar disorder (n = 161), postpartum depression (n = 135), premenstrual dysphoric disorder (n = 109), addiction (n = 31), learning specific disorder (n = 15), conduct disorder (n = 10), autism (n = 9), and schizophrenia (n = 9).
The results of the second generalized linear model are described in Table 3 and illustrated in Fig. 4 . The mental illnesses associated with insomnia development were addiction (OR = 7.69 [1.67–35.39]), GAD (OR = 3.67 [2.79–4.82]), depression (OR = 3.37 [2.56–4.44]), ADHD (OR = 2.35 [1.14–4.86]), social anxiety (OR = 2.21 [1.34–3.64]), and bipolar disorder (OR = 2.21 [1.15–4.27]). Between the first and second timepoints of data collection, 878 patients (9.07 % of the initial non-insomniac sample) reported a new diagnosis of insomnia. 698 (79.5 %) of them had reported a previous diagnosis of mental illness in the first survey.Fig. 4 Previously diagnosed mental illnesses reported in timepoint 1 and their OR for insomnia development in the adjusted model of the multivariate analysis of the longitudinal arm.
ADHD: attention deficit/hyperactivity disorder; MDD: major depressive disorder; OR: odds ratio; PSTD: post-traumatic stress disorder; empty dot: OR; continuous line: confidence interval (95 %).
Fig. 4
3313 eligible patients answered the third survey, which showed 73 new cases of insomnia (2.20 %). No association was found between previous diagnosis of mental illnesses and development of insomnia in the analysis of the sample that reached the third timepoint. Fig. 5 illustrates the JSS Score mean of four different groups of analysis, which found no statistical difference neither among inter-group scores in the same timepoint nor longitudinally for any group. Unreliable answers (e.g., individuals who reported previous diagnosis of insomnia in timepoint 1 but no previous diagnosis of insomnia in timepoint 2) were excluded from this analysis (n = 633, 19.1 %).Fig. 5 JSS score means by timepoint in different groups of insomnia diagnosis patterns in HCW.
G0: individuals who did not report any previous diagnosis of insomnia in all the timepoints (n = 2402); G1: individuals who reported previous diagnosis of insomnia in timepoint 1 (n = 47); G2: individuals who did not report previous diagnosis of insomnia in timepoint 1 but reported it in timepoint 2 (n = 158); G3: individuals who did not report previous diagnosis of insomnia in both timepoints 1 and 2 but reported it in timepoint 3 (n = 73). The scale of JSS score was enlarged for better visualization.
Fig. 5
The model used to test the cross-influence between the GSI index of BSI and insomnia development reached a good fit (CFI = 0.982, TLI = 0.931, RMSEA = 0.073 95 % CI [0.058–0.089], SRMR = 0.053) and all the regressions coefficients were significant (Fig. 5). Nevertheless, 3 out 4 effects were close to the boundary of no to small effects. The effect of insomnia diagnosis in timepoint 2 on BSI score in timepoint 3 was the only exception, considered of small size. Most of the effects on a variable from the subsequent timepoint are due to its own state at the current timepoint. The general correlations between GSI scores and insomnia diagnosis are small, varying from 0.21 (timepoints 1 and 3) to 0.32 (timepoint 2) (Fig. 6 ).Fig. 6 Cross-influence between GSI index of BSI and insomnia development in timepoints 1, 2 and 3.
GSI: Global Severity Index; BSI: Brief Symptoms Inventory; INS_1: Individuals who reported insomnia in timepoint 1; INS_2: Individuals with newly diagnosed insomnia in timepoint 2; INS_3: Individuals with newly diagnosed insomnia in timepoint 3; GSI_1: GSI score in timepoint 1; GSI_2: GSI score in timepoint 2; GSI_3: GSI score in timepoint 3. **p < 0.01; ***p < 0.001.
Fig. 6
4 Discussion
In this study, the sleep pattern of Brazilian HCW was measured and associated with many covariates in a large cohort. It is noteworthy that sleep disturbances in HCW have already been observed in previous pandemics, such as the ones related to SARS-CoV-1 and MERS-CoV (Almutairi et al., 2018; Voitsidis et al., 2020). Even now, meta-analyses have observed the increase in sleep disorders (Pappa et al., 2020; Cénat et al., 2021). Previous studies have already found risk factors for poor sleep symptoms in Turkish HCW, especially related to sex, education, professional experience, working with inpatient care or in frontline against COVID-19, and being a nurse.
This study adds to previous data because it focuses on a large cohort of high risk Brazilian professionals for sleep disorders during a stressful period of two years in the COVID-19 pandemic, which were proven to be at even higher risk of sleep disturbances.
As expected, previous diagnosis of insomnia was the largest predictor of poor sleep patterns, while other clinical covariates (such as education and marital status) were found to have no correlation with them. Although covariates such as sex and professional characteristics were correlated with differences in JSS scores in the cross-sectional analysis, only previous mental illness was associated with a higher risk of insomnia development during the pandemic. This offers a contrast with previous data, especially in Turkey, which found these covariates to be strongly related to poor quality of sleep. This difference may be explained by differences in population and evaluation of psychiatric comorbidities, which, in the longitudinal arm of this study, had a higher impact in sleep quality. It is noteworthy that the results of the cross-sectional arm of this study are very similar to previous data on risk factors for sleep disturbances in HCW.
For a satisfactory critical analysis of the results of this study, it is important to differentiate the continuous sleep pattern measure (the JSS score) from the categorical covariate (insomnia development). First, there was a high percentage of unreliable answers (almost 20 %), which could be explained by transient insomnia in this sample of HCW. About the JSS scores, they did not significantly vary across time in all groups of recently diagnosed insomnia. It was able to infer specific time periods of insomnia diagnosis in the longitudinal analysis by observing the shift in the answers for the question “have you already been diagnosed with insomnia?” among timepoints. Despite the lack of correlation between JSS score means and recent insomnia diagnosis, the cross-sectional analysis showed a precise correspondence of previous insomnia diagnosis and sleep problems. This may lead to the conclusion that, in the long-term, patients with a formal diagnosis of insomnia are more prone to sustain high levels of sleep problems as expected. Further evaluation is needed to conclude if JSS scores tend to rise in the long-term follow-up of HCW with the diagnosis of insomnia.
In the longitudinal analysis, the only risk factor found for insomnia development in HCW was the previous diagnosis of a mental illness — especially addiction, ADHD and anxiety and mood disorders. Age, gender, education, and marital status were not important risk factors in this cohort. These results are correspondent to previous studies that have related sleep disturbances with other psychiatric illnesses (Mantua et al., 2018; Tong et al., 2018). After discrimination of different previous mental illness diagnosis, it was found that individuals with addiction were at a larger risk of insomnia development, followed by those with GAD, social anxiety, bipolar disorder, ADHD, and depression. Considering the stressful characteristics of the moment and the disruption in the health and specifically psychiatric care, it is reasonable to assume a possibility of symptoms worsening of the previously diagnosed psychiatric disorder. The predominance of anxiety and mood disorders in the list of comorbidities may indicate an overlap of pathogenesis between them and insomnia, although a causal relationship remains largely unclear. Despite previous diagnosis of addiction being the most important risk factor for insomnia development, this study prospects but did not detail the substance of addiction.
Years before, it was known that HCW were more prone to these disturbances, but there was scarce data about the specific risk factors that could lead to a formal diagnosis of insomnia. In this study, while age, gender, marital status, and education do not seem to be related with insomnia, there was a clear distinction between the HCW with previous diagnosis of mental illnesses and the ones without it. Understanding these results can help us to guide our efforts toward populations with higher risk of insomnia development, and to prepare ourselves for possible future outbreaks.
4.1 Limitations
This study has limitations, including a high level of attrition bias. The subjects were included via professional entries of the Ministry of Health of Brazil invitees, which limits its generalizability in other countries and in non-HCW patients. However, it also has strong qualities, such as conformity with CHERRIES checklist for online questionnaires, a high number of respondents, strict statistical analysis, and a national level of divulgation by official institutions.
Among the strengths, the results of this study help to elucidate the most important risk factors in HCW during the COVID-19 pandemic for insomnia development, which was the strongest predictor of poor sleep in its cross-sectional analysis. They help to understand how this population, which is known to be at a higher risk of insomnia, can have a risk stratification that directs efforts toward the prevention and early treatment of sleep disturbances. Further investigation of the probable delay in medical seek among HCW is advised, as well as a deeper study on the effects of substance addiction on sleep among HCW.
5 Conclusion
The COVID-19 pandemic is known to be related with a higher incidence of sleep disturbances, especially in HCW. In this study, it was found that previous insomnia diagnosis is the covariate most closely associated with sleep problems, and previous mental health illnesses were the major risk factors for insomnia development in HCW during the COVID-19 pandemics. Further investigation of mental health symptoms and long-term effects of insomnia in this population is advised.
Funding resources
This work was supported by Panamerican Health Organization (PAHO, grant number SCON2020-00202) and 10.13039/501100003593 Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPQ, grant number 401542/2020-3) in task force with Brazilian Association of Psychiatry (ABP), Brazilian Association of Impulsivity and Dual Pathology (ABIPD), and Saúde Mental Baseada em Evidências (SAMBE) Research Group. The funding sources had no direct involvement in the study design, collection/analysis of data, writing, nor the decision to submit it to publishing.
CRediT authorship contribution statement
Gustavo dos Santos Alves Maria: Conceptualization, Methodology, Validation, Formal analysis, Investigation, Data curation, Writing – original draft, Writing – review & editing, Visualization, Project administration. Alexandre Luiz de Oliveira Serpa: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Resources, Data curation, Writing – original draft, Writing – review & editing, Visualization. Clarice de Medeiros Chaves Ferreira: Conceptualization, Methodology, Data curation, Writing – original draft, Writing – review & editing. Vitor Douglas de Andrade: Conceptualization, Methodology, Data curation, Writing – original draft, Writing – review & editing. Alessandra Rodrigues Hansen Ferreira: Conceptualization, Methodology, Data curation, Writing – original draft, Writing – review & editing. Danielle de Souza Costa: Conceptualization, Methodology, Validation, Formal analysis, Investigation, Data curation, Writing – review & editing, Visualization, Supervision. Alexandre Paim Diaz: Conceptualization, Methodology, Validation, Formal analysis, Investigation, Data curation, Writing – review & editing, Visualization, Supervision. Antônio Geraldo da Silva: Conceptualization, Methodology, Validation, Formal analysis, Investigation, Data curation, Writing – review & editing, Visualization, Supervision. Débora Marques de Miranda: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Data curation, Writing – original draft, Writing – review & editing, Visualization, Supervision, Project administration. Rodrigo Nicolato: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Data curation, Writing – original draft, Writing – review & editing, Visualization, Supervision, Project administration. Leandro Fernandes Malloy-Diniz: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Data curation, Writing – original draft, Writing – review & editing, Visualization, Supervision, Project administration.
Conflict of interest
None.
Acknowledgements
None.
==== Refs
References
Almutairi A.F. Adlan A.A. Balkhy H.H. Abbas O.A. Clark A.M. "It feels like I'm the dirtiest person in the world.": exploring the experiences of healthcare providers who survived MERS-CoV in Saudi Arabia J. Infect. Public Health 11 2 2018 187 191 10.1016/j.jiph.2017.06.011 28676285
Bates Douglas Maechler Martin Bolker Ben Walker Steve Fitting linear mixed-effects models using lme4 J. Stat. Softw. 67 1 2015 1 48 10.18637/jss.v067.i01
Blaxton J.M. Bergeman C.S. Whitehead B.R. Braun M.E. Payne J.D. Relationships among nightly sleep quality, daily stress, and daily affect J. Gerontol. B Psychol. Sci. Soc. Sci. 72 3 2017 363 372 10.1093/geronb/gbv060 PMID: 26307483; PMCID: PMC5927097 26307483
Cénat J.M. Blais-Rochette C. Kokou-Kpolou C.K. Noorishad P.G. Mukunzi J.N. McIntee S.E. Dalexis R.D. Goulet M.A. Labelle P.R. Prevalence of symptoms of depression, anxiety, insomnia, posttraumatic stress disorder, and psychological distress among populations affected by the COVID-19 pandemic: a systematic review and meta-analysis Psychiatry Res. 295 2021 113599 10.1016/j.psychres.2020.113599 Epub 2020 Nov 26. PMID: 33285346; PMCID: PMC7689353
Chen P.J. Pusica Y. Sohaei D. Prassas I. Diamandis E.P. An overview of mental health during the COVID-19 pandemic Diagnosis 8 4 2021 403 412 10.1515/dx-2021-0046 PMID: 34214386 34214386
Derogatis L.R. Brief Symptom Inventory (BSI) [Database record] APA PsycTests 1982 10.1037/t00789-000
von Elm E. Altman D.G. Egger M. Pocock S.J. Gøtzsche P.C. Vandenbroucke J.P. STROBE Initiative The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies Apr J. Clin. Epidemiol. 61 4 2008 344 349 PMID: 18313558 18313558
Epskamp Sacha semPlot: path diagrams and visual analysis of various SEM packages' output. R package version 1.1.5 URL https://CRAN.R-project.org/package=semPlot 2022
Eysenbach G. Improving the quality of web surveys: the checklist for reporting results of internet e-surveys(Cherries) J. Med. Internet Res. 6 3 2004 e34
Gignac G.E. Szodorai E.T. Effect size guidelines for individual differences researchers Personal. Individ. Differ. 102 2016 74 78 10.1016/j.paid.2016.06.069
Gould C.E. Spira A.P. Liou-Johnson V. Cassidy-Eagle E. Kawai M. Mashal N. Beaudreau S.A. … Association of anxiety symptom clusters with sleep quality and daytime sleepiness J.Gerontol.Ser.B 73 3 2017 413 420 10.1093/geronb/gbx020
Hansen D.A. Satterfield B.C. Layton M.E. Van Dongen H.P.A. Sleep deprivation and sleep-onset insomnia are associated with blunted physiological reactivity to stressors Mil. Med. 186 Suppl. 1 2021 246 252 10.1093/milmed/usaa464 PMID: 33499519 33499519
Hawryluck L. Gold W.L. Robinson S. Pogorski S. Galea S. Styra R. SARS control and psychological effects of quarantine, Toronto, Canada Jul Emerg. Infect. Dis. 10 7 2004 1206 1212 10.3201/eid1007.030703 PMID: 15324539; PMCID: PMC3323345 15324539
Jenkins C.D. Stanton B.A. Niemcryk S.J. Rose R.M. A scale for the estimation of sleep problems in clinical research J. Clin. Epidemiol. 41 1988 313 321 3351539
Jiang Y. Li P. Zhong L. Liu B. Gao X. Ning L. Lian Y. Liu J. The influence of changes in work stressors and coping resources on sleep disturbances: evidence from the OHSPIW cohort study Aug 13 Sleep 44 8 2021 zsab039 10.1093/sleep/zsab039 PMID: 33608703
Jin Y. Sun T. Zheng P. An J. Mass quarantine and mental health during COVID-19: a meta-analysis Dec 1 J. Affect. Disord. 295 2021 1335 1346 10.1016/j.jad.2021.08.067 Epub 2021 Sep 2. PMID: 34706447; PMCID: PMC8674683 34706447
Joaquim R.M. Pinto A.L.C.B. Guatimosim R.F. Bereavement and psychological distress during COVID-19 pandemics: the impact of death experience on mental health Curr.Res.Behav.Sci. 2 2021 1 7 10.1016/j.crbeha.2021.100019
Kline R. Principles And Practice of Structural Equation Modeling 4th ed. 2015 The Guilford New York
Kuznetsova A. Brockhoff P.B. Christensen R.H.B. lmerTest package: tests in linear mixed effects models J. Stat. Softw. 82 13 2017 1 26 10.18637/jss.v082.i13
Lenth Russell V. emmeans: estimated marginal means, aka least-squares means. R package version 1.7.2 URL https://CRAN.R-project.org/package=emmeans 2022
Li Y. Qin Q. Sun Q. Sanford L.D. Vgontzas A.N. Tang X. Insomnia and psychological reactions during the COVID-19 outbreak in China J. Clin. Sleep Med. 16 8 2020 1417 1418 10.5664/jcsm.8524 PMID: 32351206; PMCID: PMC7446072 32351206
Long J.A. jtools: analysis and presentation of social scientific data. R package version 2.1.0 URL https://cran.r-project.org/package=jtools 2020
Mantua J. Helms S.M. Weymann K.B. Capaldi V.F. Lim M.M. Sleep quality and emotion regulation interact to predict anxiety in veterans with PTSD Behav. Neurol. 2018 2018 1 10 10.1155/2018/7940832
Maunder R. Hunter J. Vincent L. Bennett J. Peladeau N. Leszcz M. Sadavoy J. Verhaeghe L.M. Steinberg R. Mazzulli T. The immediate psychological and occupational impact of the 2003 SARS outbreak in a teaching hospital CMAJ 168 10 2003 1245 1251 PMID: 12743065; PMCID: PMC154178 12743065
Mok E. Chung B.P. Chung J.W. Wong T.K. An exploratory study of nurses suffering from severe acute respiratory syndrome (SARS) Int. J. Nurs. Pract. 11 4 2005 150 160 10.1111/j.1440-172X.2005.00520.x PMID: 15985093; PMCID: PMC7165571 15985093
Olfson M. Wall M. Liu S.M. Morin C.M. Blanco C. Insomnia and impaired quality of life in the United States Sep 11 J. Clin. Psychiatry 79 5 2018 17m12020 10.4088/JCP.17m12020 PMID: 30256547
Pappa S. Ntella V. Giannakas T. Giannakoulis V.G. Papoutsi E. Katsaounou P. Prevalence of depression, anxiety, and insomnia among healthcare workers during the COVID-19 pandemic: a systematic review and meta-analysis Aug Brain Behav. Immun. 88 2020 901 907 10.1016/j.bbi.2020.05.026 Epub 2020 May 8. Erratum in: Brain Behav Immun. 2021 Feb;92:247. PMID: 32437915; PMCID: PMC7206431 32437915
Pérez-Mengual N. Aragonés-Barbera I. Moret-Tatay C. Moliner-Albero A.R. The relationship of fear of death between neuroticism and anxiety during the COVID-19 pandemic Front. Psychiatry 12 2021 648498 10.3389/fpsyt.2021.648498 PMID: 33959053; PMCID: PMC8093560
Reis Cátia Mestre Catarina Tecedeiro Miguel Paiva Teresa Translation, cross-cultural adaptation and psychometric properties of the Jenkins Sleep Scale in a sample of Portuguese shift workers Laboratório de Psicologia 12 2014 89 98 10.14417/lp.894
Ritchie Hannah Mathieu Edouard Rodés-Guirao Lucas Appel Cameron Giattino Charlie Ortiz-Ospina Esteban Hasell Joe Macdonald Bobbie Beltekian Diana Roser Max Coronavirus pandemic (COVID-19) Published online at OurWorldInData.org. Retrieved from https://ourworldindata.org/coronavirus 2020 [Online Resource]
Rogers J.P. Chesney E. Oliver D. Psychiatric and neuropsychiatric presentations associated with severe coronavirus infections: a systematic review and meta-analysis with comparison to the COVID-19 pandemic Lancet Psychiatry 7 7 2020 611 627 10.1016/S2215-0366(20)30203-0 32437679
Rosseel Yves lavaan: an R package for structural equation modeling J. Stat. Softw. 48 2 2012 1 36 10.18637/jss.v048.i02
Şahin M.K. Aker S. Şahin G. Karabekiroğlu A. Prevalence of depression, anxiety, distress and insomnia and related factors in healthcare workers during COVID-19 pandemic in Turkey J. Community Health 45 6 2020 1168 1177 32915381
Scovelle A.J. King T. Shields M. O'Neil A. Lallukka T. Hewitt B. Milner A. Do psychosocial job stressors differentially affect the sleep quality of men and women? A study using the HILDA survey Eur. J. Pub. Health 31 4 2021 736 738 10.1093/eurpub/ckab056 PMID: 33963842 33963842
Serpa A.L.O. Costa D.S. Ferreira C.M.C. Pinheiro M.I.C. Diaz A.P. de Paula J.J. Miranda D.M. da Silva A.G. Malloy-Diniz L.F. Brief symptoms inventory psychometric properties supports the hypothesis of a general psychopathological factor Trends Psychiatry Psychother. 2021 10.47626/2237-6089-2021-0207 Epub ahead of print. PMID: 33760429
de Sousa Moreira J.L. Barbosa S.M.B. Vieira J.G. Chaves N.C.B. Felix E.B.G. Feitosa P.W.G. da Cruz I.S. da Silva C.G.L. Neto M.L.R. The psychiatric and neuropsychiatric repercussions associated with severe infections of COVID-19 and other coronaviruses Prog. Neuro-Psychopharmacol. Biol. Psychiatry 106 2021 110159 10.1016/j.pnpbp.2020.110159 Epub 2020 Nov 2. PMID: 33147504; PMCID: PMC7605739
Tong L. Ye Y. Yan Q. The moderating roles of bedtime activities and anxiety/depression in the relationship between attention-deficit/hyperactivity disorder symptoms and sleep problems in children BMC Psychiatry 18 1 2018 10.1186/s12888-018-1879-4
Voitsidis P. Gliatas I. Bairachtari V. Papadopoulou K. Papageorgiou G. Parlapani E. Syngelakis M. Holeva V. Diakogiannis I. Insomnia during the COVID-19 pandemic in a Greek population Psychiatry Res. 289 2020 113076 10.1016/j.psychres.2020.113076 Jul
Wickwire E.M. Shaya F.T. Scharf S.M. Health economics of insomnia treatments: the return on investment for a good night's sleep Sleep Med. Rev. 30 2016 72 82 10.1016/j.smrv.2015.11.004 Epub 2015 Nov 28 PMID: 26874067 26874067
Yılmaz M. Kıraç Y. Sahin M.K. Sleep quality and related factors in a sample of Turkish healthcare workers during the COVID-19 pandemic: a cross-sectional study Int. J. Clin. Pract. 75 11 2021
| 36455718 | PMC9705011 | NO-CC CODE | 2022-12-10 23:15:28 | no | J Affect Disord. 2023 Feb 15; 323:472-481 | utf-8 | J Affect Disord | 2,022 | 10.1016/j.jad.2022.11.082 | oa_other |
==== Front
Results Eng
Results Eng
Results in Engineering
2590-1230
The Author(s). Published by Elsevier B.V.
S2590-1230(22)00456-X
10.1016/j.rineng.2022.100786
100786
Article
Modeling and analysis of monkeypox disease using fractional derivatives
Okyere Samuel ∗
Ackora-Prah Joseph
Department of Mathematics, Kwame Nkrumah University of Science and Technology, Kumasi, Ghana
∗ Corresponding author.
29 11 2022
3 2023
29 11 2022
17 100786100786
14 10 2022
10 11 2022
17 11 2022
© 2022 The Author(s)
2022
Elsevier has created a Monkeypox Information Center (https://www.elsevier.com/connect/monkeypox-information-center) in response to the declared public health emergency of international concern, with free information in English on the monkeypox virus. The Monkeypox Information Center is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its monkeypox related research that is available on the Monkeypox Information Center - including this research content - immediately available in publicly funded repositories, with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the Monkeypox Information Center remains active.
The frequency of monkeypox outbreaks and the extent of the projected outbreaks in human populations have both steadily increased. This paper proposes Atangana-Baleanu fractional-order derivatives define in Caputo sense to investigate the kinetics of Monkeypox transmission in Ghana. We determine the stability of the recommended model's equilibrium points and basic reproduction number. The solution's existence and originality, as well as the model's Hyers-Ullam stability, are proven. The models basic reproduction number was found to be R0 = 0.1940. The numerical simulation showed the fractional operator had an influence on the various compartments of the model. The dynamics of the disease in the community were shown to be influenced by fractional-order derivatives, and infections were eradicated within the first five (5) days when π = 0.2.
Keywords
Monkeypox
Fractional derivative
Basic reproduction number
Atangana-Baleanu
Equilibrium points
==== Body
pmc1 Introduction
The first human case of the zoonotic orthopox DNA virus, known as monkeypox virus, was reported in the Democratic Republic of the Congo in 1970 [15]. The monkeypox virus, a member of the orthopoxvirus genus in the Poxviridae family, is the culprit behind monkeypox. Three additional human viruses are members of this genus: the vaccinia virus, the cowpox virus, and the smallpox-causing variola virus. Smallpox and monkeypox have similar clinical manifestations, with monkeypox's early-stage lymphadenopathy serving as a defining characteristic. There have never been any reports of smallpox and monkeypox epidemics coexisting. Monkeypox is a zoonotic disease that spreads from an unexplained animal reservoir to human populations, whereas smallpox is known to only affect humans. Monkeypox is occasionally introduced into human populations as a result of encounters with animal species in the woods of western and central Africa, particularly in the Republic of Congo, the Central African Republic, Nigeria, and the Democratic Republic of the Congo [16,18]. Every monkeypox outbreak was self-contained, with human transmission pathways stopping before epidemics could develop.
Monkeypox looks to be taking over as the primary pox in humans after smallpox was eradicated. The chance of an epidemic spreading to people is now thought to be low [17,18]. Humans can contract the disease through coming into intimate contact with an animal or person who has the disease, as well as by coming into contact with contaminated objects. The virus that causes monkeypox spreads from one person to another by close contact with lesions, body fluids, infected objects like bedding, and respiratory droplets [19]. There has been an evidence of human-to - human transmission of the virus in the UK [38], and Democratic Republic of Congo [37]. A report by Thornhill et al. [15], revealed that among 528 patients with disease in 16 countries, 98% of the patients were gays. A recent prospective observational cohort study in Spain showed that 166 (92%) of the 188 patients with MPX infection were either gay or bisexual men [40]. Ghana, a West African country recorded a series of the monkeypox disease on the May 24, 2022. A report from the Minister of Health indicated that the outbreak began from a man who had recently traveled from the United States to Ghana with mild symptoms of MPX disease. Since then, there has been 84 confirmed cases through human-to-human transmission of the virus and these were identified through contact tracing [20,[41], [42], [43]].
On June 23, 2022, the World Health Organization designated monkeypox as an “emerging risk of moderate public health concern.” More than 65,000 cases of monkeypox virus infection have been reported globally in 106 countries and five geographic zones since September 2022, resulting in 26 deaths. A combination of factors, including waning smallpox immunity, laxing coronavirus disease 2019 (Covid-19) prevention measures, resuming international travel, and sexual interactions associated with large gatherings, may be to blame for the current global outbreak of monkeypox virus infection in humans [15].
There have been few studies on the disease's transmission in the past [[21], [22], [23], [24], [25]]. However, mathematical models have been used to study the transmission of diseases such as COVID-19 [[34], [35], [36]] and diseases belonging to the family Poxviridae such as smallpox [[26], [27], [28]], chickenpox [29,30] and cowpox [31]. The authors of [26] proposed a mathematical model to examine the effects of case isolation and ring vaccination for epidemic containment and test the capacity of the health system under various scenarios with available interventions in order to estimate the effects of a smallpox attack in Mumbai, India. The authors of [28] created a mathematical model to explain how smallpox spreads after being intentionally released into the environment. To examine the chickenpox in Nigeria, Madaki et al. [29] proposed a deterministic SEIR model combining the method of control used by the national chickenpox and leprosy control programs. Using models of ordinary differential equations, Qureshi [30] studied the 2013 chickenpox outbreak among school children in Schenzen, China. In order to find the model with the maximum efficiency rate, three novel models with the Mittag-Leffler type kernel (Atangana-Baleanu in the Caputo sense), the exponentially decaying type kernel (Caputo-Fabrizio), and the power law type kernel (Caputo) were formulated. Somma et al. [22] developed a mathematical model to study the spread of monkeypox among rodent and human populations. According to the concept proposed in Ref. [22], the human population is provided with a quarantine class and a public awareness campaign to prevent the spread of the disease. Peter et al., [24] created and investigated a deterministic mathematical model for the transmission of the monkeypox virus between humans and rodents in Nigeria. A model was put forth by Grant et al. [25] to investigate how the monkeypox and smallpox viruses spread throughout the Democratic Republic of the Congo. To gain a better understanding of how the epidemic spread within Zambia, Kalezhi et al. [34] used a number of existing data mining methods (classifiers) available in the Waikato Environment for Knowledge Analysis (WEKA) machine learning library.
Fractional derivatives have been proposed for application in mathematical models to understand viral infections [[1], [2], [3], [4], [5], [6],9,11,33]. Because they have more degrees of freedom than integer-order models, fractional-order models are more precise and trustworthy. The fractional-order differential equation models appear to be more compatible with this condition than the integer-order ones. This is done so that fractional derivatives and integrals can be used to characterize the inherent memory and heredity properties of many materials and processes [8]. Veeresha and Prakasha [33] employed the homotopy analysis transform approach to solve the fractional generalized Zakharov (FGZ) equations using the Atangana-Baleanu fractional derivative model. To investigate the disease's spread in Nigeria, Peter et al. [11], developed both traditional and Caputo-Fabrizio fractional-order derivative models. This is the only fractional-order derivative model of the monkeypox virus. After reviewing several publications, we found that few studies on the monkeypox virus and its modes of transmission took into account the virus's ability to spread from person to person. Although there is evidence of human-to-human transmission of the virus [34, 38, 40, 43] which is the main mode of transmission of the disease in Ghana, existing models focus on rodent-to-human transmission of the disease. There is no model that predicts or forecasts the disease's spread in Ghana. In this study, we propose Atangana-Baleanu fractional-order derivatives defined in the Caputo sense to study human-to-human transmission of the virus in the country. The Atangana-Baleanu and Caputo derivatives, which have a variety of advantageous qualities, including a nonlocal and nonsingular kernel, offer a clearer understanding of the crossover behavior in the model utilizing this operator. These characteristics may or may not allow other operators, such as Caputo and Caputo-Fabrizio, to define the dynamics of the monkeypox appropriately [10].
The remaining sections of the paper are as follows: In Section 2, we formulate and study a mathematical model based on the fractional-order derivative. In Section 3, we list the model's qualitative attributes. Along with the fundamental reproduction number, we also determine equilibrium points and their stability. In Section 4, the optimal control model is examined, and the defined model includes time-dependent optimal control. Section 5 does the numerical analysis of the model. We investigate the effect of the fractional-order operator on the various compartments. We numerically analyzed the optimal control model in Section 6. In Section 7, we assess and describe the outcomes of our proposed model.
2 Model formulation
By analyzing the model in Ref. [24], we develop a mathematical model that best captures the spread of the virus from person to person, altering [24] to incorporate those who are immune to the infection. Then, Atangana-Baleanu fractional-order derivatives defined in Caputo sense are used to describe the model. The population is partitioned into six (6) compartments: Susceptible H S, exposed (H E), infected (H A), hospitalized (H Q), recovered (H R) and immune individuals H V. The total population is given as.N(t) = H S + H E + H A + H Q + H R + H V
The rate of recruitment into the susceptible class is ξ, while the rate of natural death is ξ 1. The susceptible have a ξ 2 chance of becoming infected when they get the illness from people in H A. The parameters ξ 3 and ξ 6 are the recovery rates of an infected person and those hospitalized respectively. The parameters ξ 5 and ξ 9 are the disease-induced death rate and infectious rate respectively. The rate of transition from the infected compartment to the hospitalized compartment is given by ξ 4. The flowchart of the model is shown in Fig. 1 .Fig. 1 Flowchart of the monkeypox model.
Fig. 1
The model is described by the following fractional-order derivatives.(1) DtπHS=ξπ-ξ2πHAHSN-ξ1πHS,D0,tπHE=ξ2πHAHSN-(ξ9π+ξ1π)HE,D0,tπHA=ξ9πHE-(ξ3π+ξ4π+ξ1π+ξ5π)HA,D0,tπHQ=ξ4πHA-(ξ6π+ξ5π+ξ1π)HQ,D0,tπHR=ξ6πHQ+ξ3πHA-ξ1πHRD0,tπHV=ξπη-ξ1πHV.
With initial conditions H S(0) = H S o, H E(0) = H E o, H A(0) = H A o, H Q(0) = H Q o, H R(0) = H R o, H V(0) = H V(o).
2.1 Preliminaries
We go over the definitions of the key terms used in this work and those specified in Ref. [1] in this part.Definition 2.1 Liouville and Caputo (LC) describe the fractional derivative of order π as in [1,5](2) DtππCh(t)=1Γ(1−π)∫0t(τ−q)−πh(q)dq,0<π≤1.
Definition 2.2 [1,4] provide the Liouville-Caputo sense definition of the Atangana-Baleanu fractional derivative.(3) DtππABCh(t)=G(π)(1−π)∫0tHEπ−πτ−qπ1−πh•(q)dq,
where G(π)=1−π+πΓ(π), is the normalized function.
Definition 2.3 The Atangana—Baleanu—Caputo derivative’s pertinent fractional integral is given by the definition at [1,4](4) ItππCh(t)=(1−π)G(π)h(t)+πG(π)Γ(π)∫0τ(τ−q)q−1h•(q)dq.
They calculated both derivatives’ Laplace transforms and discovered the following:(5) LDtπ0ABCh(t)=G(π)W(f)fπ−fπ−1h(0)(1−π)fπ+π1−π.
Where L is the Laplace transform operator.Theorem 2.1 For a functionh ∈ H[y 1, y 2], the following results holds [1,7]:
DtππABCf(t)<G(π)(1−π)c(t), wherec(t)=maxy1≤t≤y2c(t).
Additionally, the derivatives of Atangana, Baleanu, and Caputo satisfy the Lipschitz criterion [1,7]:(6) D0,tππABCc1(t)−DtππABCc2(t)<ωc1(t)−c2(t).
2.2 Existence and uniqueness of the solutions
This section establishes the existence and distinctiveness of the solutions to system (1).
Using the symbol Z(X) to represent a banach space, where X = [0, b], and Y = Z(X) × Z(X) × Z(X) × Z(X) × Z(X) × Z(X) × Z(X) and the given norm HS,HE,HA,HQ,HR,HV=HS+HE+HA+HQ+HR+HV, where HS=Supτ∈XHS,HE=Supτ∈XHE,HA=Supτ∈XHA,HQ=Supτ∈XHQ, HR=Supτ∈XHR,HV=Supτ∈XHV, and using the ABC integral operator on system (1) gives(7) HS(t)-HS(0)=Dtπ[HS]=ξπ-ξ2πHAHSN-ξ1πHS,HE(t)-HE(0)=D0,tπ[HE]=ξ2πHAHSN-(ξ9π+ξ1π)HE,HA(t)-HA(0)=D0,tπ[HA]=ξ9πHE-(ξ3π+ξ4π+ξ1π+ξ5π)HA,HQ(t)-HQ(0)=D0,tπ[HQ]=ξ4πHA-(ξ6π+ξ5π+ξ1π)HQ,HR(t)-HR(0)=Dtπ[HR]=ξ6πHQ+ξ3πHA-ξ1πHR,HV(t)-HV(0)=Dtπ[HV]=ξπηπ-ξ1πHV,
Definition 1 gives,(8) HS(t)-HS(0)=1-πG(π)ψ1(π,t,HS(t))+πG(π)Γ(π)×∫0t(t-r)π-1ψ1(π,r,HS(r))dr,HE(t)-HE(0)=1-πG(π)ψ2(π,t,HE(t))+πG(π)Γ(π)×∫0t(t-r)π-1ψ2(π,r,HE(r))dr,HA(t)-HA(0)=1-πG(π)ψ3(π,t,HA(t))+πG(π)Γ(π)×∫0t(t-r)π-1ψ3(π,r,HA(r))dr,HQ(t)-HQ(0)=1-πG(π)ψ4(π,t,HQ(t))+πG(π)Γ(π)×∫0t(t-r)π-1ψ4(π,r,HQ(r))dr,HR(t)-HR(0)=1-πG(π)ψ5(π,t,HR(t))+πG(π)Γ(π)×∫0t(t-r)π-1ψ5(π,r,HR(r))dr,HV(t)-HV(0)=1-πG(π)ψ6(π,t,HV(t))+πG(π)Γ(π)×∫0t(t-r)π-1ψ6(π,r,HV(r))dr,
where(9) ψ1(π,r,HS(t))=ξπ-ξ2πHAHSN-ξ1πHS,ψ2(π,r,HE(t))=ξ2πHAHSN-(ξ9π+ξ1π)HE,ψ3(π,r,HA(t))=ξ9πHE-(ξ3π+ξ4π+ξ1π+ξ5π)HA,ψ4(π,r,HQ(t))=ξ4πHA-(ξ6π+ξ5π+ξ1π)HQ,ψ5(π,r,HR(t))=ξ6πHQ+ξ3πHA-ξ1πHRψ6(π,r,HV(t))=ξπηπ-ξ1πHV.
Additionally, the Atangana-Baleanu in Caputo derivatives only meets the Lipschitz requirement [7] if H S(t), H E(t), H A(t), H Q(t), H R(t) and H V(t) have an upper bound. Assuming that H S(t) and H S*(t) are pair functions,(10) ψ1(π,t,HS(t))−ψ1(π,t,HS*(t))=−ξ2πHAHSN+ξ1πHS(HS(t)−HS*(t)).
Considering(11) F1=−ξ2πHAHSN+ξ1πHS.
Equation (10) simplifies to(12) ψ1(π,t,HS(t))−ψ1(π,t,HS*(t))≤F1(HS(t)−HS*(t)).
Similarly,(13) ψ2(π,t,HE(t))−ψ2(π,t,HE*(t))≤F2(HE(t)−HE*(t)),ψ3(π,t,HA(t))−ψ3(π,t,HA*(t))≤F3(HA(t)−HA*(t))ψ4(π,t,HQ(t))−ψ4(π,t,HQ*(t))≤F4(HQ(t)−HQ*(t)),ψ5(π,t,HR(t))−ψ5(π,t,HR*(t))≤F5(HR(t)−HR*(t)),ψ6(π,t,HV(t))−ψ6(π,t,HE*(t))≤F6(HV(t)−HV*(t)),
where(14) F2=−(ξ9π+ξ1π)HE,F3=−(ξ3π+ξ4π+ξ1π+ξ5π)HA,F4=−(ξ6π+ξ5π+ξ1π)HQ,F5=−ξ1πHR,F6=−ξ1πHV.
Lipschitz's condition is thus valid. Now, repeatedly applying system (8) results in(15) HSn(t)-HS(0)=1-πB(π)ψ1(π,t,HSn-1(t))+πB(π)Γ(π)×∫0t(t-r)π-1ψ1(π,r,HSn-1(r))dr,HEn(t)-HE(0)=1-πB(π)ψ2(π,t,HEn-1(t))+πB(π)Γ(π)×∫0t(t-r)π-1ψ1(π,r,HEn-1(r))dr,HAn(t)-HA(0)=1-πB(π)ψ3(π,t,HAn-1(t))+πB(π)Γ(π)×∫0t(t-r)π-1ψ3(π,r,HAn-1(r))dr,HQn(t)-HQ(0)=1-πB(π)ψ4(π,t,HQn-1(t))+πB(π)Γ(π)×∫0t(t-r)π-1ψ4(π,r,HQn-1(r))dr,HRn(t)-HR(0)=1-πB(π)ψ5(π,t,HRn-1(t))+πB(π)Γ(π)×∫0t(t-r)π-1ψ5(π,r,HRn-1(r))dr,HVn(t)-HV(0)=1-πB(π)ψ1(π,t,HVn-1(t))+πB(π)Γ(π)×∫0t(t-r)π-1ψ6(π,r,HVn-1(r))dr,
with the initial conditions HS(0)=HS0,HE(0)=HE0,HA(0)=HA0,HQ(0)=HQ0,HR(0)=HR0,HV(0)=HV0.
Difference of consecutive terms yields(16) ΨHSn(t)=HSn(t)-HSn-1(t)=1-πB(π)(ψ1(π,t,HSn-1(t))-ψ1(π,t,HSn-2(t)))+πB(π)Γ(π)∫0t(t-r)π-1(ψ1(π,r,HSn-1(r))-ψ1(π,r,HSn-2(r)))dr,ΨHEn(t)=HEn(t)-HEn-1(t)=1-πB(π)(ψ2(π,t,HEn-1(t))-ψ2(π,t,HEn-2(t)))+πB(π)Γ(π)∫0t(t-r)π-1(ψ2(π,r,HEn-1(r))-ψ2(π,r,HEn-2(r)))dr,ΨHAn(t)=HAn(t)-HAn-1(t)=1-πB(π)(ψ3(π,t,HAn-1(t))-ψ3(π,t,HAn-2(t)))+πB(π)Γ(π)∫0t(t-r)π-1(ψ3(π,r,HAn-1(r))-ψ3(π,r,HAn-2(r)))dr,ΨHQn(t)=HQn(t)-HQn-1(t)=1-πB(π)(ψ4(π,t,HQn-1(t))-ψ4(π,t,HQn-2(t)))+πB(π)Γ(π)∫0t(t-r)π-1(ψ4(π,r,HQn-1(r))-ψ4(π,r,HQn-2(r)))dr,ΨHRn(t)=HRn(t)-HRn-1(t)=1-πB(π)(ψ5(π,t,HRn-1(t))-ψ5(π,t,HRn-2(t)))+πB(π)Γ(π)∫0t(t-r)π-1(ψ5(π,r,HRn-1(r))-ψ5(π,r,HRn-2(r)))dr,ΨHVn(t)=HVn(t)-HVn-1(t)=1-πB(π)(ψ6(π,t,HVn-1(t))-ψ6(π,t,HVn-2(t)))+πB(π)Γ(π)∫0t(t-r)π-1(ψ6(π,r,HVn-1(r))-ψ6(π,r,HRn-2(r)))dr
where HSn(t)=∑i=0nΨHSn(t),HEn(t)=∑i=0nΨHEn(t),HAn(t)=∑i=0nΨHAn(t),HQn(t)=∑i=0nΨHQn(t),HRn(t)=∑i=0nΨHRn(t),HVn(t)=∑i=0nΨHVn(t),.
Taking into consideration equation (12), (13) and considering ΨHSn−1(t)=HSn−1(t)−HSn−2(t),ΨHEn−1(t)=HEn−1(t)−HEn−2(t),ΨHAn−1(t)=HAn−1(t)−HAn−2(t),ΨHQn−1(t)=HQn−1(t)−HQn−2(t), ΨHRn−1(t)=HRn−1(t)−HRn−2(t), ΨHVn−1(t)=HVn−1(t)−HVn−2(t) (17) ΨHSn(t)≤1−πB(π)F1ΨHSn−1(t)πB(π)Γ(π)F1×∫0t(t−r)π−1ΨHSn−1(r)dr,ΨHEn(t)≤1−πB(π)F2ΨHEn−1(t)πB(π)Γ(π)F2×∫0t(t−r)π−1ΨHEn−1(r)dr,ΨHAn(t)≤1−πB(π)F3ΨHAn−1(t)πB(π)Γ(π)F3×∫0t(t−r)π−1ΨHAn−1(r)dr,ΨHQn(t)≤1−πB(π)F4ΨHQn−1(t)πB(π)Γ(π)F4×∫0t(t−r)π−1ΨHQn−1(r)dr,ΨHRn(t)≤1−πB(π)F5ΨHRn−1(t)πB(π)Γ(π)F5×∫0t(t−r)π−1ΨHRn−1(r)dr,ΨHVn(t)≤1−πB(π)F6ΨHVn−1(t)πB(π)Γ(π)F6×∫0t(t−r)π−1ΨHVn−1(r)dr,
Theorem 2.2 system (1) has a unique solution fort ∈ [0, b] subject to the condition 1−πB(π)Fi+πB(π)Γ(π)bπni<1,i=1,2,3,……,6 holds [44].
Proof:
Since H S(t), H E(t), H A(t), H Q(t), H R(t) and H_V(t) are bounded functions and Equation (12), (13) holds. In a recurring manner (17) reaches(18) ΨHSn(t)≤HS0(t)1−πB(π)F1+πbπB(π)Γ(π)F1n,ΨHEn(t)≤HE0(t)1−πB(π)F2+πbπB(π)Γ(π)F2n,ΨHAn(t)≤HA0(t)1−πB(π)F3+πbπB(π)Γ(π)F3n,ΨHQn(t)≤HQ0(t)1−πB(π)F4+πbπB(π)Γ(π)F4n,ΨHRn(t)≤HR0(t)1−πB(π)F5+πbπB(π)Γ(π)F5n,ΨHVn(t)≤HV0(t)1−πB(π)F6+πbπB(π)Γ(π)F6n,
and.
ΨHSn(t)→0,ΨHEn(t)→0,ΨHAn(t)→0,ΨHQn(t)→0,ΨHRn(t)→0,ΨHVn(t)→0.
Incorporating the triangular inequality and for any j, system (18) yields(19) HSn+j(t)−HSn(t)≤∑i=n+1n+jF1j=F1n+1−F1n+m+11−F1,HEn+j(t)−HEn(t)≤∑i=n+1n+jF2j=F2n+1−F2n+m+11−F2,HAn+j(t)−HAn(t)≤∑i=n+1n+jF3j=F3n+1−F3n+m+11−F3,HQn+j(t)−HQn(t)≤∑i=n+1n+jF4j=F4n+1−F4n+m+11−F4,HRn+j(t)−HRn(t)≤∑i=n+1n+jF5j=F5n+1−F5n+m+11−F5,HVn+j(t)−HVn(t)≤∑i=n+1n+jF6j=F6n+1−F6n+m+11−F6,
where Fi=1−πB(π)Fi+πB(π)Γ(π)bπFi<1.
Hence there exists unique solution for system (1).
2.3 Hyers –ulam stability
Definition 2.4 Atangana-Baleanu fractional derivative system (1) is said to be Hyers-Ulam stable if constants ℏ i < 0, i ∈ N 5 matching the following conditions exist for any ω¯i>0,i∈N5 (20) HS(t)-1-πG(π)Ψ1(π,t,HS(t))+πG(π)Γ(π)×∫0t(t-r)π-1Ψ1(π,r,HS(r))dr≤ω¯1,HE(t)-1-πG(π)Ψ2(π,t,HE(t))+πG(π)Γ(π)×∫0t(t-r)π-1Ψ2(π,r,HE(r))dr≤ω¯2,HA(t)-1-πG(π)Ψ3(π,t,HA(t))+πG(π)Γ(π)×∫0t(t-r)π-1Ψ3(π,r,HA(r))dr≤ω¯3,HQ(t)-1-πG(π)Ψ4(π,t,HQ(t))+πG(π)Γ(π)×∫0t(t-r)π-1Ψ4(π,r,HQ(r))dr≤ω¯4,HR(t)-1-πG(π)Ψ5(π,t,HR(t))+πG(π)Γ(π)×∫0t(t-r)π-1Ψ5(π,r,HR(r))dr≤ω¯5,HV(t)-1-πG(π)Ψ6(π,t,HV(t))+πG(π)Γ(π)×∫0t(t-r)π-1Ψ6(π,r,HV(r))dr≤ω¯6,
and there exist {H•S(t),H•E(t),H•A(t),H•Q(t),H•R(t)H•V(t)},
where(21) H•S(t)=1−πG(π)Ψ1(π,t,HS(t))+πG(π)Γ(π)×∫0t(t−r)π−1Ψ1(π,r,H•S(r))dr,H•E(t)=1−πG(π)Ψ2(π,t,HE(t))+πG(π)Γ(π)×∫0t(t−r)π−1Ψ2(π,r,H•E(r))dr,H•A(t)=1−πG(π)Ψ3(π,t,HA(t))+πG(π)Γ(π)×∫0t(t−r)π−1Ψ3(π,r,H•A(r))dr,H•Q(t)=1−πG(π)Ψ4(π,t,HQ(t))+πG(π)Γ(π)×∫0t(t−r)π−1Ψ4(π,r,H•Q(r))dr,H•R(t)=1−πG(π)Ψ5(π,t,HR(t))+πG(π)Γ(π)×∫0t(t−r)π−1Ψ5(π,r,H•R(r))dr,H•V(t)=1−πG(π)Ψ6(π,t,HV(t))+πG(π)Γ(π)×∫0t(t−r)π−1Ψ6(π,r,H•V(r))dr,
such that.
HS(t)−H•S(t)≤μ1ω¯1,HE(t)−H•E(t)≤μ2ω¯2,HA(t)−H•A(t)≤μ3ω¯3,HQ(t)−H•Q(t)≤μ4ω¯4,HR(t)−H•R(t)≤μ5ω¯5,HV(t)−H•V(t)≤μ6ω¯6.
3 Equilibrium point and local stability
3.1 Equilibrium points
We examine the equilibrium points/steady states of the model in this section. The disease-free equilibrium (DFE) and the endemic equilibrium (EE) are the two steady states of system (1). When there is no infection in the population, or when H A = H Q = 0, the steady state solution is disease-free equilibrium. Solving system (1) after equating the right side of the system to zero results in(22) DFE=(HSo,HEo,HAo,HQo,HRo,HVo)=ξπξ1π,0,0,0,0,ξπηπξ1π
The endemic equilibrium point, EE=(HS*,HE*,HA*,HQ*,HR*,HV*) is(23) HS*=ξπNξ2πHA*+ξ1πN,HE*=ξ2πHA*HS*(ξ9π+ξ1π)N,HA*=ξ9πHE*(ξ3π+ξ4π+ξ1π+ξ5π),HQ*=ξ4πHA*(ξ6π+ξ5π+ξ1π),HR*=ξ6πHQ*+ξ3πHA*ξ1π,HV*=ξπηπξ1π.
The basic reproduction number is the total number of secondary cases that a single infected person might cause during the course of the infection, in a susceptible population [1]. It is a crucial factor that determines whether or not the disease will spread throughout a population. The infected compartments in this model are H E, H A and H Q. Denote F and V, respectively, as the right-hand side of system (1) corresponding to the infected compartments using the next-generation operator method [1].
dzdt=F(z)−V(z)
Where.
F=ξ2πHAHSNξ9πHEξ4πHA and. V=(ξ9π+ξ1π)HE(ξ3π+ξ4π+ξ1π)HA(ξ6π+ξ5π+ξ1π)Q
The matrix F(x) and V(x) calculated at the equilibrium point of disease-free is given as.
F=0ξ2πHSoN0ξ9π000ξ4π0 andV=ξ9π+ξ1π000ξ3π+ξ4π+ξ1π+ξ5π000(ξ6π+ξ5π+ξ1π)
FV−1=0ξ2πξ3π+ξ4π+ξ1π+ξ5π0ξ9πξ+ξ1π000ξ4πξ3π+ξ4π+ξ1π+ξ5π0
The basic reproductive number is the largest positive eigenvalue of FV −1 and is given as(24) R0=ξ2πξ9π(ξ3π+ξ4π+ξ1π+ξ5π)(ξ9π+ξ1π).
3.2 Local stability of the disease-free equilibrium
The following theorem establishes the necessary condition for the local stability of the disease-free steady state.Theorem 3.1 In the event that it exists, the disease-free equilibrium is locally asymptotically unstable forR0 > 1 and stable for R 0 < 1.
The Jacobian matrix of system (1), evaluated at the disease - free equilibrium point is given as(25) J=−ξ1π0−ξ2π0000−(ξ9π+ξ1π)ξ2π0000ξ9π−(ξ3π+ξ4π+ξ1π+ξ5π)00000ξ4π−(ξ6π+ξ5π+ξ1π)0000ξ3πξ6π−ξ1π000000−ξ1π.
We must demonstrate the negative real components of each and every eigenvalue of system (25). The first four (4) eigenvalues are φ 1,2,3 = −ξ 1, and φ4=−(ξ6π+ξ5π+ξ1π). The remaining ones are derived from the sub-matrix (26), which is created by leaving off the first, fourth, fifth, and sixth rows and columns of system (25). Thus, we have(26) JEO=−(ξ9π+ξ1π)ξ2πξ9π−(ξ3π+ξ4π+ξ1π+ξ5π).
System (26) characteristic equation, is given as(27) φ2+f1φ+f2=0,
wheref1=ξ3π+ξ4π+ξ5π+ξ9π+2ξ1πf2=(ξ3π+ξ4π+ξ5π+ξ1π)(ξ9π+ξ1π)1−R02.
The stability of the aforementioned characteristic equation is then assessed using the Routh—Hurwitz stability criterion. According to the Routh-Hurwitz stability criterion, if both the f 1 > 0 and f 2 > 0 conditions are satisfied, all of the characteristic equation's roots have negative real portions, indicating a stable equilibrium. When R 0 < 1, the coefficients f 1 > 0 and f 2 > 0 are clearly greater than zero. As a result, our equilibrium is secure.
3.3 Local stability of the endemic equilibrium point
The endemic equilibrium point is a stable, positive situation in which the disease is still present in the populace.Theorem 3.2 The model has a unique endemic equilibrium point if the following Routh-Hurwitz conditions are satisfied:g1 > 0, g 3 > 0 and g 1 g 2 > g 3 [32]
Proof. At the endemic equilibrium point (4), the Jacobian matrix is(28) JE*=J110J13000J21J22J230000ξ9πJ3300000ξ4πJ440000ξ3πξ6πξ1π000000ξ1π,
whereJ11=−ξ2πHA*N−ξ1π,J33=−(ξ3π+ξ6π+ξ1π+ξ5π),J13=−ξ2πHS*NJ21=ξ2πHA*N,J23=ξ2πHS*N,J44=−(ξ6π+ξ5π+ξ1π),.
The three eigenvalues of the Jacobian matrix (28) are represented by the diagonal elements J 44 and ξ1π repeated roots. The remaining values are generated by removing the fourth, fifth, and sixth columns and rows of (28). This results in(29) JE*=J110J13J21J22J230ξ9πJ33
The characteristic equation of Jacobian (29) is given as(30) Ω3+g1Ω2+g2Ω+g3=0
Where(31) g1=−(J33+J11+J22)g2=J33(J11+J22)+J11J22−ξ9πJ23g3=ξ9π(J11J23−J13J21)−J11J22J33
If the requirements g 1 > 0, g 3 > 0, and g 1 g 2 > g 3 are met, then the characteristic equation (30) has negative real roots, which denotes a stable equilibrium according to the Routh—Hurwitz stability criterion.
4 Numerical analysis
This section validates the fractional-order monkeypox model using Ghana's demographic information, and published parameter values. The parameter values are given in Table 1 .Table 1 Parameter description and values.
Table 1Parameter Description Value, Year−1 Source
ξ2 Contact rate between infected human and susceptible human 0.022325 [11]
ξ Recruitment rate 29.08 [12, 13]
ξ6 The rate at which critically ill individuals recovers 0.036246 [11]
ξ3 The rate at which infected individuals recovers due to natural immunity 0.088366 [11]
ξ4 The rate at which infected individuals become critically ill 0.5 [11]
ξ9 The rate at which the exposed becomes infectious 0.016744 [11]
ξ5 Monkeypox disease-induced death rate 0.003286 [11]
ξ1 Natural mortality rate 0.4252912 × 10−4 [12]
η immunity rate 0.1 [11]
When the parameter values given in Table 1 are used, the result of the basic reproduction number is R 0 = 0.1940, demonstrating that the disease is not endemic in Ghana.
On May 24, 2022, when the first five cases of monkeypox are confirmed by Ghanaian authorities [14], we start our simulation. The population of Ghana is 30.8 million, according to the 2021 Population and Housing Census [12]. The total population under study, N = 30.8 million, is taken into account in our model, which is in agreement. H S(0) = 30,799,995, H E(0) = 0, H A(0) = 5, H Q(0) = 0,H R(0) = 0, H V(0)=0, are the initial conditions we chose. The results of the simulation are displayed in Fig. 2, Fig. 3, Fig. 4, Fig. 5, Fig. 6, Fig. 7, Fig. 8 , which depicts the dynamic behavior of the susceptible, exposed, infected, hospitalized, deceased, recovered and immune individuals respectively for the period of 100 days.Fig. 2 Dynamics of the susceptible compartment.
Fig. 2
Fig. 3 Dynamics of the exposed compartment.
Fig. 3
Fig. 4 Dynamics of the infected compartment.
Fig. 4
Fig. 5 Dynamics of the hospitalized compartment.
Fig. 5
Fig. 6 Dynamics of the deceased.
Fig. 6
Fig. 7 Dynamics of the recovered compartment.
Fig. 7
Fig. 8 Dynamics of the immune individuals.
Fig. 8
4.1 Discussion
In Fig. 2, Fig. 3, Fig. 4, Fig. 5, Fig. 6, Fig. 7, Fig. 8, are reported, the solutions of system (1) for 5 different values of π ∈ [0, 1] at step-size 0.2 for a period of 100 days. The decrease in the susceptible population is directly proportional to a reduction in the value of the fractional operator π. There is an early peak in the number of exposed and hospitalized individuals as the fractional operator value is reduced (see Fig. 3, Fig. 5). The number of infected individuals is seen to decay within the first 5 days, and this is directly related to the operator value (see Fig. 4). The number of people who die, on the other hand, increases as π decreases (see Fig. 6). The recovered data exhibits an early peak with a fall in the operator value and also shows the crossover effect of the model (see Fig. 7). The immune individuals increases with a decrease in the value of π (see Fig. 8).
5 Conclusion
This work seeks to examine the transmission of the monkeypox disease among humans in a community. Since contacts from human-to-rodents are very rare in the Country. This study examines the spread of the monkeypox virus within a society using the fractional-order derivative described in the Atangana-Baleanu in Caputo sense. The new model includes individuals immune to the virus and also report on the dynamics of the deceased which available model fails to do so. We investigated the qualitative characteristics of the model, including its basic reproduction number, equilibrium points, and equilibrium point stability. Along with Hyers-Ulam stability, the solutions' existence and distinctiveness were demonstrated. Based on the basic reproduction number R 0 = 0.1940, it was determined that the disease was not endemic. The numerical simulation revealed that, the fractional operator had an impact on the model's distinct compartments. Classical models (π = 1.0) exhibited fewer deaths, however, this wasn't the case as the fractional operator was reduced. The dynamics of the infected compartment were shown to be influenced by fractional-order derivatives, and infections were eradicated within the first five (5) days when π = 0.2. Control measures could also help curb the disease and prevent future occurrences. Hence, control measures are recommended for future studies.
Author statement
The contribution of each author regarding the manuscript “Modeling and Analysis of Monkeypox using Fractional Derivatives” are as follows: Samuel Okyere: Conceptualization, Methodology, Formal analysis, Writing - Original Draft, Software, Joseph Ackora Prah: Conceptualization, Methodology, Supervision, Proofreading, Revision.
Funding
The research did not receive funding from any sources
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Data availability
No data was used. All parameter values are duly cited and referenced
==== Refs
References
1 Sweilam N.H. Al-Mekhlafi S.M. Baleanu D. Optimal control for a fractional tuberculosis infection model including the impact of diabetes and resistant strains J. Adv. Res. 17 2019 125 137 10.1016/j.jare.2019.01.007 ISSN 2090-1232 31193340
2 Ullah S. Khan M.A. Farooq M. Hammouch Z. Baleanu D. A fractional model for the dynamics of tuberculosis infection using Caputo-Fabrizio derivative Am. Inst. Math. Sci. 13 3 2020 975 993 10.3934/dcdss.2020057
3 Shaikh A.S. Shaikh I.N. Nisar K.S. A mathematical model of COVID-19 using fractional derivative: outbreak in India with dynamics of transmission and control Adv. Differ. Equ. 373 2020 10.1186/s13662-020-02834-3
4 Atangana A. Baleanu D.D. New fractional derivatives with nonlocal and nonsingular kernels: theory and application to the heat transfer model Therm. Sci. 20 2 2016 763 769 10.2298/TSCI160111018A
5 Moghaddam B.P. Yaghoobi S. Machado J.T. An extended predictor-corrector algorithm for variable-order fractional delay differential equations J. Comput. Nonlinear Dynam. 1 2016 1 11 10.1115/1.4032574
6 Naik P.A. Yavuz M. Qureshi S. Zu J. Townley S. Modeling and analysis of COVID-19 epidemics with treatment in fractional derivatives using real data from Pakistan Eur. Phys. J. Plus 135 10 2020 795 10.1140/epjp/s13360-020-00819-5 33145145
7 Coll C. Herrero A. Sanchez E. Thome N. A dynamic model for a study of diabetes Math. Comput. Model. 50 5–6 2009 713 716 10.1016/j.mcm.2008.12.027
8 Sweilam N.H. Al-Mekhlafi S.M. Baleanu D. A hybrid fractional optimal control for a novel Coronavirus (2019-nCov) mathematical model J. Adv. Res. 32 2021 149 160 10.1016/j.jare.2020.08.006 32864171
9 Habenom H. Aychluh M. Suthar D.L. Al-Mdallal Q. Purohit S.D. Modeling and analysis on the transmission of covid-19 Pandemic in Ethiopia Alex. Eng. J. 61 7 2022 5323 5342 10.1016/j.aej.2021.10.054 ISSN 1110-0168
10 Khan M.A. Atangana A. Modeling the dynamics of novel coronavirus (2019-nCov) with fractional derivative Alex. Eng. J. 59 4 2020 2379 2389 10.1016/j.aej.2020.02.033 ISSN 1110-0168
11 Peter O.J. Oguntolu F.A. Ojo M.M. Abdulmumin O. Oyeniyi A.O. Jan R. Khan I. Fractional order mathematical model of monkeypox transmission dynamics Phys. Scripta 97 8 2022 084005 10.1088/1402-4896/ac7ebc
12 Okyere S. Ackora-Prah J. A mathematical model of transmission dynamics of SARS-CoV-2 (COVID-19) with an underlying condition of diabetes Int. J. Math. Math. Sci. 2022 2022 10.1155/2022/7984818 Article ID 7984818, 15 pages
13 Indexmundi Ghana Birth Rate 2021 www.indexmundi.com accessed 11th December, 2021
14 Nyabor J. Monkeypox Cases in Ghana Rise to 18 2022 https://citinewsroom.com/2022/06/monkeypox-cases-in-ghana-rise-to-18/
15 Thornhill J.P. Barkati S. Walmsley S. Rockstroh Juergen Antinori Andrea Harrison Luke B. Palich Romain Nori Achyuta Reeves Iain Habibi Maximillian S. Apea Vanessa Boesecke Christoph Vandekerckhove Linos Yakubovsky Michal Sendagorta Elena Blanco Jose L. Florence Eric Moschese Davide Maltez Fernando M. Goorhuis Abraham Pourcher Valerie Migaud Pascal Noe Sebastian Pintado Claire Maggi Fabrizio Hansen Ann-Brit E. Hoffmann Christian Lezama Jezer I. Mussini Cristina Cattelan AnnaMaria Makofane Keletso Tan Darrell Nozza Silvia Nemeth Johannes Klein Marina B. Orkin Chloe M. Monkeypox virus infection in humans across 16 countries — april–june 2022 N. Engl. J. Med. 387 8 2022 679 691 10.1056/NEJMoa2207323 35866746
16 Sklenovská N. Van Ranst M. Emergence of monkeypox as the most important orthopoxvirus infection in humans Front. Public Health 4 6 2018 241 10.3389/fpubh.2018.00241
17 Reed K.D. Melski J.W. Graham M.B. Regnery R.L. Sotir M.J. Wegner M.V. The detection of monkeypox in humans in the Western Hemisphere N. Engl. J. Med. 350 4 2004 342 350 10.1056/NEJMoa032299 14736926
18 Grant R. Nguyen L.L. Breban R. Modelling human-to-human transmission of monkeypox Bull. World Health Organ. 98 9 2020 638 640 10.2471/BLT.19.242347 Epub 2020 Jul 8. PMID: 33012864; PMCID: PMC7463189 33012864
19 World Health Organization Monkeypox 2022 https://www.who.int/news-room/fact-sheets/detail/monkeypox
20 Centers for Disease Prevention and Control Monkeypox” 2022 https://www.cdc.gov/poxvirus/monkeypox/response/2022/world-map.html
21 Peter O.J. Oguntolu F.A. Ojo M.M. Oyeniyi A.O. Jan R. Khan I. Fractional order mathematical model of monkeypox transmission dynamics Phys. Scripta 97 2022 Number 8 https://iopscience.iop.org/article/10.1088/1402-4896/ac7ebc/pdf
22 Somma S.A. Akinwade N.I. Chado U.D. A mathematical model of monkey pox virus transmission dynamics Ife J. Sci. AJOA 21 1 2019 elSSN: 0794-4896 https://www.ajol.info/index.php/ijs/article/view/185244
23 Usman S. Adamu I. Modeling the transmission dynamics of the monkeypox virus infection with treatment and vaccination interventions J. Appl. Math. Phys. 5 12 2017 https://www.scirp.org/journal/papercitationdetails.aspx?paperid=81078&JournalID=2436
24 Peter O.J. Kumar S. Kumari N. Oguntolu F.A. Oshinubi K. Musa R. Transmission dynamics of Monkeypox virus: a mathematical modelling approach Modeling Earth Systems and Environment, 1–12 2021 Advance online publication 10.1007/s40808-021-01313-2
25 Grant R. Nguyen L.L. Breban R. Modelling human-to-human transmission of monkeypox Bull. World Health Organ. 98 9 2020 638 640 10.2471/BLT.19.242347 33012864
26 Mohanty B. Costantino V. Narain J. Modelling the impact of a smallpox attack in India and influence of disease control measures BMJ Open 10 2020 e038480 10.1136/bmjopen-2020-038480
27 Bacaër N. Daniel Bernoulli, d'Alembert and the inoculation of smallpox (1760) A Short History of Mathematical Population Dynamics 2011 Springer London 10.1007/978-0-85729-115-8_4
28 Meltzer M.I. Damon I. LeDuc J.W. Millar J.D. Modeling potential responses to smallpox as a bioterrorist weapon Emerg. Infect. Dis. 7 6 2001 959 969 10.3201/eid0706.010607 2001 11747722
29 Madaki Y.U. Manu H.A. Gwani A.A. Edeghagba E.E. Application of mathematical modelling on the spread of chicken pox disease (A case study of Nayinawa clinic Damaturu, Yobe state) Scholars J. Phys. Math. Statis. 7 2020 260 271 10.36347/sjpms.2020.v07i10.004
30 Qureshi S. Abdullahi Yusuf A. Modeling chickenpox disease with fractional derivatives: from caputo to atangana-baleanu 2020 Chaos, Solit. Fractals 122 2019 111 118 10.1016/j.chaos.2019.03.020 ISSN 0960-0779
31 Johnson R.F. Yellayi S. Cann J.A. Johnson A. Smith A.L. Paragas J. Jahrling P.B. Blaney J.E. Cowpox virus infection of cynomolgus macaques as a model of hemorrhagic smallpox Virology 418 Issue 2 2011 102 112 10.1016/j.virol.2011.07.013 ISSN 0042-6822 21840027
32 Okyere S. Ackora-Prah J. A mathematical model of transmission dynamics of SARS-CoV-2 (COVID-19)” with an underlying condition of diabetes Int. J. Math. Math. Sci. 2022 2022 10.1155/2022/7984818 Article ID 7984818, 15 pages
33 Veeresha P. Prakasha D.G. Solution for fractional generalized Zakharov equations with Mittag-Leffler function Res. Eng. 5 2020 100085 10.1016/j.rineng.2019.100085 2590-1230
34 Kalezhi J. Chibuluma M. Chembe C. Chama V. Lungo F. Kunda D. Modelling Covid-19 infections in Zambia using data mining techniques Res. Eng. 13 2022 100363 10.1016/j.rineng.2022.100363 2590-1230
35 O'Connor S. Mathew S. Dave F. Tormey D. Parsons U. Gavin M. Nama P.M. Moran R. Rooney M. McMorrow R. Bartlett J. Pillai S.C. COVID-19: rapid prototyping and production of face shields via flat, laser-cut, and 3D-printed models Res. Eng. 14 2022 100452 10.1016/j.rineng.2022.100452 2590-1230
36 Narayan S.S. Saha S. Bhattacharjee A. ”A 2D FSI mathematical model of blood flow to analyze the hyper-viscous effects in atherosclerotic COVID patients” Res. Eng. 12 2021 2021 10.1016/j.rineng.2021.100275 100275, ISSN 2590-1230
37 Nolen L. Osadebe L. Katomba J. Likofata J. Mukadi D. Monroe B. Extended human-to-human transmission during a monkeypox outbreak in the Democratic Republic of the Congo Emerg. Infect. Dis. 22 6 2016 1014 1021 10.3201/eid2206.150579 27191380
38 Vaughan A. Aarons E. Astbury J. Brooks T. Chand M. Flegg P. Hardman A. Harper N. Jarvis R. Mawdsley S. McGivern M. Morgan D. Morris G. Nixon G. O'Connor C. Palmer R. Phin N. Price D.A. Russell K. Said B. Dunning J. Human-to-Human transmission of monkeypox virus, United Kingdom, october 2018 Emerg. Infect. Dis. 26 4 2020 782 785 10.3201/eid2604.191164 32023204
40 Tarín-Vicente E.J. Alemany A. Agud-Dios M. Clinical presentation and virological assessment of confirmed human monkeypox virus cases in Spain: a prospective observational cohort study Lancet 400 2022 661 669 10.1016/S0140-6736(22)01436-2 35952705
41 Newsroom Citi Greater Accra Has More than 50% of Monkeypox Cases 2022 https://citinewsroom.com/2022/07/greater-accra-has-more-than-50-of-monkeypox-cases/
42 Adadi P. Mensah E.O. Abdul-Razak S. Letter to the Editor: the outbreak of monkeypox (MPX) in Ghana Journal of Medical Virology 2022 Wiley Online Library 10.1002/jmv.28171
43 Awuni F. Monkeypox: One dead in Upper East, 13 contacts being monitored – GHS 2022 https://citinewsroom.com/2022/07/monkeypox-one-dead-in-upper-east-13-contacts-being-monitored-ghs/
44 Bonyah E. Sagoe A.K. Kumar D. Deniz S. Fractional optimal control dynamics of coronavirus model with Mittag–Leffler law Ecol. Complex. 45 100880 2021 1476 10.1016/j.ecocom.2020.100880 945X
| 36467285 | PMC9705013 | NO-CC CODE | 2022-12-09 23:14:54 | no | Results Eng. 2023 Mar 29; 17:100786 | utf-8 | Results Eng | 2,022 | 10.1016/j.rineng.2022.100786 | oa_other |
==== Front
Am J Physiol Heart Circ Physiol
Am J Physiol Heart Circ Physiol
AJPHEART
American Journal of Physiology - Heart and Circulatory Physiology
0363-6135
1522-1539
American Physiological Society Rockville, MD
36367689
H-00578-2022
H-00578-2022
10.1152/ajpheart.00578.2022
Rapid Report
Cardiovascular Consequences of COVIDSARS-CoV-2 infection downregulates myocardial ACE2 and potentiates cardiac inflammation in humans and hamsters
MECHANISMS OF CARDIAC INJURY IN COVID-19
Viveiros Anissa 1 2
Noyce Ryan S. 3 4
Gheblawi Mahmoud 1
https://orcid.org/0000-0001-8607-2950
Colombo Daniele 5
Bilawchuk Leanne M. 3 4
Clemente-Casares Xavier 4
Marchant David J. 3 4
https://orcid.org/0000-0002-9357-0912
Kassiri Zamaneh 1
https://orcid.org/0000-0001-6664-9257
Del Nonno Franca 5
Evans David H. 3 4
https://orcid.org/0000-0002-9154-9028
Oudit Gavin Y. 1 2 6
1Department of Physiology, Faculty of Medicine and Dentistry, University of Alberta, Edmonton, Alberta, Canada
2Mazankowski Alberta Heart Institute, University of Alberta, Edmonton, Alberta, Canada
3Li Ka Shing Institute of Virology, University of Alberta, Edmonton, Alberta, Canada
4Department of Medical Microbiology and Immunology, University of Alberta, Edmonton, Alberta, Canada
5Pathology Unit, IRCCS Istituto Nazionale per le Malattie Infettive “Lazzaro Spallanzani”, Rome, Italy
6Division of Cardiology, Department of Medicine, Faculty of Medicine and Dentistry, University of Alberta, Edmonton, Alberta, Canada
Correspondence: G. Y. Oudit ([email protected]).
1 12 2022
11 11 2022
11 11 2022
323 6 H1262H1269
12 10 2022
2 11 2022
3 11 2022
Copyright © 2022 the American Physiological Society.
2022
American Physiological Society
Myocardial pathologies resulting from SARS-CoV-2 infections are consistently rising with mounting case rates and reinfections; however, the precise global burden is largely unknown and will have an unprecedented impact. Understanding the mechanisms of COVID-19-mediated cardiac injury is essential toward the development of cardioprotective agents that are urgently needed. Assessing novel therapeutic strategies to tackle COVID-19 necessitates an animal model that recapitulates human disease. Here, we sought to compare SARS-CoV-2-infected animals with patients with COVID-19 to identify common mechanisms of cardiac injury. Two-month-old hamsters were infected with either the ancestral (D614) or Delta variant (B.1.617.2) of SARS-CoV-2 for 2 days, 7 days, and/or 14 days. We measured viral RNA and cytokine expression at the earlier time points to capture the initial stages of infection in the lung and heart. We assessed myocardial angiotensin-converting enzyme 2 (ACE2), the entry receptor for the SARS-CoV-2 virus, and cardioprotective enzyme, as well as markers for inflammatory cell infiltration in the hamster hearts at days 7 and 14. In parallel, human hearts were stained for ACE2, viral nucleocapsid, and inflammatory cells. Indeed, we identify myocardial ACE2 downregulation and myeloid cell burden as common events in both hamsters and humans infected with SARS-CoV-2, and we propose targeting downstream ACE2 downregulation as a therapeutic avenue that warrants clinical investigation.
NEW & NOTEWORTHY Cardiac manifestations of COVID-19 in humans are mirrored in the SARS-CoV-2 hamster model, recapitulating myocardial damage, ACE2 downregulation, and a consistent pattern of immune cell infiltration independent of viral dose and variant. Therefore, the hamster model is a valid approach to study therapeutic strategies for COVID-19-related heart disease.
ACE2
; COVID-19
; heart
; inflammation
Gouvernement du Canada | Canadian Institutes of Health Research (CIHR) 10.13039/100000024 PJT-451105 Call for PapersTrue
==== Body
pmcINTRODUCTION
Cardiovascular complications in severe COVID-19, namely, microvascular dysfunction, myocarditis, conduction abnormalities, and heart failure, represent an emerging global health crisis. Furthermore, cardiovascular injury resulting from SARS-CoV-2, the causative virus of COVID-19, is complicated by the postacute sequelae as also linked to an increased incidence of inflammatory heart disease and thrombotic disorders; thus, cardiac injury in COVID-19 is likely currently underestimated (1). Reports of myocardial damage and the fact that angiotensin-converting enzyme 2 (ACE2), the indispensable viral entry receptor, is expressed in the heart and other affected organs suggest a probable mechanism of direct infection (1, 2). Therefore, in light of this public health emergency, it is of paramount importance to delineate the pathological mechanisms of myocardial inflammation and injury in COVID-19 and to discover novel cardioprotective interventions.
Studying SARS-CoV-2-mediated myocardial injury requires animal models to test therapeutic interventions, a feat challenged by the limited ability of small animal models to adequately recapitulate this human disease. Generally, mouse models are studied extensively and are readily available; however, they have limited utility for COVID-19 studies since murine ACE2 does not effectively bind to the SARS-CoV-2 spike protein (3). Strategies to overcome this limitation have been developed, including modifying the viral spike protein to bind mouse ACE2 or developing humanized mice that express the human ACE2 protein (3). Despite these approaches, these mice can develop additional symptoms following SARS-CoV-2 infection, such as lethal encephalitis in humanized mice that are not captured in humans; yet, they do not develop cardiac symptoms. In addition, extensive manipulation of animals and the viral spike protein limits translational significance (3). Alternatively, the Syrian hamster model superficially resembles human COVID-19 disease, with low mortality, is readily available, and succumbs to infection by unmodified SARS-CoV-2; therefore, is a valuable tool for COVID-19 research (4). To date, however, it is unclear if hamsters exhibit similar cardiac manifestations as in human patients with COVID-19.
Here, we provide evidence that the Syrian hamster is a suitable animal model to study the cardiovascular manifestations of COVID-19, thus providing a means of studying therapeutic interventions to prevent or treat myocardial injury in patients with COVID-19.
MATERIALS AND METHODS
Cells and Viruses
Vero cells (ATCC CCL-81) were maintained in minimum essential medium (MEM) supplemented with 100 U/mL penicillin, 100 U/mL streptomycin, 0.25 µg/mL amphotericin B, and 10% fetal bovine serum. An ancestral (D614) SARS-CoV-2 strain (GISAID No. EPI_ISL_425177) and a SARS-CoV-2-Delta variant of concern strain (B.1.617.2) were used in these studies.
In Vivo Hamster Infections
Animal experiments were approved by the University of Alberta Animal Care and Use Committee (AUP00001847 and AUP00003869). All SARS-CoV-2 infection studies were conducted in a certified BSL3 containment facility at the University of Alberta. Briefly, 2-mo-old male Syrian hamsters were inoculated intranasally with the ancestral SARS-CoV-2 variant at a dose of 2.0 × 103 PFU in a total volume of 100 µL (50 µL per nare). Nasal swabs were performed on days 1, 3, and 6 after challenge and collected for histopathology on day 7 or day 14 postinfection. To determine cytokine responses following virus infection, another set of hamsters were inoculated at a viral dose of 1.0 × 106 PFU with SARS-CoV-2-Delta (B.1.617.2) and heart and lung tissues were collected on day 2 and day 7. Control hamsters were inoculated with MEM containing no virus. Animals were randomized to either the infected or control groups. The hamsters were monitored daily for signs of infection and morbidity (Fig. 1A). Animals who lost greater than 20% of their initial body weight were to be humanely euthanized and excluded from the study; however, no animals exceeded this threshold.
Figure 1. SARS-CoV-2 infection induces lung injury and upregulates cytokine expression in hamsters. A: experimental timeline for Delta (collected at days 2 and 7 for RT-PCR) and ancestral (collected at days 7 and 14 for histopathology) SARS-CoV-2-infected hamsters. Vehicle-inoculated control hamsters were analyzed in parallel. B: anthropometric data of body weight and lung weight of ancestral SARS-CoV-2-infected hamsters. Body weight was measured and recorded daily and reflected as percent of weight change relative to day 0. Time points were compared using two-way repeated-measures ANOVA and Sidak’s multiple comparisons test. Lungs were harvested and weighed at the end of the 14-day study period. C: SARS-CoV-2 nasal swabs and plaque measurement D: gross histopathological assessment of hamster lungs using hematoxylin and eosin staining (H&E) and Masson’s trichrome staining at day 7 after ancestral SARS-CoV-2 infection E: staining for SARS-CoV-2 nucleocapsid protein in the lungs at day 7 after viral challenge F: SARS-CoV-2 viral RNA copies in the lung following SARS-CoV-2 Delta challenge G: expression of Il-1β, Tnf-α, Il-6, and Il-10 in hamster lungs at day 2 and day 7 following Delta SARS-CoV-2 infection. Cytokines are visualized as a relative expression compared with controls. Data are represented as means ± SE, and each point represents biological replicates (n = 4 hamsters/group). Unpaired Student’s t test was performed for comparisons of controls to SARS-CoV-2-treated animals. One-way ANOVA with Dunnett’s multiple comparisons test or Kruskal Wallis test with Dunn’s multiple comparisons were used to compare parametric or nonparametric data, respectively; *P < 0.05, **P < 0.01.
Virus Titrations
For virus culture, 2 × 105 cells were seeded into each well of a 12-well tissue culture plate 1 day before titration. Tenfold serial dilutions of the virus stock or nasal swab were plated in duplicate on Vero CCL-81 cells and cultured for 3 days at 37°C in MEM containing 0.5% carboxymethylcellulose (Sigma). Cells were fixed and stained with a solution containing 0.13% (wt/vol) crystal violet, 11% formaldehyde (vol/vol), and 5% ethanol (vol/vol) to visualize plaques.
Human Samples
Patients who succumbed to COVID-19 from the Lazio region were autopsied in Rome, Italy. In parallel, hearts were procured from age- and sex-matched donors following cardioplegic arrest according to the Human Organ Procurement and Exchange (HOPE) protocol. Control hearts were obtained from brain-dead donors (DBD) with no known history of cardiovascular disease. Transmural myocardial sections were formalin fixed and paraffin embedded for histological analysis. All protocols are approved by the Health Research Ethics Board of the University of Alberta.
Histological Staining
Formalin-fixed paraffin-embedded tissues were sectioned onto slides at 5-µm thickness. Hematoxylin and eosin (H&E) staining was performed according to a standard protocol. Briefly, slides were dewaxed and rehydrated by decreasing alcohol gradient. Nuclei were stained with Harris hematoxylin for 15 s, rinsed, and differentiated with 1% acid alcohol. Sections were rinsed in Scott’s tap water substitute (20 mM sodium bicarbonate, 166 mM magnesium sulfate) and then stained with eosin. Sections were dehydrated, cleared, and mounted. Masson’s trichrome staining was performed using a commercially available kit (Abcam ab150686). Briefly, deparaffinized and rehydrated slides were incubated in preheated Bouin’s fluid, cooled, and rinsed. The sections were next stained in Weigert’s iron hematoxylin working solution (equal parts of solutions A and B) and rinsed under tap water. The slides were next stained in Biebrich Scarlet/Acid Fuchsin solution and washed with distilled water. Slides were differentiated in phosphomolybdic/phosphotungstic acid solution until collagen was no longer red and transferred directly to aniline blue solution. Slides were rinsed and then differentiated in 1% acetic acid. Slides were quickly dehydrated, cleared in xylene, and mounted with a resinous mounting medium. Images were captured with a Leica DM4000 B LED microscope system.
Immunohistochemical Staining
Immunohistochemical (IHC) staining was performed on formalin-fixed paraffin-embedded tissues sectioned at 5-µm thickness and then dewaxed and rehydrated by ethanol gradient. Heat-induced epitope retrieval was achieved with preheated sodium citrate buffer, consisting of 10 mM sodium citrate and 0.05% Tween 20 (pH 6.0). Slides were blocked with 10% goat serum in 1% bovine serum albumin (BSA) and incubated with primary antibodies diluted in 1% BSA for ACE2 (1:50, R&D Systems AFF933), SARS-CoV-2 nucleocapsid (1:1,000, Bioss bs-41408R), CD15 (1:50, Abcam ab135377), CD68 [1:100, Thermo Fisher MA5-13324 (human); 1:100, AbD Serotec MCA1597 (hamster)], CD4 [1:100, Abcam ab133616 (human); 1:50, Millipore Sigma MABF415 (hamster)], and CD8 (1:50, Biolegend 200702). Antibody specificity was validated by staining control hamsters and human hearts, which do not have immune cell infiltration or virus. Antibody specificity for ACE2 was validated previously (5). Endogenous peroxidases were blocked with 10% H2O2, and slides was then incubated with horseradish peroxidase (HRP)-conjugated secondary antibodies (1:1,000, Cell Signaling Technologies) and visualized with freshly prepared 3,3′-diaminobenzidine substrate (Abcam). Slides were counterstained with regressive Harris hematoxylin, differentiated with 1% acid alcohol, then dehydrated by alcohol gradient, cleared, and mounted with organic mounting media. Staining was quantified as a percent area that the staining occupies over the total area of the image [staining area (%)] and as a count of positive staining regions per square millimeter (count/mm2). Imaging and quantification were performed blinded to the experimental groups.
RT-PCR
RNA was extracted from the lung and left ventricle of control and SARS-CoV-2-Delta-infected hamsters using TRIzol-chloroform. cDNA was reverse transcribed from 1 µg of RNA template using SuperScript II Reverse Transcriptase (Invitrogen). Real-time quantitative polymerase chain reaction (RT-PCR) with TaqMan premixed assays (ThermoFisher Scientific) and 25 ng of cDNA was used to quantify Ace2 (Cg04585346_m1), Tnf-α (Cg04607188_g1), Il-1β (Cg04576706_g1), Il-6 (Cg04486380_m1), Il-10 (Cg04628513_m1) with Hprt (Cg04448432_m1) as the housekeeping gene. Viral RNA copy number was quantified using the 2019-nCoV RUO kit (Integrated DNA Technologies) and a serial dilution of the 2019-nCoV_N positive control (Integrated DNA Technologies). Analyses of 2.5 ng of cDNA for the lung and 25 ng of cDNA for the heart were done.
Protein Extraction
Immunoblot was performed from protein samples extracted in TRIzol following removal of the aqueous phase for RNA extraction. DNA was precipitated from the organic phase with one third volume of ethanol and then centrifuged at 2,000 g at 4°C for 5 min. Proteins were precipitated from the supernatant with ispropanol (1:1) and then centrifuged at 12,000 g at 4°C. The pellet was washed with 0.3 M guanidine hydrochloride in 95% ethanol and centrifuged at 12,000 g, then repeated, and then washed in ethanol by the same procedure. The supernatant was decanted, and the protein pellet was dried under vacuum and then resolubilized in 500 µL of CelLytic buffer (Sigma-Aldrich) supplemented with protease inhibitors (Roche). Samples were sonicated on ice using a tipultrasonicator (Sonic Dismembrator Model 100, Fisher Scientific) set to level 2 in 4 × 20-s bursts with 30-s intervals.
Immunoblot
Extracted proteins were quantified with DC Protein Assay (BioRad), and 90 µg of protein was resolved by SDS-PAGE and then transferred to polyvinylidene fluoride membranes in transfer buffer, consisting of 25 mM Tris, 192 mM glycine, and 20% methanol (pH 8.3). Membranes were blocked in 5% nonfat milk and incubated with ACE2 (1:1,000, Abcam ab108252) primary antibody overnight and subsequently detected with HRP-conjugated secondary antibodies (1:4,000, Cell Signaling Technology) and Clarity ECL substrate (Bio-Rad). Lanes were normalized to MemCode total protein stain (Thermo Fisher Scientific). Band densitometry was quantified with Image Studio Software (LI-COR Biosciences).
Statistical Analysis
Statistics were performed with SPSS software, and graphs were created with GraphPad Prism. Data are represented as means ± SE. Unpaired Student’s t test was performed for comparisons of controls to patients with COVID-19. In comparisons exceeding two groups, cases that followed normal distribution were subject to one-way ANOVA with Dunnett’s multiple comparisons tests, and nonparametric data sets were subject to Kruskal—Wallis test with Dunn’s multiple comparisons.
RESULTS
Anthropometric assessment of SARS-CoV-2 infection in hamsters demonstrated weight loss compared with baseline over the course of the study and increased lung weight at the end of the 14-day period following ancestral SARS-CoV-2 infection (Fig. 1B). Nasal swabs of control and ancestral SARS-CoV-2-infected animals were measured on days 1, 3, and 6 to assay viral burden (Fig. 1C). Histologically, infected hamster lungs revealed multifocal parenchymal damage and thickening of the alveolar septa, indicating diffuse alveolar damage (DAD) (Fig. 1D). Trichrome staining corroborated alveolar thickening; however, staining did not demonstrate substantive parenchymal fibrosis (Fig. 1D). SARS-CoV-2 viral nucleocapsid staining was absent in control hamster lungs; however, SARS-CoV-2-treated hamsters displayed positive perivascular staining for viral proteins (Fig. 1E). To examine earlier time points, we next analyzed control and Delta SARS-CoV-2-infected hamsters at days 2 and 7, which revealed positive amplification for SARS-CoV-2 viral RNA (Fig. 1F) and increased expression of Il-1β, Tnf-α, and Il-10 in lung homogenates (Fig. 1G).
We next examined myocardial changes in the hamster heart consequent to SARS-CoV-2 infection. Infected hamsters have infrequent regions of increased mononuclear infiltrates and focal fibrosis that was absent in control animals (Fig. 2A). Furthermore, the viral nucleocapsid was detected in the heart in ancestral SARS-CoV-2-infected animals (Fig. 2B). This finding aligns with the positive amplification of SARS-CoV-2 viral RNA copies in the heart of SARS-CoV-2 Delta-infected animals, supporting myocardial infection as strain and dose independent (Fig. 2C). Following a basic assessment of myocardial damage and discovering SARS-CoV-2 viral proteins and RNA in the hamster heart, we aimed to investigate the consequence of infection on ACE2 levels. We first assessed the ancestral SARS-CoV-2 by histology, which demonstrated a reduction in ACE2 that persisted at both time intervals (days 7 and 14) (Fig. 2D). Interestingly, when we examined early time points, Ace2 expression was unchanged between control and Delta SARS-CoV-2-infected animals (Fig. 2E); however, ACE2 protein levels were significantly reduced compared with controls as in the ancestral SARS-CoV-2-infected animals (Fig. 2F). Because of the prominent myocardial inflammation present in certain COVID-19 cases (6), we next aimed to identify and quantify the immune cells of the heart following infection. Staining for mature neutrophils (CD15), macrophages (CD68), and T-cell antigens (CD4 and CD8) demonstrated a macrophage- and neutrophil-dominant pattern, with a mild increase in lymphocyte staining compared with that in control animals (Fig. 2, G and H). Congruently, expression of proinflammatory cytokines was mildly elevated in the heart, namely, Il-1β, and a trend toward an increase of Tnf-α (Fig. 2I).
Figure 2. SARS-CoV-2 infection downregulates myocardial angiotensin-converting enzyme 2 (ACE2) and induces immune cell infiltration in hamsters. A: histological assessment of hamster hearts using hematoxylin and eosin staining (H&E) and Masson’s trichrome staining 7 days following ancestral SARS-CoV-2 challenge B: SARS-CoV-2 nucleocapsid staining in the hamster heart at 14 days after ancestral SARS-CoV-2 infection C: SARS-CoV-2 viral RNA copies in the heart following SARS-CoV-2 Delta challenge. D: representative images and pooled analysis of ACE2 staining in the hamster heart at day 7 (empty squares) and day 14 (filled squares) following inoculation with ancestral SARS-CoV-2. ACE2 staining area (%) and the number of areas with positive staining (count/mm2) are quantified. E: Ace2 mRNA expression by RT-PCR in the hamster heart at day 2 and day 7 after Delta SARS-CoV-2 infection. F: Western blot analysis and quantification of immunoreactivity (band densitometry for protein levels) of ACE2 following Delta SARS-CoV-2 inoculation. Immunoblots were visualized at a standard exposure (STD) or overexposed (OE) to visualize low protein levels of ACE2. Band densitometry was quantified and normalized to MemCode total protein stain (MEM). G: representative immune cell staining for neutrophils (CD15), macrophages (CD68), and T cells (CD4 and CD8) in the hearts of vehicle (control) and ancestral SARS-CoV-2-inoculated hamsters at day 14, and immune cell quantification (H). Day 7 (empty squares) and day 14 (filled squares) are pooled for analysis. I: expression of Tnf-α and Il-1β in hamster heart at day 2 and day 7 following Delta SARS-CoV-2 infection. Il-6 and Il-10 were below the limit of detection in the heart. Cytokines are visualized as a relative expression compared with controls. Data are represented as means ± SE, and each point represents biological replicates (n = 4–6 hamsters/group). Unpaired Student’s t test was performed for comparisons of controls to SARS-CoV-2-treated animals. One-way ANOVA with Dunnett’s multiple comparisons test or Kruskal Wallis test with Dunn’s multiple comparisons was used to compare parametric or nonparametric data, respectively; *P < 0.05, **P < 0.01.
We next compared the hamster phenotype with myocardial samples of patients who died from COVID-19 (Table 1). Basic histological staining revealed extensive mononuclear infiltrates and interstitial and vascular fibrosis (Fig. 3A). Viral nucleocapsid staining was predominantly perivascular in the myocardium (Fig. 3B). ACE2 was significantly downregulated in hearts of patient with COVID-19 compared with sex- and age-matched controls (Fig. 3C). Finally, the myocardium from patients with COVID-19 demonstrated a significant inflammatory cell burden biased toward neutrophils and macrophages consistent with a pathological diagnosis of myocarditis in 60% of the autopsied hearts (Fig. 3, D and E; Table 1).
Figure 3. Severe COVID-19 leads to myocardial angiotensin-converting enzyme 2 (ACE2) downregulation and immune cell infiltration in humans. A: routine histological assessment of human control and COVID-19 hearts with hematoxylin and eosin staining (H&E) and Masson’s trichrome staining. B and C: SARS-CoV-2 nucleocapsid and ACE2 staining in the human heart. ACE2 staining area (%) and the number of areas with positive staining (count/mm2) are quantified. D: immune cell staining for neutrophils (CD15), macrophages (CD68), and T cells (CD4 and CD8) in the hearts of control donors and patients with COVID-19, and immune cell quantification (E). Data are represented as means ± SE, and each point represents individual control donors (n = 3–4) or COVID-19 patients (n = 8–10). Unpaired Student’s t test was performed for comparisons of controls to patients with COVID-19; *P < 0.05, **P < 0.01.
Table 1. Patient clinical characteristics
Control COVID-19
Demographics
n 6 10
Age, yr 54.3 [47–67] 69.2 [44–86]
Sex, male 3 (50.0) 7 (70.0)
Comorbidities, n (%)
Preexisting cardiac conditions
Dilated cardiomyopathy 0 (0) 1 (10.0)
Ischemic cardiomyopathy 0 (0) 2 (20.0)
Hypertension 1 (16.7) 2 (20.0)
Diabetes mellitus 0 (0) 1 (10.0)
Obesity 1 (16.7) 0 (0)
CKD 0 (0) 2 (20.0)
COPD 0 (0) 2 (20.0)
Acute injury, n (%)
Myocarditis/pericarditis 0 (0) 6 (60.0)
Arrhythmia 0 (0) 4 (40.0)
Coagulopathy 0 (0) 6 (60.0)
AKI 0 (0) 8 (80.0)
ARDS 0 (0) 10 (100)
Continuous variables are reported by mean with the range in brackets: age. Categorical variables are reported by count with percentage in parenthesis: sex, comorbidities, and diagnoses.
Obesity is defined as a BMI ≥ 30 kg/m2. AKI, acute kidney injury; ARDS, acute respiratory distress syndrome; BMI, body mass index; CKD, chronic kidney disease; COPD, chronic obstructive pulmonary disease.
DISCUSSION
Our data corroborate pulmonary histological findings in hamsters that mirror findings from human patients with COVID-19 (7), with a common mechanism of myocardial ACE2 downregulation that aligns with previous work from the SARS-CoV epidemic (8). Although humans had a greater abundance of immune cell infiltrates, this difference likely reflects the enhanced severity of humans who died of COVID-19 compared with animals that survived and recovered. Nevertheless, the pattern of immune cell burden coincided in humans and hamsters, suggesting SARS-CoV-2 infection facilitates a consistent macrophage- and neutrophil-dominant recruitment. These findings are unexpected, as cell-mediated immune responses driven by T lymphocytes are implicated in typical cases of viral myocarditis, such as coxsackievirus B-mediated myocardial inflammation (6). It remains unclear if SARS-CoV-2 predominately mediates direct cardiac injury because of ACE2 tropism or if myocarditis is consequent of indirect, cytokine-activated cardiotoxicity. However, we provide evidence of SARS-CoV-2 viral nucleocapsid and positive viral RNA amplification in the lungs and heart independent of viral strain and dose that appears preferentially localized to vessels. This corroborates the detectable myocardial SARS-CoV-2 viral load in the hearts of patients with COVID-19 (6) and work supporting a direct viral infection of the myocardium in a subset of postmortem autopsy hearts in patients with SARS, which demonstrated increased fibrosis, inflammation, and a reduction of myocardial ACE2 (8). Furthermore, the localization of viral nucleocapsid and ACE2 is supported by single-nucleus RNA sequencing studies that identify pericytes, vascular smooth muscle cells, and fibroblasts as harboring the greatest ACE2 expression in the human left ventricle (9).
Predominant myeloid recruitment aligns with tangential evidence emerging from the COVID-19 pandemic. Interstitial macrophages loaded with viral particles reside in the myocardium of a patient with COVID-19-mediated cardiogenic shock (10), and macrophages are the primary cell for SARS-CoV viral replication (11). Consistently, macrophage-derived cytokine interleukin 6 (IL-6) is part of a significant predictive signature of COVID-19 mortality (12). Although Il-6 was below the limit of detection in hamster lung and myocardial tissue, we observed the paradoxical upregulation of anti-inflammatory cytokine Il-10 that was found to predict increased disease severity (13).
As concomitant cardiovascular disease (CVD) in patients with COVID-19 is a risk factor for severe disease, SARS-CoV-2 itself mediates myocardial damage, and SARS-CoV-2 exploits ACE2 for cellular entry, it is essential to delineate the role of ACE2 in CVD and COVID-19 (14–16). ACE2 is protective in CVD to counteract the proinflammatory, hypertensive arm of the canonical renin-angiotensin system (RAS) mediated by angiotensin II (ANG II). Specifically, ACE2 deactivates ANG II to Ang-(1–7), a peptide that acts on the Mas receptor to promote vasodilation and anti-inflammatory effects (1, 14). A disintegrin and metalloprotease 17 (ADAM17) facilitates proteolytic ectodomain shedding of membrane-bound proteins, including ACE2; thus, deactivating and releasing it into the systemic circulation (17, 18). Indeed, ADAM17 is aberrantly activated in CVD and may lead to a deficiency in membrane ACE2, which is associated with worsened outcomes in hypertension, heart failure, and coronary artery disease (1).
SARS-CoV-2 also activates ADAM17 to foster ACE2 loss (19), which is suggested to be critically involved in COVID-19 pathogenesis as coronaviruses that cause the common cold do not activate ADAM17 (19). We previously proposed that increased ACE2 in healthy, aged males may confer susceptibility to SARS-CoV-2 infection (20); however, the downstream consequence of SARS-CoV-2 is a reduction of membrane ACE2 (measured as an increase in plasma ACE2) as a result of ADAM17 activity (1, 19). Consistently, progressive elevation in soluble plasma ACE2 in an intraindividual serial sampling of hospitalized patients with COVID-19 predicted cumulative mortality, similarly with increased surrogate markers of ADAM17 activity (1). This highlights the double-edged sword of ACE2 in COVID-19, as higher ACE2 may increase SARS-CoV-2 viral load in the initial stages of infection, where viral load is positively associated with disease severity (20); however, postacute ADAM17 activity and viral endocytosis promote ACE2 proteolytic shedding (1). This suggests ADAM17 inhibition is likely a promising therapeutic target in SARS-CoV-2 infection to circumvent loss of protective membrane-bound ACE2.
Taken together, our comparative study demonstrates that hamsters exhibit similar downstream effects of SARS-CoV-2 infection as human patients with COVID-19, recapitulating myocardial damage, ACE2 downregulation, and a consistent pattern of immune cell infiltration. Therefore, ACE2 is a double-edged sword in COVID-19, such that increased ACE2 may enhance infection susceptibility in the initial stages, yet maintaining tissue levels of cardioprotective ACE2 will likely ameliorate myocardial injury. Critical to resolving the COVID-19 pandemic aftermath is screening novel therapeutic strategies; thus, the hamster model provides a means to target both the immune cell burden and loss of membrane ACE2—two mechanisms that drive disease pathogenesis.
GRANTS
This study was supported by Canadian Institutes of Health Research Grant PJT-451105.
DISCLOSURES
Z. Kassiri is an editor of American Journal of Physiology-Heart and Circulatory Physiology and was not involved and did not have access to information regarding the peer-review process or final disposition of this article. An alternate editor oversaw the peer-review and decision-making process for this article. None of the other authors had any conflicts of interest, financial or otherwise, to disclose.
AUTHOR CONTRIBUTIONS
X.C-C., D.J.M., Z.K., F.D.N., D.H.E., and G.Y.O. conceived and designed research; A.V., R.S.N., M.G., D.C., L.M.B., and F.D.N. performed experiments; A.V., M.G., D.C., Z.K., and F.D.N. analyzed data; A.V., X.C-C., D.J.M., Z.K., D.H.E., and G.Y.O. interpreted results of experiments; A.V. and R.S.N. prepared figures; A.V. drafted manuscript; A.V., R.S.N., D.C., L.M.B., X.C-C., D.J.M., Z.K., F.D.N., D.H.E., and G.Y.O. edited and revised manuscript; A.V., R.S.N., M.G., D.C., L.M.B., X.C-C., D.J.M., Z.K., F.D.N., D.H.E., and G.Y.O. approved final version of manuscript.
==== Refs
REFERENCES
1. Wang K, Gheblawi M, Nikhanj A, Munan M, MacIntyre E, O'Neil C, Poglitsch M, Colombo D, Del Nonno F, Kassiri Z, Sligl W, Oudit GY. Dysregulation of ACE (angiotensin-converting enzyme)-2 and renin-angiotensin peptides in SARS-CoV-2 mediated mortality and end-organ injuries. Hypertension 79 : 365–378, 2022. doi:10.1161/HYPERTENSIONAHA.121.18295. 34844421
2. Chen L, Li X, Chen M, Feng Y, Xiong C. The ACE2 expression in human heart indicates new potential mechanism of heart injury among patients infected with SARS-CoV-2. Cardiovasc Res 116 : 1097–1100, 2020 [Erratum in Cardiovasc Res 116: 1994, 2020]. doi:10.1093/cvr/cvaa078. 32227090
3. Munoz-Fontela C, Dowling WE, Funnell SGP, Gsell PS, Riveros-Balta AX, Albrecht RA, et al . Animal models for COVID-19. Nature 586 : 509–515, 2020. doi:10.1038/s41586-020-2787-6. 32967005
4. Chan JF-W, Zhang AJ, Yuan S, Poon VK-M, Chan CC-S, Lee AC-Y, Chan W-M, Fan Z, Tsoi H-W, Wen L, Liang R, Cao J, Chen Y, Tang K, Luo C, Cai J-P, Kok K-H, Chu H, Chan K-H, Sridhar S, Chen Z, Chen H, To KK-W, Yuen K-Y. Simulation of the clinical and pathological manifestations of Coronavirus Disease 2019 (COVID-19) in a golden Syrian hamster model: implications for disease pathogenesis and transmissibility. Clin Infect Dis 71 : 2428–2446, 2020. doi:10.1093/cid/ciaa325.32215622
5. Huang J, Hume AJ, Abo KM, Werder RB, Villacorta-Martin C, Alysandratos KD, Beermann ML, Simone-Roach C, Lindstrom-Vautrin J, Olejnik J, Suder EL, Bullitt E, Hinds A, Sharma A, Bosmann M, Wang R, Hawkins F, Burks EJ, Saeed M, Wilson AA, Mühlberger E, Kotton DN. SARS-CoV-2 infection of pluripotent stem cell-derived human lung alveolar type 2 cells elicits a rapid epithelial-intrinsic inflammatory response. Cell Stem Cell 27 : 962–973.e7, 2020. doi:10.1016/j.stem.2020.09.013. 32979316
6. Tschöpe C, Ammirati E, Bozkurt B, Caforio ALP, Cooper LT, Felix SB, Hare JM, Heidecker B, Heymans S, Hubner N, Kelle S, Klingel K, Maatz H, Parwani AS, Spillmann F, Starling RC, Tsutsui H, Seferovic P, Van Linthout S. Myocarditis and inflammatory cardiomyopathy: current evidence and future directions. Nat Rev Cardiol 18 : 169–193, 2021. doi:10.1038/s41569-020-00435-x. 33046850
7. Angeles Montero-Fernandez M, Pardo-Garcia R. Histopathology features of the lung in COVID-19 patients. Diagn Histopathol (Oxf) 27 : 123–127, 2021. doi:10.1016/j.mpdhp.2020.11.009. 33312229
8. Oudit GY, Kassiri Z, Jiang C, Liu PP, Poutanen SM, Penninger JM, Butany J. SARS-coronavirus modulation of myocardial ACE2 expression and inflammation in patients with SARS. Eur J Clin Invest 39 : 618–625, 2009. doi:10.1111/j.1365-2362.2009.02153.x. 19453650
9. Tucker NR, Chaffin M, Bedi KC Jr, Papangeli I, Akkad AD, Arduini A, Hayat S, Eraslan G, Muus C, Bhattacharyya RP, Stegmann CM; Human Cell Atlas Lung Biological Network; Margulies KB, Ellinor PT; Human Cell Atlas Lung Biological Network Consortium Members. Myocyte-specific upregulation of ACE2 in cardiovascular disease: implications for SARS-CoV-2-mediated myocarditis. Circulation 142 : 708–710, 2020. 32795091
10. Tavazzi G, Pellegrini C, Maurelli M, Belliato M, Sciutti F, Bottazzi A, Sepe PA, Resasco T, Camporotondo R, Bruno R, Baldanti F, Paolucci S, Pelenghi S, Iotti GA, Mojoli F, Arbustini E. Myocardial localization of coronavirus in COVID-19 cardiogenic shock. Eur J Heart Fail 22 : 911–915, 2020. doi:10.1002/ejhf.1828. 32275347
11. Cheung CY, Poon LLM, Ng IHY, Luk W, Sia S-F, Wu MHS, Chan K-H, Yuen K-Y, Gordon S, Guan Y, Peiris JSM. Cytokine responses in severe acute respiratory syndrome coronavirus-infected macrophages in vitro: possible relevance to pathogenesis. J Virol 79 : 7819–7826, 2005. doi:10.1128/JVI.79.12.7819-7826.2005. 15919935
12. Del Valle DM, Kim-Schulze S, Huang H-H, Beckmann ND, Nirenberg S, Wang B, Lavin Y, Swartz TH, Madduri D, Stock A, Marron TU, Xie H, Patel M, Tuballes K, Van Oekelen O, Rahman A, Kovatch P, Aberg JA, Schadt E, Jagannath S, Mazumdar M, Charney AW, Firpo-Betancourt A, Mendu DR, Jhang J, Reich D, Sigel K, Cordon-Cardo C, Feldmann M, Parekh S, Merad M, Gnjatic S. An inflammatory cytokine signature predicts COVID-19 severity and survival. Nat Med 26 : 1636–1643, 2020. doi:10.1038/s41591-020-1051-9. 32839624
13. Han H, Ma Q, Li C, Liu R, Zhao L, Wang W, Zhang P, Liu X, Gao G, Liu F, Jiang Y, Cheng X, Zhu C, Xia Y. Profiling serum cytokines in COVID-19 patients reveals IL-6 and IL-10 are disease severity predictors. Emerg Microbes Infect 9 : 1123–1130, 2020. doi:10.1080/22221751.2020.1770129. 32475230
14. Viveiros A, Rasmuson J, Vu J, Mulvagh SL, Yip CYY, Norris CM, Oudit GY. Sex differences in COVID-19: candidate pathways, genetics of ACE2, and sex hormones. Am J Physiol Heart Circ Physiol 320 : H296–H304, 2021. doi:10.1152/ajpheart.00755.2020. 33275517
15. South AM, Diz DI, Chappell MC. COVID-19, ACE2, and the cardiovascular consequences. Am J Physiol Heart Circ Physiol 318 : H1084–H1090, 2020. doi:10.1152/ajpheart.00217.2020. 32228252
16. Hoffmann M, Kleine-Weber H, Schroeder S, Kruger N, Herrler T, Erichsen S, Schiergens TS, Herrler G, Wu NH, Nitsche A, Müller MA, Drosten C, Pöhlmann S. SARS-CoV-2 cell entry depends on ACE2 and TMPRSS2 and is blocked by a clinically proven protease inhibitor. Cell 181 : 271–280.e8, 2020. doi:10.1016/j.cell.2020.02.052. 32142651
17. Patel VB, Clarke N, Wang Z, Fan D, Parajuli N, Basu R, Putko B, Kassiri Z, Turner AJ, Oudit GY. Angiotensin II induced proteolytic cleavage of myocardial ACE2 is mediated by TACE/ADAM-17: a positive feedback mechanism in the RAS. J Mol Cell Cardiol 66 : 167–176, 2014. doi:10.1016/j.yjmcc.2013.11.017.24332999
18. Lambert DW, Yarski M, Warner FJ, Thornhill P, Parkin ET, Smith AI, Hooper NM, Turner AJ. Tumor necrosis factor-alpha convertase (ADAM17) mediates regulated ectodomain shedding of the severe-acute respiratory syndrome-coronavirus (SARS-CoV) receptor, angiotensin-converting enzyme-2 (ACE2). J Biol Chem 280 : 30113–30119, 2005. doi:10.1074/jbc.M505111200. 15983030
19. Haga S, Yamamoto N, Nakai-Murakami C, Osawa Y, Tokunaga K, Sata T, Yamamoto N, Sasazuki T, Ishizaka Y. Modulation of TNF-alpha-converting enzyme by the spike protein of SARS-CoV and ACE2 induces TNF-alpha production and facilitates viral entry. Proc Natl Acad Sci USA 105 : 7809–7814, 2008. doi:10.1073/pnas.0711241105. 18490652
20. Viveiros A, Gheblawi M, Aujla PK, Sosnowski DK, Seubert JM, Kassiri Z, Oudit GY. Sex- and age-specific regulation of ACE2: insights into severe COVID-19 susceptibility. J Mol Cell Cardiol 164 : 13–16, 2022. doi:10.1016/j.yjmcc.2021.11.003. 34774871
| 36367689 | PMC9705018 | NO-CC CODE | 2022-12-09 23:25:55 | no | Am J Physiol Heart Circ Physiol. 2022 Dec 1; 323(6):H1262-H1269 | utf-8 | Am J Physiol Heart Circ Physiol | 2,022 | 10.1152/ajpheart.00578.2022 | oa_other |
==== Front
Am J Infect Control
Am J Infect Control
American Journal of Infection Control
0196-6553
1527-3296
Association for Professionals in Infection Control and Epidemiology, Inc. Published by Elsevier Inc.
S0196-6553(22)00707-6
10.1016/j.ajic.2022.09.023
Commentary
Journal Club: Remote infection control assessments in long-term care facilities during COVID-19 pandemic in Texas, 2020
Merrill Katreena PhD, RN, CIC, FAPIC a⁎
Piatek Dana MPH, MSN, RN, CIC, FAPIC b
Hebden Joan MS, RN, CIC, FAPIC, FSHEA c
a Brigham Young University
b Pennsylvania Department of Health
c University of Maryland School of Medicine, Department of Epidemiology and Public Health, Baltimore, MD
⁎ Address correspondence to Katreena Merrill PhD, RN, CIC, FAPIC, Brigham Young University
29 11 2022
12 2022
29 11 2022
50 12 13981400
© 2022 Association for Professionals in Infection Control and Epidemiology, Inc. Published by Elsevier Inc. All rights reserved.
2022
Association for Professionals in Infection Control and Epidemiology, Inc.
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
==== Body
pmcJOURNAL CLUB REVIEW: REMOTE INFECTION CONTROL ASSESSMENTS IN LONG TERM CARE FACILITIES SUMMARY
This article, Remote Infection Control Assessments in Long-term Care Facilities (LTCF) during COVID-19 Pandemic in Texas, 2020,1 reported the results of remote infection control assessments (tele-ICARs) performed predominately in nursing homes and/or skilled nursing facilities (NH/SNF) and assisted living facilities (ALF) in both a proactive and responsive fashion by the Texas Department of State Health Services (DSHS) during the SARS-CoV-2 pandemic. The results from 438 LTCFs tele-ICARs conducted using a standardized assessment tool during the first 8 months of the COVID-19 pandemic (March – October 2020) were analyzed.
The primary objective was to identify differences in infection prevention and control (IPC) knowledge and practices across LTCF types and inform educational efforts for these facilities. An additional objective was to determine the feasibility of conducting tele-ICARs. Findings from proactive and outbreak responsive tele-ICARs include gaps in the proper use of Centers for Disease Control and Prevention (CDC) guidelines for hand hygiene, disinfectant use and education, communal dining and activities, and compliance auditing for hand hygiene and personal protective equipment (PPE) practices. LTCFs with responsive tele-ICARs reported a significant difference in preference for alcohol-based sanitizer to soap and water hand hygiene, had suspended communal dining and group activities and had designated cohort isolation space when compared to LTCFs who had proactive tele-ICARs. Proactive tele-ICARs performed prior to identified COVID-19 cases were less likely to have dedicated space to house or cohort residents with infection. Significantly more ALF than NH/SNF had not suspended communal dining activities nor provided designated cohort isolation space. These findings can help inform IPC practice in LTCFs with focused education on CDC hand hygiene recommendations, the importance of social distancing, appropriate disinfectant use, and the preparation of designated spaces for the care of infected residents before an outbreak. The authors concluded that performing tele-ICARs in LTCFs enables public health agencies to provide direct and individualized feedback to facilities and identify state-wide opportunities for effective interventions in response to SARS-CoV-2.
ARTICLE OVERVIEW
Residents in LTCFs are increasingly vulnerable to infection, and these facilities are often found deficient in infection and prevention control practices.2 The ICAR tool was developed in response to the Ebola outbreak as a systematic method to assess infection prevention preparedness in skilled nursing homes.3 During the COVID-19 pandemic, a modified remote ICAR (tele-ICAR) was developed. These assessments can be conducted before cases are identified in the LTCFs (proactive) or in response to a facility outbreak (responsive).
The Texas DSHS offered proactive and responsive tele-ICARs to LTCFs across their eight public health regions. During the 8-month study period, 438 LTCFs voluntarily participated (264 proactive and 174 responsive). Infection prevention knowledge gaps identified in both proactive and responsive assessments included a preference for soap and water (rather than alcohol-based hand sanitizer as per CDC recommendation), the contact time for disinfectants, need for suspension of communal dining, the value and necessity for infection prevention audits, need to curtail communal group activities and preparation of dedicated space to cohort residents with COVID-19 infections. As previously summarized, ALF was significantly less likely to have stopped communal dining or have a space to cohort COVID -19 residents when compared to NH/SNF. Further, LTCFs who had proactive tele-ICARs were less likely than those that received responsive tele-ICARs to be compliant with published core IPC guidelines, for example, alcohol-based hand sanitizer over soap and water for hand hygiene or recommended SARS-CoV-2 prevention guidelines, for example, suspension of communal dining and group activities, identification of a dedicated cohort isolation space. The authors concluded that despite conducting the ICARs remotely, they gleaned important information about infection prevention knowledge and practice gaps and provided important and timely education regarding recommended SARS-CoV-2 prevention and outbreak mitigation guidelines.
DISCUSSION AND LIMITATIONS
This article was well-written and methodologically sound. A brief but sufficient up-to-date background on infection prevention in LTCFs and the ICRA process was provided. The authors met their objectives and provided feasibility data on how to conduct tele-ICARs successfully. Acronyms were first summarized to help those unfamiliar with these facilities to clarify their meaning.
The research questions were explicitly stated. The authors solicited a volunteer sample of LTCFs to participate in the tele-ICARs who were not experiencing a COVID-19 outbreak (proactive) and those who were experiencing an outbreak (responsive). It was unclear how many proactive and responsive LTCFs were eligible to participate, and no response rate was reported. As this was a convenience sample, it was not as rigorous a design as conducting a random sample of LTCFs.4 One limitation may be that LTCFs who volunteered for the tele-ICARs were either better or less prepared than LTCFs who did not volunteer to participate. The authors did not mention Institutional Review Board approval (IRB); however, as this project was initiated as a process improvement initiative, IRB oversight was unnecessary.
The authors provided a copy of the ICAR tool domains and measures in the appendix. This may be particularly helpful for infection preventionists (IPs) who want to replicate this work. The data was analyzed using a Fisher's Exact test followed by Bonferroni correction for multiple comparisons. Significant associations were added to a logistic regression model to characterize the relationship. Tables were provided with the results. However, the main tables in the article only included Fisher's exact test p values without Bonferroni correction. This may be somewhat misleading for those who do not thoroughly explore the article.
The authors provide a good discussion that relates the article's findings to the literature and the relationship to IPC practice in LTCFs. Most limitations were addressed, including that different epidemiologists and infection preventionists conducted the tele-ICARs. While this study was conducted in Texas, it contributes to the knowledge about gaps in LTCF IPC practices and provides a novel approach for conducting IPC assessments during a pandemic. Information from this study may assist LTCFs with developing and implementing best practice procedures that incorporate the CDC guidelines for hand hygiene, PPE compliance monitoring, selection of disinfectants, and social distancing during outbreaks. Interestingly, these results are similar to a previously published article from the CDC Tele-ICAR Team.5 In that study, which included 629 LTCFs Tele-ICAR consultations across 19 states, 68% of the facilities identified gaps in core IPC practices: 39% of the facilities failed to follow CDC recommendations for hand hygiene, and 24% had gaps in environmental cleaning policies and procedures (failure to follow contact times appropriately). In addition, they reported that communal practices were infrequently suspended: 6% stopped group activities inside the facility and field trips, 7% stopped communal dining, and only 3% of facilities encouraged residents to remain in their rooms.5
IMPLICATIONS AND CONCLUSIONS FOR INFECTION PREVENTIONISTS
This article identified that LTCFs need increased education regarding IPC practices, especially during a pandemic. To ensure that this vulnerable population is being cared for with IPC best practices, ICAR assessments may be helpful. In most cases, on-site ICARs are recommended.6 However, this article described a successful method for conducting tele-ICARs. While a tele-ICAR is not intended to replace the on-site assessment, IPs could use the remote method during a pandemic or as an interim or follow-up assessment when on-site visits are not feasible. The ICAR further assists the facility in addressing the root causes of incorrect or suboptimal IPC practices, which promotes the protection of residents and staff from not only COVID but other infectious diseases as well.5
While this study was conducted in Texas LTCFs, the knowledge and practice gaps identified are similar to those described in the literature and can be used to help IPs tailor LTCF educational efforts. This study revealed that the most common gap in LTCFs was a preference for soap and water hand hygiene compared to the CDC-recommended alcohol-based hand sanitizer. They hypothesized that this gap might result from misunderstanding the difference between community and healthcare settings. These findings underscore the need for IPs to continue to educate LTCFs on the CDC guidelines for hand hygiene.
Another finding from this study, which validates a gap identified by previous investigators, was a lack of understanding about disinfectants, particularly the importance of contact time. This is a significant knowledge gap for these facilities where rigorous environmental disinfection protocols are necessary to prevent the transmission of epidemiologically important pathogens. Failure to adhere to the manufacturer's instructions for use and the recommended contact time may interfere with the product's efficacy. The authors suggested that multidisciplinary team communication and sustained education, including the environmental services vendor and staff, are needed to address this gap.7
A unique aspect of LTCFs is communal dining and group activities. While this approach promotes social interaction for the residents, it presents a risk when IPC mitigation strategies are required, especially during a pandemic. At the beginning of the pandemic, the CDC recommended that group activities be canceled.8 However, as the pandemic continued, guidance was continually updated. The constant flow of information can be challenging in LTCFs, where staff are already overworked, and understaffing is addressed by agency personnel unfamiliar with the facility. IPs can help LTCFs adapt to innovative approaches to communal activities such as staggering mealtimes and social distancing. Involving the team, especially nursing assistants, would help create strategies that keep patients safe while not overburdening staff.
This study also identified the lack of a cohortation plan. Like communal dining, LTCFs are designed to provide a casual ‘home’ environment where private rooms are not routinely available. Cohorting infected residents can be highly disruptive to residents and staff and significantly affect workload. Developing a cohort plan in advance of an outbreak is ideal and a required element of LTCFs emergency preparedness policy. IPs may need to be creative in their approach and recognize that private rooms are generally not available or feasible, and identifying a unit or part of a unit for cohortation is necessary.
Acknowledgments
This Research Committee Commentaries: Implementation Science Commentary is part of the ongoing activity of the Association for Professionals in Infection Control and Epidemiology Research Committee. The primary purpose of this section in AJIC is to direct and support readers regarding how to evaluate the relevance and implications of a study for implementation in their own setting. The Research Committee presents an assessment of the study quality and study relevance of select articles.
Conflict of interest: No conflict of interest to disclose for the authors.
==== Refs
References
1 Singer R Rodriguez G Garcia B Nutt A Merengwa E. Remote infection control assessments in long-term care facilities during COVID-19 pandemic in Texas, 2020 Am J Infect Control 50 2022 1398 1400
2 United States Government Accountability Office Infection Control Deficiencies Were Widespread and Persistent in Nursing Homes Prior to COVID-19 Pandemic 2020 GAO-20-576R Washington, DC
3 Ostrowsky BE Weil LM Olaisen RH Real-time virtual infection prevention and control assessments in skilled nursing homes, New York, March 2020 – A pilot project Infect Control & Hosp Epidemiol 43 2021 351 357 33736719
4 Turner DP. Sampling methods in research design Headache 60 2020 8 12 31913516
5 Walters MS Prestel C Fike L Remote infection control assessments of US nursing homes during the COVID-19 pandemic, April to June 2020 J Am Med Dir Assoc 23 2022 909 916.e2 35504326
6 National Center for Immunization and Respiratory Diseases (NCIRD), Division of viral diseases, centers for disease control and prevention. Infection Prevention and Control Assessment Tool for Nursing Homes Preparing for COVID- 19. 2021. Accessed October 15, 2021. https://www.cdc.gov/coronavirus/2019-ncov/hcp/assessment-tool-fornursing-homes.html.
7 Quinn MM Henneberger PK and members of the National Institute for Occupational Safety and Health (NIOSH), National Occupational Research Agenda (NORA) Cleaning and Disinfecting in Healthcare Working Group. Cleaning and disinfecting environmental surfaces in health care: Toward an integrated framework for infection and occupational illness prevention Am J Inf Control 43 2015 424 434
8 Centers for Disease Control and Prevention. Interim infection prevention and control recommendations to prevent SARS-CoV-2 spread in nursing homes for nursing homes & long-term care facilities. 2020. Accessed July 22, 2021. https://www.cdc.gov/coronavirus/2019-ncov/hcp/long-term-care.html.
| 36460418 | PMC9705192 | NO-CC CODE | 2022-12-01 23:19:33 | no | Am J Infect Control. 2022 Dec 29; 50(12):1398-1400 | utf-8 | Am J Infect Control | 2,022 | 10.1016/j.ajic.2022.09.023 | oa_other |
==== Front
J Formos Med Assoc
J Formos Med Assoc
Journal of the Formosan Medical Association
0929-6646
0929-6646
Formosan Medical Association, Elsevier
S0929-6646(22)00436-3
10.1016/j.jfma.2022.11.015
Original Article
Evolution of neutralizing antibodies and cross-activity against different variants of SARS-CoV-2 in patients recovering from COVID-19
Liu Wang-Da ab
Wang Jann-Tay ac∗1
Chao Tai-Ling d
Ieong Si-Man d
Tsai Ya-Min d
Kuo Po-Hsien e
Tsai Ming-Jui f
Chen Yi-Jie a
Li Guei-Chi a
Ho Shu-Yuan g
Chen Hui-Hou g
Huang Yu-Shan a
Hung Chien-Ching afh
Chen Yee-Chun ai
Chang Sui-Yuan dg∗∗1
Chang Shan-Chwen aj
a Department of Internal Medicine, National Taiwan University Hospital and National Taiwan University College of Medicine, Taipei, Taiwan
b Department of Medicine, National Taiwan University Cancer Center, Taipei, Taiwan
c Institute of Infectious Diseases and Vaccinology, National Health Research Institutes, Zhunan, Taiwan
d Department of Clinical Laboratory Sciences and Medical Biotechnology, National Taiwan University College of Medicine, Taipei, Taiwan
e Department of Internal Medicine, National Taiwan University Hospital Biomedical Park Hospital, Hsinchu, Taiwan
f Department of Internal Medicine, National Taiwan University Hospital Yun-Lin Branch, Yun-Lin County, Taiwan
g Department of Laboratory Medicine, National Taiwan University Hospital and National Taiwan University College of Medicine, Taipei, Taiwan
h Department of Tropical Medicine and Parasitology, National Taiwan University College of Medicine, Taipei, Taiwan
i Center of Infection Control, National Taiwan University Hospital, Taipei, Taiwan
j School of Medicine, National Taiwan University College of Medicine, Taipei, Taiwan
∗ Corresponding author. Department of Internal Medicine, National Taiwan University Hospital, 7 Chung-Shan South Rd., Taipei City 10002, Taiwan.
∗∗ Corresponding author. Department of Laboratory Medicine, National Taiwan University Hospital, 7 Chung-Shan South Rd., Taipei City 10002, Taiwan.
1 Wang JT and Chang SY contributed equally to this work.
29 11 2022
29 11 2022
12 9 2022
17 11 2022
22 11 2022
.
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Background
Patients recovering from COVID-19 may need vaccination against SARS-CoV-2 because acquired immunity from primary infection may wane, given the emergence of new SARS-CoV-2 variants. Understanding the trends of anti-spike IgG and neutralizing antibody titers in patients recovering from COVID-19 may inform the decision made on the appropriate interval between recovery and vaccination.
Methods
Participants aged 20 years or older and diagnosed with COVID-19 between January and December, 2020 were enrolled. Serum specimens were collected every three months from 10 days to 12 months after the onset of symptom for determinations of anti-spike IgG and neutralizing antibody titers against SARS-CoV-2 Wuhan strain with D614G mutation, alpha, gamma and delta variants.
Results
Of 19 participants, we found a decreasing trend of geometric mean titers of anti-spike IgG from 560.9 to 217 and 92 BAU/mL after a 4-month and a 7-month follow-up, respectively. The anti-spike IgG titers declined more quickly in the ten participants with severe or critical disease than the nine participants with only mild to moderate disease between one month and seven months after SARS-CoV-2 infection (−8.49 vs - 2.34-fold, p < 0.001). The neutralizing activity of the convalescent serum specimens collected from participants recovering from wild-type SARS-CoV-2 infection against different variants was lower, especially against the delta variants (p < 0.01 for each variant with Wuhan strain as reference).
Conclusion
Acquired immunity from primary infection with SARS-CoV-2 waned within 4–7 months in COVID-19 patients, and neutralizing cross-activities against different SARS-CoV-2 variants were lower compared with those against wild-type strain.
Keywords
Plaque reduction neutralization test (PRNT)
Humoral immunity
Spike protein
Variant of concern
Delta variant
B.1.617.2
==== Body
pmcIntroduction
The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection has become pandemic and caused a high number of morbidities and mortalities since 2020.1 The occurrences of SARS-CoV-2 reinfection following primary infection in this pandemic have raised public health concerns.2, 3, 4
A longitudinal study for antibody kinetics after primary SARS-CoV-2 infection showed that neutralizing antibodies could be detected even 300 days after primary infection with only a slight decrease.5 Another study also revealed that the protection effectiveness against reinfection with SARS-CoV-2 was 85% or greater, which could last for seven months.6 However, breakthrough infections occurred in individuals with waning immunity that was acquired from either infection or vaccination; moreover, reinfections with different SARS-CoV-2 variants were also reported.7
Until now, vaccination of patients recovering from SARS-CoV-2 infection has been recommended. However, the appropriate timing of vaccination for these populations remains debating.8 Given the emergence of new SARS-CoV-2 variants, a shorter interval between recovery and vaccination might be necessary. In this study, we aimed to evaluate the trends of anti-spike protein IgG and neutralizing antibody levels of patients who had recovered from coronavirus disease 2019 (COVID-19) and to assess the cross-neutralizing activities of the convalescent sera against different variants of SARS-CoV-2.
Methods
Study population
Participants aged 20 years or older and diagnosed with SARS-CoV-2 infection by real-time polymerase-chain-reaction (RT-PCR) assay at the National Taiwan University Hospital (NTUH) from January, 2020 to December, 2020 were enrolled. Medical records were reviewed to obtain the information on age, gender, underlying comorbidities, clinical features, laboratory profile, serial cycle-threshold (CT) values of SARS-CoV-2 RT-PCR and treatment of COVID-19 of each participant. COVID-19 disease severity was classified according to the COVID-19 treatment guidelines by the National Institutes of Health as asymptomatic, mild, moderate, severe, and critical disease.9 Serum specimens of the participants were collected once or twice per week during their hospital stays, while, during follow-up, serum specimens were obtained at out-patient clinics every three or six months after discharge. The patients were followed-up until death, receiving any SARS-CoV-2 vaccine, or loss to follow-up, whichever occur first. All serum samples were inactivated at 56 °C for 30 min and stored at −20 °C before testing. The serum specimens were tested for neutralizing antibodies and IgG against SARS-CoV-2 spike protein at the following time points, including ten days, one month, four months, seven months, ten months, and 12 months after presentations of initial COVID-19-related symptoms. The study was approved by the Research Ethics Committee of NTUH (NTUH 202002002RIND) and written informed consent was obtained from the participants.
Neutralization assays
Plaque reduction neutralization test (PRNT) was performed on the sequentially collected serum specimens to determine the neutralizing antibody titers against SARS-CoV-2. The serum specimens used in these assays were heat-inactivated at 56 °C for 30 min, and then 2-fold serially diluted in serum-free DMEM media, from 1:80 to 1:1280. PRNT was performed in triplicate in 24-well tissue culture plates. The clinical isolates of SARS-CoV-2 used in the assay included SARS-CoV-2/NTU03/TWN/human/2020 (EPI_ISL 413592), which exhibits the D614G mutation, SARS-CoV-2/NTU49/TWN/human/2020 (EPI_ISL 1010728) as alpha variant, SARS-CoV-2/CGU56/TWN/human/2021 (EPI_ISL 2249615) as gamma variant, and SARS-CoV-2/NTU92/TWN/human/2021 (EPI_ISL 3979387) as delta variant. SARS-CoV-2 (50–100 plaque-forming units, pfu) was incubated with diluted test sera for 1 h at 37 °C before adding to the Vero E6 cell monolayer for another 1 h. Subsequently, virus-serum mixtures were removed and the cell monolayer was washed once with phosphate buffered saline before covering with DMEM media containing 2% fetal bovine serum (FBS) and 1% methylcellulose for 5–7 days. The cells were fixed with 10% formaldehyde overnight. After removal of overlay media, the cells were stained with 0.7% crystal violet and the plaques were counted.
To facilitate conversion of 50% PRNT (PRNT50) to International Unit (IU/mL), the WHO international standard reference panel (20/268 [including reference samples 20/150, 20/148, 20/144, and 20/140]) from the National Institute for Biological Standards and Control (NIBSC; Potters Bar, UK) was used to generate an equation for converting PRNT50 to IU/mL (y=(x-9.3313)/1.453, where y is the value of IU/mL and x is the value of the PRNT50). In this study, a PRNT50 titer lower than 80 will be labeled as 40 (21.1 IU/mL) for further statistical analysis.
Detection of IgG against SARS-CoV-2 spike protein
SARS-CoV-2 spike (S) protein-specific IgG in serum samples was determined using SARS-CoV-2 IgG II Quant assay (Abbott, Abbott Park, Illinois, U.S.A.) according to the manufacturer's instructions. An IgG level higher than 50 AU/mL (7.1 BAU/mL) was considered positive.
Statistical analysis
The geometric mean titers (GMTs) of SARS-CoV-2 anti-spike IgG and neutralizing antibodies were calculated in log-transformed data for statistics. The GMTs of IgG against SARS-CoV-2 spike protein between participants of different severity categories were analyzed using Student's t-test. Neutralizing antibody titers against different variants were analyzed using paired t-tests, with the titer against Wuhan strain as reference. Pearson's product–moment correlation was performed to evaluate the relationship between anti-spike IgG and neutralizing titer. A two-tailed P value less than 0.05 was considered statistically significant. All analyses were performed using Stata/SE software, Version 11.0 (https://www.stata.com).
Results
A total of 19 participants were enrolled, including nine with mild to moderate disease (Group 1), five with severe disease (Group 2) and five with critical disease who developed respiratory failure that led to intubation and mechanical ventilator support (Group 3). The demographic and clinical features of the participants are shown in Table 1 . Participants with severe or critical disease tended to be older and have more comorbidities. In Group 1, one (11.1%) participant received treatment with lopinavir/ritonavir while the other eight (88.9%) participants received hydroxychloroquine. All of the five participants of Group 2 received a 5-day course of remdesivir. For participants with critical disease, four (80%) received high-dose corticosteroids while two (40%) received tocilizumab. None of the participants died of COVID-19, and there were no cases of reinfection during the observation period. Of the 19 participants, full-length virus genome sequences were determined for 12 participants using the virus isolated from their respiratory specimens. The viral sequences were deposited in the GISAID database (Supplementary Table 1). The spike sequences of these strains were very similar to those of the original SARS-CoV-2 strain, with seven having additional D614G mutation and one S254F mutation and none having deletions at V69, H70, and Y144.Table 1 Demographic and baseline clinical features of enrolled patients.
Table 1Characteristics Total (n = 19) Mild/Moderate (n = 9) Severe (n = 5) Critical (n = 5)
Demographic
Median age (IQR), years 48 (32–62) 35 (26–38) 58 (57–65) 61 (55–66)
Male gender 10 (52.6) 3 (33.3) 3 (60) 4 (80)
Underlying disease
Hypertension 6 (31.6) 1 (11.1) 3 (60) 2 (40)
Type 2 diabetes mellitus 1 (5.3) 0 0 1 (20)
Coronary artery disease 2 (10.5) 0 1 (20) 1 (20)
Congestive heart failure 2 (10.5) 0 2 (40) 0
Malignancy 1 (5.3) 0 0 1 (20)
HIV infection 2 (10.5) 2 (22.2) 0 0
Clinical features
Fever 12 (63.2) 5 (55.6) 4 (80) 3 (60)
Cough 13 (68.4) 6 (66.7) 4 (80) 3 (60)
Dyspnea 6 (31.6) 2 (22.2) 1 (20) 3 (60)
Rhinorrhea 5 (26.3) 4 (44.4) 0 1 (20)
Malaise/Myalgia 6 (31.6) 3 (33.3) 2 (40) 1 (20)
Diarrhea 4 (21.1) 2 (22.2) 0 2 (20)
Lab data
Hb (g/dL) 13.7 (12.5–15.1) 13.4 (12–15) 13.9 (12.7–14.8) 13.9 (12.7–15.2)
Platelet (K/μL) 229 (169–288) 223 (196–282) 193 (169–190) 278 (255–327)
WBC (K/μL) 5.96 (4.71–7.23) 5.15 (3.96–6.15) 6.1 (4.84–7.23) 7.26 (5.74–7.38)
Lymphocyte (cells/μL) 1288 (939–1580) 1454 (1119–1721) 1056 (1012–1121) 1222 (723–1092)
Neutrophil (cells/μL) 4239 (2983–5136) 3179 (2039–4492) 4675 (3432–5719) 5709 (4527–6170)
CRP (mg/dL) 4.46 (0.15–11.35) 0.17 (0.1–0.17) 7.21 (3.34–11.44) 9.43 (3.67–11.88)
Procalcitonin (ng/mL) 5.94 (0.03–0.08) 0.03 (0.02–0.04) 0.09 (0.04–0.08) 25.1 (0.08–50.12)
Ferritin (ng/mL) 1933.16 (138.11–1358.1) 178.89 (24.22–354.3) 1023.9 (376.26–852.6) 6578.29 (1849.73–11306.85)
Albumin (g/dL) 4 (3.5–4.4) 4.4 (4.2–4.7) 3.86 (3.7–3.9) 3.5 (3.4–3.5)
ALT U/L) 28 (13–46) 20 (8–28) 33 (25–40) 37 (32–46)
LDH (U/L) 289 (151–403) 153 (133–177) 292 (207–361) 557 (448–666)
BUN (mg/dL) 14 (10.4–16.5) 14 (10.5–16.5) 12.4 (10.1–13.9) 15.6 (12.1–21.7)
Creatinine (mg/dL) 0.9 (0.6–0.9) 0.8 (0.6–0.9) 0.8 (0.7–0.9) 1.2 (0.7–0.9)
CK (U/L) 130 (47–146) 60 (35–63) 144 (55–244) 252 (71–433)
Na (mmol/L) 137 (135–139) 138 (136–139) 135 (134–136) 137 (135–141)
K (mmol/L) 3.5 (3.3–3.7) 3.7 (3.6–3.8) 3.3 (3.1–3.4) 3.4 (3.2–3.5)
Treatment
LPV/r 1 (5.3) 1 (11.1) 0 0
RDV 8 (42.1) 0 5 (100) 3 (60)
HCQ 8 (42.1) 7 (77.8) 0 1 (20)
Steroid 4 (21.1) 0 0 4 (80)
tocilizumab 2 (10.5) 0 0 2 (40)
Outcome
Duration of virus shedding (days) 24 (19–43) 37 (26–46) 22 (17–23) 19 (19–58)
Mortality (people) 0 0 0 0
SARS-CoV-2 identification
Wild type without D614G mutation 5 (26.3) 2 (22.2) 0 3 (60)
Wild type with D614G mutation 7 (36.8) 4 (44.4) 2 (40) 1 (20)
Not identified 7 (36.8) 3 (33.3) 3 (60) 1 (20)
The GMT of anti-spike IgG of the participants in Groups 1 and 2 reached the peak one month after the onset of symptoms, with the level of 125.9 BAU/mL (95% confidence interval [CI], 73.8–214.8) and 869.5 (95% CI, 472.9–1598.6), respectively, while the antibody level reached the peak (1032 BAU/mL) during the first 10 days after symptoms onset of Group 3 (95% confidence interval [CI], 18.4–57971.9). Waning of anti-spike IgG titers from all participants was noted, with GMT decreasing from 560.9 to 217 and 92 BAU/mL after a 4-month and a 7-month follow-up, respectively. Of note, the anti-spike IgG titers declined more quickly in the participants with severe or critical disease than those with only mild to moderate disease between one month and seven months after SARS-CoV-2 infection (−8.49 vs −2.34-fold, p < 0.001) (Fig. 1 ).Figure 1 Evolution of anti-spike IgG of patients with different severity.
Fig. 1
The GMT of neutralizing antibodies against SARS-CoV-2 with D614G mutation showed a similar trend, which decreased from 382.8 (95% CI, 199.6–734.3) at the first month to 80.9 (95% CI, 32.7–200.1) IU/mL at the seventh month after symptom onset. The neutralizing activity against different variants of the serum specimens, especially delta variants, obtained from participants recovering from wild-type SARS-CoV-2 infection regardless of the presence of D614G mutation was lower than that against wild-type virus. The GMTs of neutralizing antibodies against alpha, gamma and delta variants of the serum specimens collected at the first month after primary infection were much lower than that against wild-type virus (p < 0.01 for each variant) (Table 2 ). Throughout the follow-up period, the neutralizing activity against delta variant was lowest among the four types of SARS-CoV-2 tested in this study of participants of three different severity categories (Fig. 2 ).Table 2 Neutralizing activity of different SARS-CoV-2 variants to convalescent serum collected from patients recovered from Wuhan strain.
Table 2 Wuhan Alpha Gamma Delta
Total (N = 75) 2.08 Ref 1.83 p < 0.001 (0.14–0.37) 1.73 p < 0.001 (0.22–0.43) 1.69 p < 0.001 (0.28–0.47)
Time
0–10 days (N = 17) 1.96 Ref 1.75 p = 0.13 (−0.09-0.59) 1.68 p = 0.02 (0.06–0.57) 1.76 p = 0.06 (−0.01-0.47)
1 month (N = 19) 2.58 Ref 2.22 p = 0.01 (0.12–0.62) 2.01 p = 0.001 (0.23–0.72) 2.01 p < 0.001 (0.34–0.7)
4 months (N = 10) 2.19 Ref 1.95 p = 0.24 (−0.12-0.44) 1.79 p = 0.05 (−0.01-0.58) 1.68 p = 0.03 (0.04–0.79)
7 months (N = 9) 1.91 Ref 1.7 p = 0.19 (−0.13-0.55) 1.71 p = 0.19 (−0.12-0.52) 1.45 p = 0.03 (0.07–0.85)
10 months (N = 10) 1.82 Ref 1.68 p = 0.11 (−0.06-0.47) 1.63 p = 0.15 (−0.11-0.58) 1.44 p = 0.006 (0.15–0.68)
12 months (N = 10) 1.69 Ref 1.48 p = 0.04 (0.02–0.45) 1.4 p = 0.03 (0.03–0.57) 1.53 p = 0.03 (0.03–0.36)
Disease severity
Severe (N = 20) 2.47 Ref 2.1 p = 0.01 (0.12–0.79) 2.17 p = 0.002 (0.15–0.6) 2.01 p < 0.001 (0.32–0.71)
Moderate (N = 22) 2.23 Ref 2.02 p = 0.01 (0.06–0.43) 1.73 p < 0.001 (0.25–0.67) 1.7 p < 0.001 (0.36–0.74)
Mild (N = 33) 1.7 Ref 1.52 p = 0.008 (0.04–0.25) 1.45 p = 0.01 (0.06–0.36) 1.48 p = 0.007 (0.05–0.29)
Note. Neutralization titer was recorded in log IU/mL.
Figure 2 Neutralizing activity of different SARS-CoV-2 variants to convalescent serum obtained from patients recovered from Wuhan strain (p < 0.001 for alpha vs Wuhan, gamma vs Wuhan, and delta vs Wuhan).
Fig. 2
Of all samples with a neutralizing antibody titer higher than 48.63 IU/mL, which was the lower limit of detection of neutralizing antibody titer in our study, only one was tested negative for anti-spike IgG. Pearson's product–moment correlation analysis revealed a statistically significantly positive correlation between the quantitative levels of anti-spike IgG and the titers of neutralizing antibodies against the virus strain with D614G mutation (r = 0.78, p < 0.0001). Such correlations remained even though the neutralizing activities against other SARS-CoV-2 variants were lower (alpha variant, r = 0.64, p < 0.001; gamma variant, r = 0.58, p < 0.001; delta variant, r = 0.67, p < 0.001) (Fig. 3 ).Figure 3 Correlation between anti-spike IgG titer and neutralizing activity against different SARS-CoV-2 variants. (A. Wuhan strain with D614 mutation, r = 0.78, p < 0.001; B. alpha variant, r = 0.64, p < 0.001; C. gamma variant, r = 0.58, p < 0.001; D. delta variant, r = 0.67, p < 0.001).
Fig. 3
Discussion
In this longitudinal follow-up study of the anti-spike IgG and neutralizing antibody titers in patients recovering from COVID-19, we found that both titers waned over time and the neutralizing activities against different SARS-CoV-2 variants were much lower than those against the original strain causing the first wave of SARS-CoV-2 infection.
Until now, our understanding of the durability of antibodies against SARS-CoV-2 acquired from primary infection with SARS-CoV-2 remains relatively limited. Our study demonstrated that sustained antibody responses could be detected more than six months after primary infection, which is in line with the findings of other studies.10 , 11 We also found that a stronger antibody response was associated with a more severe disease status, which echoed the study by den Hartog et al.12 However, those with more severe disease had a more rapid decline of antibody titers, as previous studies have described.13 An increasing age was shown to be associated with a stronger neutralizing antibody response.14 , 15 The patients with milder disease in our study was much younger than the patients enrolled in the study by den Hartog et al., which might explain this difference. Moreover, the patients with severe or critical disease in our cohort tended to have more comorbidities such as malignancies, which might interfere with the antibody response.
Our study revealed that serum specimens obtained from patients who recovered from infection with wild-type strain showed lower neutralizing activities against different variants, especially against the delta variant. Previous studies have demonstrated that the delta variant was resistant to neutralization by certain monoclonal antibodies, including bamlanivimab, and the convalescent serum specimens obtained from affected individuals were fourfold less potent against the delta variant than against the alpha variant.16 A recent study has shown that the level of neutralizing activity from either vaccination or convalescent cohorts was predictive of seroprotection against severe COVID-19, and a sustained protective effect might persist despite the decrease of neutralizing activity.17 Rosati et al. demonstrated that SARS-CoV-2 antibodies induced by wild-type strain potently recognized alpha-spike-receptor-binding-domain (RBD) but only showed slightly lower affinity to delta-spike-RBD in pseudovirus tests.18 In addition, Betton et al. demonstrated that the neutralizing activity of serum specimens from patients recovering from infection with wild-type strain lasted at least for six months, regardless of a decreased IgG level, which could even confer a cross-protection against the D614G, alpha, beta, and gamma variants.19
In the real-world setting, previous SARS-CoV-2 infection within 13 months was shown to be an independent protective factor against reinfection of delta strain, and patients with reinfection tended to have milder disease.20, 21, 22 However, even though patients infected with wild-type SARS-CoV-2 had more sustained and higher neutralizing antibody titers than those in vaccinated people, neutralizing antibodies from convalescent sera of patients infected with wild type, alpha, beta or delta variant had low cross-reactivity to omicron variants.23, 24, 25 Therefore, given the high transmissibility of the circulating delta/omicron variants in the community, there should be no delay in vaccination among patients recovering from COVID-19 since a waning immunity has been observed. In a retrospective cohort study from Israel, in which 149,032 patients recovering from SARS-CoV-2 infection were followed for 270 days, the reinfection rate among unvaccinated patients were significantly higher than those with vaccination after primary infection.26 Moreover, Hall et al. also demonstrated that a booster vaccination for previously infected individuals provided a sustained protection, which provides supportive evidence for the recommendation of a booster dose of SARS-CoV-2 vaccination.27
Our study demonstrated a strong correlation between anti-spike IgG level and neutralizing activity. However, the correlation between anti-spike IgG level and neutralizing activity with the real-world protection remains debating. A predictive model by Khoury DS et al. showed 20% of convalescent serum specimens (54 IU/mL) can approximate 50% vaccine efficacy.17 In our study, the mean neutralizing activity of the serum specimens obtained 12 months after primary infection fell below this standard. However, the mean neutralizing activity against other variants such as the alpha or delta strain showed a faster waning in those obtained at month seven or even earlier. Another method to predict clinical efficacy is the BAU conversion model by Feng et al., which suggested that anti-spike IgG titers higher than 264 and 899 BAU/mL were correlated with 80% and 90% vaccine efficacy against the alpha variants, respectively.28 However, in our cohort, the serum specimens obtained seven months after primary infection also failed to meet this standard. Hence, a shorter interval between recovery from COVID-19 and a booster vaccination should be considered.
This is the first study that demonstrates the longitudinal follow-up of anti-spike IgG among patients with different severity in the early stage of the pandemics in Taiwan, which is previously less discussed. However, there are still several limitations in our study. First, the case number in this study was small and some participants missed blood sampling, which might dampen the power of statistical analyses. Secondary, not all the virus strains from the participants were available though there were no documented cases of infection with alpha strain in Taiwan before April, 2021. Therefore, there was a strong epidemic evidence of wild-type strain infection in our cohort even though the presence of D614G mutation was not completely clarified. Nevertheless, a recent study by Weissman et al. reported that even though D614G mutation increased SARS-CoV-2 transmission efficiency, such mutation increased the susceptibility of the virus strain to neutralization by receptor-binding domain (RBD) monoclonal antibodies and convalescent sera.29 Third, there were no domestic cases of reinfection in Taiwan during the study period, which might preclude us from evaluating the association between waning antibody titers and protection. Last, we used the cut-off level of 21 IU/mL as the lower limit of detection for neutralizing activity in the estimation of Pearson's correlation model for all neutralizing titers below 42 IU/mL, which might interfere the identification of correlation between IgG and neutralizing activity. Nevertheless, a strong correlation was still observed as previous study.30
Conclusions
Our study found waning antibody titers within 4–7 months after primary infection with SARS-CoV-2. Moreover, the neutralizing cross-activities against different SARS-CoV-2 variants were lower compared with that against wild-type strain virus. Our findings imply that vaccination of individuals who have recovered from COVID-19 should be considered in the ongoing pandemic with emergence of new SARS-CoV-2 variants. An interval with no longer than seven months could be considered, though the optimal timing for vaccination warrants further evaluations.
Declaration of competing interest
None to declare.
Appendix A Supplementary data
The following is the Supplementary data to this article:Multimedia component 1
Multimedia component 1
Acknowledgement
This work was financially supported by the ‘Research Center for Epidemic Prevention Science’ from Exploration of Novel Therapies by the National Science and Technology Council (NSTC), Taiwan R.O.C. under Grant no. NSTC 111-2321-B-002-017-.
Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.jfma.2022.11.015.
==== Refs
References
1 Su S.Y. Lee W.C. Monitoring the peaks of multiwave COVID-19 outbreaks J Microbiol Immunol Infect 2021 S1684 1182(21)00144-4
2 Abu-Raddad L.J. Chemaitelly H. Bertollini R. National Study Group for COVID-19 Epidemiology Severity of SARS-CoV-2 reinfections as compared with primary nfections N Engl J Med 385 2021 2487 2489 34818474
3 Selhorst P. van Ierssel S.H. Michiels J. Mariën J. Bartholomeeusen K. Dirinck E. Symptomatic severe acute respiratory syndrome coronavirus 2 reinfection of a healthcare worker in a Belgian nosocomial outbreak despite primary neutralizing antibody response Clin Infect Dis 73 2021 e2985 e2991 33315049
4 Akinbami L.J. Biggerstaff B.J. Chan P.A. McGibbon E. Pathela P. Petersen L.R. Reinfection with SARS-CoV-2 among previously infected healthcare personnel and first responders Clin Infect Dis 2021 ciab952 [Epub ahead of print]
5 Scheiblauer H. Nübling C.M. Wolf T. Khodamoradi Y. Bellinghausen C. Sonntagbauer M. Antibody response to SARS-CoV-2 for more than one year - kinetics and persistence of detection are predominantly determined by avidity progression and test design J Clin Virol 146 2022 105052
6 Abu-Raddad LJ, Chemaitelly H, Coyle P, Malek JA, Ahmed AA, Mohamoud YA, et al. SARS-CoV-2 antibody-positivity protects against reinfection for at least seven months with 95% efficacy. EClinicalMedicine. 202;35:100861.
7 Jeffery-Smith A. Rowland T.A.J. Patel M. Whitaker H. Iyanger N. Williams S.V. Reinfection with new variants of SARS-CoV-2 after natural infection: a prospective observational cohort in 13 care homes in England Lancet Healthy Longev 2 2021 e811 e819 34873592
8 Centers for Disease Control and Prevention. Interim Clinical Considerations for Use of COVID-19 Vaccines Currently Approved or Authorized in the United States. Available at: https://www.cdc.gov/vaccines/covid-19/clinical-considerations/covid-19-vaccines-us.html#CoV-19-vaccination.
9 National Institutes of Health. COVID-19 Treatment Guidelines: Clinical Spectrum of SARS-CoV-2 Infection. Available at: https://www.covid19treatmentguidelines.nih.gov/overview/clinical-spectrum Accessed 26th Jan 2022.
10 Wei J. Matthews P.C. Stoesser N. Maddox T. Lorenzi L. Studley R. Anti-spike antibody response to natural SARS-CoV-2 infection in the general population Nat Commun 12 2021 6250 34716320
11 Gallais F. Gantner P. Bruel T. Velay A. Planas D. Wendling M.J. Evolution of antibody responses up to 13 months after SARS-CoV-2 infection and risk of reinfection EBioMedicine 71 2021 103561
12 den Hartog G. Vos E.R.A. van den Hoogen L.L. van Boven M. Schepp R.M. Smits G. Persistence of antibodies to severe acute respiratory syndrome coronavirus 2 in relation to symptoms in a nationwide prospective study Clin Infect Dis 73 2021 2155 2162 33624751
13 Ward H. Cooke G.S. Atchison C. Whitaker M. Elliott J. Moshe M. Prevalence of antibody positivity to SARS-CoV-2 following the first peak of infection in England: serial cross-sectional studies of 365,000 adults Lancet Reg Health Eur 4 2021 100098
14 Wang X. Guo X. Xin Q. Pan Y. Hu Y. Li J. Neutralizing antibody responses to severe acute respiratory syndrome coronavirus 2 in coronavirus disease 2019 inpatients and convalescent patients Clin Infect Dis 71 2020 2688 2694 32497196
15 Karuna S. Li S.S. Grant S. Walsh S.R. Frank I. Casapia M. Neutralizing antibody responses over time in demographically and clinically diverse individuals recovered from SARS-CoV-2 infection in the United States and Peru: a cohort study PLoS Med 18 2021 e1003868
16 Planas D. Veyer D. Baidaliuk A. Staropoli I. Guivel-Benhassine F. Rajah M.M. Reduced sensitivity of SARS-CoV-2 variant Delta to antibody neutralization Nature 596 2021 276 280 34237773
17 Khoury D.S. Cromer D. Reynaldi A. Schlub T.E. Wheatley A.K. Juno J.A. Neutralizing antibody levels are highly predictive of immune protection from symptomatic SARS-CoV-2 infection Nat Med 27 2021 1205 1211 34002089
18 Rosati M. Terpos E. Ntanasis-Stathopoulos I. Agarwal M. Bear J. Burns R. Sequential analysis of binding and neutralizing antibody in COVID-19 convalescent patients at 14 Months after SARS-CoV-2 infection Front Immunol 12 2021 793953 Nov 26
19 Betton M. Livrozet M. Planas D. Fayol A. Monel B. Védie B. Sera neutralizing activities against severe acute respiratory syndrome coronavirus 2 and multiple variants 6 months after hospitalization for Coronavirus Disease 2019 Clin Infect Dis 73 2021 e1337 e1344 33851216
20 Kim P. Gordon S.M. Sheehan M.M. Rothberg M.B. Duration of SARS-CoV-2 natural immunity and protection against the delta variant: a retrospective cohort Study Clin Infect Dis 2021 ciab999 [Epub ahead of print]
21 Bates T.A. McBride S.K. Winders B. Schoen D. Trautmann L. Curlin M.E. Antibody response and variant cross-neutralization after SARS-CoV-2 breakthrough infection JAMA 327 2 2022 179 181 Jan 11 34914825
22 Abu-Raddad L.J. Chemaitelly H. Ayoub H.H. Yassine H.M. Benslimane F.M. Al Khatib H.A. Association of prior SARS-CoV-2 infection with risk of breakthrough infection following mRNA vaccination in Qatar JAMA 326 2021 1930 1939 34724027
23 Schmidt F. Muecksch F. Weisblum Y. Da Silva J. Bednarski E. Cho A. Plasma neutralization of the SARS-CoV-2 omicron variant N Engl J Med 386 6 2022 599 601 35030645
24 Rössler A. Riepler L. Bante D. von Laer D. Kimpel J. SARS-CoV-2 Omicron variant neutralization in serum from vaccinated and convalescent Persons N Engl J Med 386 7 2022 698 700 35021005
25 Dupont L. Snell L.B. Graham C. Seow J. Merrick B. Lechmere T. Neutralizing antibody activity in convalescent sera from infection in humans with SARS-CoV-2 and variants of concern Nat Microbiol 6 11 2021 1433 1442 34654917
26 Hammerman A. Sergienko R. Friger M. Beckenstein T. Peretz A. Netzer D. Effectiveness of the BNT162b2 vaccine after recovery from COVID-19 N Engl J Med 386 13 2022 1221 1229 35172072
27 Hall V. Foulkes S. Insalata F. Kirwan P. Saei A. Atti A. Protection against SARS-CoV-2 after COVID-19 vaccination and previous infection N Engl J Med 386 13 2022 1207 1220 35172051
28 Feng S. Phillips D.J. White T. Sayal H. Aley P.K. Bibi S. Correlates of protection against symptomatic and asymptomatic SARS-CoV-2 infection Nat Med 27 2021 2032 2040 34588689
29 Weissman D. Alameh M.G. de Silva T. Collini P. Hornsby H. Brown R. D614G spike mutation increases SARS CoV-2 susceptibility to neutralization Cell Host Microbe 29 1 2021 23 31.e4 33306985
30 Dolscheid-Pommerich R. Bartok E. Renn M. Kümmerer B.M. Schulte B. Schmithausen R.M. Correlation between a quantitative anti-SARS-CoV-2 IgG ELISA and neutralization activity J Med Virol 94 2022 388 392 34415572
| 36496300 | PMC9705194 | NO-CC CODE | 2022-12-07 23:16:36 | no | J Formos Med Assoc. 2022 Nov 29; doi: 10.1016/j.jfma.2022.11.015 | utf-8 | J Formos Med Assoc | 2,022 | 10.1016/j.jfma.2022.11.015 | oa_other |
==== Front
Lancet Reg Health Southeast Asia
Lancet Reg Health Southeast Asia
The Lancet Regional Health. Southeast Asia
2772-3682
2772-3682
The Author(s). Published by Elsevier Ltd.
S2772-3682(22)00138-X
10.1016/j.lansea.2022.100121
100121
Articles
Effectiveness of heterologous third and fourth dose COVID-19 vaccine schedules for SARS-CoV-2 infection during delta and omicron predominance in Thailand: a test-negative, case-control study
Intawong Kannikar a1
Chariyalertsak Suwat a∗1
Chalom Kittipan b
Wonghirundecha Thanachol b
Kowatcharakul Woravut c
Thongprachum Aksara a
Chotirosniramit Narain d
Teacharak Worachet e
Pimpinan khammawan e
Waneesorn Jarurin f
Iamsirithaworn Sopon g
a Faculty of Public Health, Chiang Mai University, Chiang Mai, Thailand
b Chiang Mai Provincial Health Office, Ministry of Public Health, Chiang Mai, Thailand
c Sansai Hospital, Ministry of Public Health, Chiang Mai, Thailand
d Faculty of Medicine, Chiang Mai University, Chiang Mai, Thailand
e Nakornping Hospital, Ministry of Public Health, Chiang Mai, Thailand
f Regional Medical Sciences Center 1, Chiang Mai, Thailand
g Department of Disease Control, Ministry of Public Health, Bangkok, Thailand
∗ Corresponding author at: Faculty of Public Health, Chiang Mai University, 239, Huay Kaew Road, Muang District, Chiang Mai Thailand, 50200 (S. Chariyalertsak)
1 Contributed equally
29 11 2022
29 11 2022
1001214 8 2022
4 10 2022
15 11 2022
© 2022 The Author(s). Published by Elsevier Ltd.
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Background
The Coronavirus disease 2019 (COVID-19) pandemic has evolved quickly, with numerous waves of different variants of concern resulting in the need for countries to offer continued protection through booster vaccination. To ensure adequate vaccination coverage, Thailand has proactively adopted heterologous vaccination schedules. While randomised controlled trials have assessed homologous schedules in detail, limited data has been reported for heterologous vaccine effectiveness (VE).
Methods
Utilising a unique active surveillance network established in Chiang Mai, Northern Thailand, we conducted a test-negative case control study to assess the VE of heterologous third and fourth dose schedules against SARS-CoV-2 infection among suspect-cases during Oct 1–Dec 31, 2021 (delta-predominant) and Feb 1–Apr 10, 2022 (omicron- predominant) periods.
Findings
After a third dose, effectiveness against delta infection was high (adjusted VE 97%, 95% CI 94 – 99%) in comparison to moderate protection against omicron (adjusted VE 31%, 95% CI 26 – 36%). Good protection was observed after a fourth dose (adjusted VE 75%, 95% CI 71 – 80%). VE was consistent across age groups for both delta and omicron infection. The VE of third or fourth doses against omicron infection were equivalent for the three main vaccines used for boosting in Thailand, suggesting coverage, rather than vaccine type is a much stronger predictor of protection.
Interpretation
Appropriately timed booster doses have a high probability of preventing COVID-19 infection with both delta and omicron variants. Our evidence supports the need for ongoing national efforts to increase population coverage of booster doses.
Funding
This research was supported by the National Research Council of Thailand (NRCT) under The Smart Emergency Care Services Integration (SECSI) project to Faculty of Public Health Chiang Mai University.
==== Body
pmcIntroduction
As of November 9, 2022, the Coronavirus disease 2019 (COVID-19) pandemic caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has led to more than 638 million confirmed cases globally with more almost 200 million in Asia and 4.7 million in Thailand alone.1 Almost 6.6 million deaths were reported worldwide, with almost 1.5 million deaths across Asia and over 33,000 in Thailand.1 COVID-19 pandemic has also created an enormous burden on healthcare, on people and on the economy.2 , 3 While public health measures like wearing of masks, social distancing and appropriate hygiene measures were able to limit the spread of SARS-CoV-2, it was the rapid development and deployment of vaccines which reduced the impact of COVID-19 substantially.4 , 5
WHO has licensed 11 COVID-19 vaccines to date and globally over 12 billion doses have been administered.6 While these vaccines have had an enormous impact in countries that have achieved high coverage rates, as at May 22, 2022, only 57 countries have vaccinated 70% of their population and almost one billion people in lower-income countries remain unvaccinated.7 There are six approved COVID-19 vaccines in Thailand8 and a sustained effort by the government has resulted in 83% of the population being fully vaccinated (two doses) and an additional 47% receiving three doses or above as of October 21, 2022.1 , 9
Randomised controlled clinical trials demonstrated several COVID-19 vaccines to be safe and immunogenic in homologous schedules, with high efficacy against both symptomatic infection and severe outcomes such as hospitalisation and death. This enabled the rapid emergency use authorisation and initial rollout of the following vaccines in Thailand: CoronaVac (Sinovac) in March 202110 ChAdOx1 nCoV-19 (AstraZeneca) in June 202111 and BNT162b2 (Pfizer-BioNTech) in October 2021.12
Subsequently, most of the initial vaccinations in Thailand were given as two doses of CoronaVac for people aged 18-59 years old or two doses of ChAdOx1 nCoV-19 for those who aged 60 or above. People living with chronic medical conditions such as coronary artery disease, chronic kidney disease, cancer patients on chemotherapy, were given priority for vaccination. With the arrival of BNT162b2, doses were initially targeted to younger age groups (>12 years). Due to challenges in vaccine supply and to manage concerns around the effectiveness and duration of CoronaVac, “mix and match” vaccine schedules were implemented from July 2021 onwards including third dose (boosters) with ChAdOx1 nCoV-19 and BNT162b2. Small number of fourth doses (second booster) were administered to high-risk individuals throughout Q4 2021 but were implemented more widely beginning in January 2022, using BNT162b2, ChAdOx1 nCoV-19 and Spikevax (Moderna) in part to address additional concerns around potential immune escape by the omicron variant. The majority of fourth doses vaccines in Thailand have been administered to individuals receiving a two-dose CoronaVac primary series. The initial global clinical trials evaluated efficacy of vaccines (using homologous schedules) against early variants of concern. The most widely used vaccines in real world studies showed high and equivalent effectiveness, especially against severe COVID-19 outcomes.13 Some studies have reported higher neutralizing-antibody response with heterologous boosters as compared to homologous boosters.14, 15, 16, 17 However, there is limited data available on the real-world vaccine effectiveness (VE) of heterologous schedules, particularly against the newer omicron variants.18 , 19 Reports of waning antibody titres and protection against infections, particularly with the omicron variant(s) have driven the rollout of booster doses globally, and hence, it is critical to evaluate the effectiveness of additional doses. The evaluation becomes even more relevant for Asian countries where heterologous schedules have been widely used.
The current study draws on a unique active surveillance network20 established in Chiang Mai, located in Northern Thailand, with a population of 1.6 million. The comprehensive system allows serial assessment of VE for SARS-CoV-2 infection of heterologous schedules during delta-predominant and omicron-predominant periods in the same population. The primary objectives of the study were to evaluate the effectiveness of heterologous three dose COVID-19 vaccine schedules for SARS-CoV-2 infection during delta-predominant period, and to evaluate the effectiveness of heterologous three dose and four dose COVID-19 vaccine schedules for SARS-CoV-2 infection during omicron-predominant period.
Methods
Study population
Residents of Chiang Mai, Thailand, aged 18 years or older, presenting to any community-based testing facilities for a SARS-CoV-2 test during Oct 01–Dec 31, 2021 (delta-predominant) and Feb 01–April 10, 2022 (omicron-predominant) were assessed for eligibility to be included in the study. Molecular testing revealed 96.5% delta and 95.6% omicron lineage during Oct 1–Dec 31, 2021 and Feb 1–April 10, 2022 periods respectively. Tests done in Jan 2022 were excluded due to mixed delta-omicron lineage among samples (omicron 75%, Delta 25%). The data capture ended on April 10, 2022, which was the last date when the community testing ended in Chiang Mai, and shifted to self-testing.
Subjects were included in the study if they met suspect-case criteria, i.e. either close-contacts of COVID-19 cases or attended an event where a COVID-19 outbreak was detected or had symptoms suggestive of COVID-19. Those with uncertain exposure were excluded. Non-Thai residents (foreigners and migrants) were excluded as the vaccination and other data for this group may be incomplete.
The patient selection flow is presented under Supplementary Figures 1a and 1b.
Data sources
We have previously published the details on creating and implementing the information systems used in this study.19 In brief: COVID-19 cases are reported in the surveillance system of Chiang Mai Provincial Health Office (Epid-CM platform), under the Communicable Disease Control Act (B.E. 2558) which mandates national reporting of all COVID-19 cases. After detection of COVID-19 case, the patient details, including laboratory results are entered into the system under a unique individual ID. Epid-CM is synchronized with Chiang Mai hospital management platform for COVID-19 (CMC-19 platform). Data on progression of the disease and treatments are recorded in each hospital information system. Death cases are reported to Chiang Mai Provincial Health Office and recorded in Epid-CM.
Community-based testing sites were initiated in the city of Chiang Mai since April 2021 and provided free COVID-19 tests. Those tested included close contacts of COVID-19 cases, attendees of an event with a COVID-19 outbreak, or those with suggestive respiratory symptoms. Health personnel performed either RT-PCR (November 2021 to January 2022) or antigen testing (February 2022 onwards), and results were uploaded in Epid-CM.
All national vaccination records are available from the Ministry of Public Health Immunization Center (MOPH IC) database maintained by the Ministry of Public Health, Thailand.
Ethical considerations
The study was conducted on routine data collected as part of the national COVID-19 response under the Communicable Disease ACT (B.E. 2558) and was exempted from ethics review. Data were de-identified at source and analysed by Chiang Mai Provincial Health Office and Faculty of Public Health, Chiang Mai University.
Study design
A test-negative, case-control analysis was conducted to evaluate the effectiveness of heterologous three dose COVID-19 vaccine schedules for SARS-CoV-2 infection during delta-predominant period, and heterologous three dose and four dose COVID-19 vaccine schedules for SARS-CoV-2 infection during omicron-predominant period among suspect-cases. “Cases” were defined as those with a positive SARS-CoV-2 result, and “controls” were those with negative SARS-CoV-2 result, either by RT-PCR or medically administered antigen testing. The type of COVID-19 vaccine, and date of vaccination were extracted from MOPH-IC. Subjects who received their COVID-19 vaccination within 14 days of the test date were excluded to allow time for the development of adequate immune responses.
Statistical analysis
Descriptive statistics were reported separately for the cases and controls, stratified by delta and omicron predominance. Continuous variables were summarised as mean and SD or median and IQR depending on the distribution. Categorical variables were summarised as frequency and percentages. Between group comparisons were done using Mann-Whitney U test or Kruskal Wallis test for continuous variables and Chi-squared test for categorical variables.
Associations between SARS-CoV-2 infection and heterologous vaccination schedule were estimated by comparing the odds of vaccination (exposed) vs no vaccination (unexposed) separately for delta and omicron-predominant periods. The odds ratio (OR) was used to estimate VE, where VE = (1 – OR) × 100% with 95% CI. Separate VEs were calculated for different vaccination schedules and stratified by age group. Forest plots were used to visualise the VEs. Multivariable logistic regression models were used to estimate ORs for SARS-CoV-2 infection, adjusting for age in years, gender, calendar day of test (in weekly units), separately for delta and omicron-predominant periods. All statistical analyses were conducted using Stata (version 15.0 SE, College station, TX: StataCorp LP). Significance tests were 2-sided and a p-values <0.05 were considered statistically significant.
Role of the funding source
This research was supported by the National Research Council of Thailand (NRCT) under The Smart Emergency Care Services Integration (SECSI) project to Faculty of Public Health Chiang Mai University. The funder of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication.
Results
Population demographics
There was a total of 19,235 COVID-19 cases and 156 deaths during the delta predominant period, and 296,064 COVID-19 cases and 175 deaths during the omicron-predominant period in Chiang Mai province. For the final analysis, a total of 27,301 subjects (2,130 cases and 25,171 controls) and 36,170 subjects (14,682 cases and 21,488 controls) were included during the delta and omicron-predominant periods respectively (Supplementary Figures 1a and1b).
During the delta-predominant period, cases and controls were of similar ages, while during the omicron-predominant period, cases were slightly younger than controls, primarily due to a higher proportion of individuals aged 18-29 years identified as cases during this period. In both periods, subjects who received boosters were significantly older compared to those who received only one or two doses. During the omicron-predominant period, subjects who received four doses were more likely to be in 30–59 age group, while those who received three doses were more likely to be in 60–69 age group. (Supplementary Tables 1a &1b). This is reflective of the vaccine roll out in Thailand, where individuals aged <60 years received their first doses earlier from April 2021, while those aged >60 years mostly received their first dose beginning in Jun 2021. A slightly higher proportion of females were included in both periods and overall, proportions were relatively consistent across doses with the exception of a higher proportion of males being unvaccinated during the delta-predominant period or receiving only single dose of vaccine during omicron-predominant period.
Multiple schedules of vaccines were used for the primary schedule and vaccine use varied between the two periods. Overall, controls were more likely to have received both two-dose and booster vaccinations, and cases had significantly longer median intervals since their last vaccinations during the omicron-predominant period for their second dose (117 vs 104 days), third dose (58 vs 50 days) and fourth dose (44 vs 39 days) (Table 1 b). Notably during the delta-predominant period, only five (0.4%) cases were reported among 1,197 subjects receiving third doses, (Table 1 a). During the omicron-predominant period, amongst 12,366 subjects with recorded third dose vaccinations, approximately 46% received BNT162b2, 35% received ChAdOx1 nCoV-19, 18% received mRNA-1273 and less than 0.5% received inactivated vaccine (Sinovac). For the 823 subjects with a recorded fourth dose in the omicron-predominant period, approx. 50% received BNT162b2, 44% received mRNA-1273 and approx. 6% received ChAdOx1 nCoV-19 (Table 1b).Table 1b Characteristics of subjects included in analysis of association of vaccination with SARS-CoV-2 infection in adults, during Omicron-predominant period (Feb 1–Apr 10, 2022)
Total (N=36170) SARS-CoV-2 status
Variable Negative (Controls) Positive (Cases) p-value
Number (%) 214881 (59.4) 14682 (40.6) -
Age, years
Median (IQR) 33 (24,46) 35 (25,47) 31 (23,44) <0.01
Age group, n (%)
18-29 14981 (41.42) 8107 (37.7) 6874 (46.8) <0.01
30-39 7712 (21.3) 4732 (22.0) 2980 (20.3)
40-49 6276 (17.3) 4034 (18.7) 2242 (15.3)
50-59 4125 (11.4) 2686 (12.5) 1439 (9.8)
60-69 2340 (6.5) 1471 (6.8) 869 (5.9)
≥70 736 (2.0) 458 (2.1) 278 (1.9)
Gender, n (%)
Male 16783 (46.4) 10011 (46.6) 6772 (46.1) 0.38
Female 19387 (53.6) 11477 (53.4) 7910 (53.9)
Vaccination Status, n (%)
Unvaccinated 4610 (12.7) 2681 (12.5) 1929 (13.1) <0.01
Vaccinated One dose only 470 (1.3) 272 (1.3) 198 (1.3)
Vaccinated two doses only 17897 (49.5) 9934 (46.2) 7963 (54.2)
Vaccinated three doses only 12369 (34.2) 7950 (37.0) 4419 (30.1)
Vaccinated four doses 824 (2.3) 651 (3.0) 173 (1.2)
Type of primary vaccine series (n=17893)1, n (%)
SV/SP-AZ 7780 (43.5) 4464 (44.9) 3316 (41.7) <0.01
AZ-PFZ/Mod 3661 (20.5) 1904 (19.2) 1757 (22.1)
SV-SV or SP-SP 3080 (17.2) 1713 (17.2) 1367 (17.2)
PFZ-PFZ 2340 (13.1) 1237 (12.5) 1103 (13.9)
Mod-Mod 516 (2.9) 294 (2.9) 222 (2.8)
AZ-AZ 396 (2.2) 249 (2.5) 147 (1.9)
SV/SP-PFZ/Mod 120 (0.7) 72 (0.7) 48 (0.6)
Type of third vaccine dose (n=12366)2, n (%)
PFZ 5700 (46.1) 3670 (46.2) 2030 (45.9) 0.50
AZ 4418 (35.7) 2874 (36.2) 1544 (35.0)
Mod 2212 (17.9) 1378 (17.3) 834 (18.9)
SV/SP 36 (0.29) 28 (0.3) 8 (0.2)
Type of fourth vaccine dose (n=823)3, n (%)
Mod 356 (43.3) 278 (42.7) 78 (45.4) 0.68
PFZ 414 (50.3) 329 (50.5) 85 (49.4)
AZ 53 (6.4) 44 (6.7) 9 (5.2)
Median (IQR) time since last vaccination4, days 85 (53,122) 80 (48,112) 93 (61,137) <0.01
Median (IQR) time since last vaccination (two doses only) 110 (84,150) 104 (81,140) 117 (89,167) <0.01
Median (IQR) time since last vaccination (three doses only) 53 (35,75) 50 (32,74) 58 (40,77) <0.01
Median (IQR) time since last vaccination (four doses) 40 (26,57) 39 (26,56) 44 (27,62) 0.06
Vaccination coverage
Completed primary vaccine series, n (%) 31090 (85.9) 18535 (86.3) 12555 (85.5) 0.04
Completed third vaccine dose, n (%) 13193 (36.5) 8601 (40.0) 4592 (31.3) <0.01
Completed fourth vaccine dose, n (%) 824 (2.3) 651 (3.0) 173 (1.2) <0.01
SV=CoronaVac (Sinovac), SP=Sinopharm, AZ=ChAdOx1 nCoV-19 (AstraZeneca), PFZ=BNT162b2 (Pfizer-BioNTech), Mod= mRNA-1273 (Moderna)
1 Vaccine type missing in 4 (0.02%)
2 Vaccine type missing in 3 (0.02%), 36 subjects received homologous vaccine schedules, 2 with BNT162b228 three dose, 6 with ChAdOx1 nCoV-19 three doses, 28 with CoronaVac three doses
3 Vaccine type missing in 1 (0.12%)
4 Among 31,299 who received at least 1 dose and dates of vaccination available
Table 1a Characteristics of subjects included in analysis of association of vaccination with SARS-CoV-2 infection in adults, during delta-predominant period (Oct 1–Dec 31, 2021)
Total (N=27301) SARS-CoV-2 status
Variable Negative (Controls) Positive (Cases) p-value
Number (%) 25171 (92.2) 2130 (7.8) -
Age, years
Median (IQR) 37 (28, 49) 37 (28,49) 37 (26,50) 0.36
Age group, n (%)
18-29 8202 (30.0) 7491 (29.7) 711 (33.4) <0.01
30-39 7147 (26.2) 6667 (26.5) 480 (22.5)
40-49 5334 (19.5) 4948 (19.7) 386 (18.1)
50-59 3707 (13.6) 3465 (13.8) 242 (11.4)
60-69 2097 (7.7) 1905 (7.6) 192 (9.0)
≥70 814 (2.9) 695 (2.7) 119 (5.6)
Gender, n (%)
Male 12745 (46.7) 11693 (46.5) 1052 (49.4) <0.01
Female 14556 (53.3) 13478 (53.6) 1078 (50.6)
Vaccination Status, n (%)
Unvaccinated 9697 (35.5) 8570 (34.0) 1127 (52.9) <0.01
Vaccinated One dose only 3360 (12.3) 3001 (11.9) 359 (16.9)
Vaccinated two doses only 13045 (47.8) 12406 (49.3) 639 (30.0)
Vaccinated three doses 1199 (4.4) 1194 (4.7) 5 (0.2)
Type of primary vaccine series (n=13045), n (%)
SV/SP-AZ 6609 (50.7) 6333 (51.1) 276 (43.2) NA
SV-SV or SP-SP 4770 (36.6) 4485 (36.1) 285 (44.6)
AZ-AZ 1326 (10.2) 1251 (10.1) 75 (11.7)
PFZ-PFZ 194 (1.5) 191 (1.5) 3 (0.5)
AZ-PFZ/Mod 133 (1.0) 133 (1.1) 0
SV/SP-PFZ/Mod 11 (0.1) 11 (0.1) 0
Mod-Mod 2 (0.02) 2 (0.02) 0
Type of third vaccine dose (n=1197)1, n (%)
PFZ 210 (17.6) 210 (17.6) 0 0.40
AZ 879 (73.4) 874 (73.3) 5 (100)
Mod 108 (9.0) 108 (9.1) 0
Median (IQR) time since last vaccination,2days 53 (28,84) 53 (28,84) 48 (29,79) 0.22
Median (IQR) time since last vaccination (two doses only)3 53 (32,79) 53 (32,79) 49 (30,74) 0.35
Vaccination coverage
Completed primary vaccine series, n (%) 14244 (52.2) 13600 (54.0) 644 (30.2) <0.01
Completed third vaccine dose, n (%) 1199 (4.4) 1194 (4.7) 5 (0.2) <0.01
SV=CoronaVac (Sinovac), SP=Sinopharm, AZ=ChAdOx1 nCoV-19 (AstraZeneca), PFZ=BNT162b2 (Pfizer-BioNTech), Mod=mRNA-1273 (Moderna)
1 Vaccine type missing in 2 (0.02%)
2 Date of last vaccination missing in 258 (1.5%)
3 Date of last vaccination missing in 96 (0.7%%)
Vaccine effectiveness
Majority (99.7%) of the boosted subjects received heterologous vaccine schedules. Effectiveness against delta infection was minimal after receiving only a single dose of vaccine. After adjusting for age, gender and calendar week of test, a two-dose primary vaccine series had a VE of 63% (95% CI 59-67%) against delta while a third dose increased the VE to 97% (95%CI 94-99%) (Fig 1 a, Supplementary Table 2a). The VE against delta was consistent across different age groups for either two or three doses. (Fig 1a)Figure 1 a: Effectiveness of two (●) and three (○) dose vaccination regimens against SARS-CoV-2 infection during delta-predominant period (Oct 1–Dec 31, 2021). *Age stratified vaccine effectiveness (VE) adjusted for gender and calendar time. Figure 1b: Effectiveness of three (●) and four (○) dose vaccination regimens against SARS-CoV-2 infection during Omicron-predominant period (Feb 1–Apr 10, 2022). AZ=ChAdOx1 nCoV-19 (AstraZeneca), PFZ=BNT162b2 (Pfizer-BioNTech), Moderna= mRNA-1273 (Moderna), VE = vaccine effectiveness. *Age stratified VE adjusted for gender and calendar time. *VE by vaccine type adjusted for age, gender, calendar time and preceding vaccine series type
During the omicron-predominant period, one or two doses of vaccine provided little to no protection against omicron infection while adjusted VE for a three dose vaccine series was 31% (95% CI 26-36%) and a four dose vaccine series was 75% (95% CI 71-80%) (Fig 1b, Supplementary Table 2b). Three or four dose VE against omicron was consistent across age group 18-50 years but limited case numbers prevented accurate assessment of older ages.
Due to the very small numbers of cases observed, it was possible to calculate only the VE for the ChAdOx1 nCoV-19 vaccine as a third dose in the delta-predominant period with an adjusted VE of 93% (95% CI 82-97) against infection (delta variant). During the omicron-predominant period, adjusted VE for the effectiveness of the third dose did not differ significantly by type of vaccine for the three main vaccines used in Chiang Mai (26-31%) (Fig 1b, Supplementary Table 2b). Similarly, adjusted VE for the effectiveness of the fourth dose also did not differ significantly by type of vaccine (71-73%) (Fig 1b, Supplementary Table 2b).
Discussion
While the number of COVID-19 cases and deaths globally is unacceptably high, the impact of vaccinations is undisputable, when implemented appropriately. As vaccination schedules have rapidly evolved to third and fourth doses, to manage new variants and concerns around waning immunity, the availability of data to support decision makers has struggled to keep pace. The current study provides urgently needed evidence to support the continued rollout booster schedules in Thailand and Asia, and for the first time provides VE for fourth dose schedules incorporating inactivated vaccines into the primary series. Our results corroborate with findings that VE against infection for omicron variant is consistently lower than VE against the delta variant.19 , 21 Heterologous booster vaccines elicit a stronger immune response, but like homologous regimes, antibodies decay over time.22 , 23 From a local context, our findings are consistent with a recent study conducted in Bangkok (Thailand) by Sritipsukho and colleagues evaluating the three dose schedules against the delta variant.24 In that study, the adjusted VE among individuals who received two and three doses of the vaccines was 65% (95% CI, 56–72), and 91% (95% CI 84–95), respectively during the delta-predominant period, which is in agreement with the results of the current study: 63% (two doses) and 97% (three doses).
Previous studies have demonstrated little to no protection against omicron infection for two doses and only mild protection against severe outcomes.21 We also observed minimal protection against omicron infection after one or two doses. We see moderate protection against omicron infection after a third dose (30-40%) and good protection (>70%) after a fourth dose. Recent studies have reported lower hospitalization rates and severe COVID-19 related outcomes due to the omicron variant 25 , 26, this could partly be due to T-cell response which underpin good protection against severe infection and death, irrespective of newer variants of concern.27
The VE of third or fourth vaccine doses against omicron infection were equivalent for the three main vaccines used for boosting in Thailand. Although we were unable to compare VE of third dose against delta by vaccine type, the study by Sritipsukho and colleagues, found comparable protection from a third dose of ChAdOx1 nCoV-19 and BNT162b2.24 The equivalent VE observed for the main booster vaccines used in Thailand is consistent with a recent global analysis of VE data demonstrating equivalence of mRNA and vector vaccines as two dose schedules against earlier variants of concern.13 This strongly suggests that accelerating booster vaccinations and increasing coverage by using any vaccines available, particularly among those aged 60 and older or those with co-morbidities is a valid strategy.
The current study has few limitations. The subjects included in the study were those meeting suspect-case definition and hence, those tested due to epidemiological linkages may or may not have presented with symptoms suggestive of COVID-19. Therefore, the results may not be completely generalisable to those who are symptomatic. In addition to age, gender and calendar time, other key baseline confounders such as chronic comorbidities, socioeconomic status, and prior COVID-19 infection were not examined in this analysis. Due to the variations in the vaccine roll-out, the current study did not differentiate between vaccine eligible population, as compared to those who received recent vaccination and are not yet eligible for next vaccination.
The strengths of the study include the use of harmonised databases with complete record capture for vaccination status. The community-based testing sites provided facilities for people who stayed in the city and surrounding districts where around 40-50% of the COVID-19 cases were reported and is therefore largely representative of the resident population.
The current study provides important findings to support the administration of additional doses of COVID-19 vaccines. Despite the high effectiveness observed for all vaccine booster schedules evaluated, coverage rates in Thailand and in most of the world, are much lower than needed, particularly among the elderly and higher risk segments of the population. Maintaining high vaccination coverage is important in the face of new variants of concern and waning immunity over time. As the pandemic evolves and COVID-19 is eventually declared an endemic disease complacency will likely increase and it may be difficult for governments to continue to support a strong emphasis on preventative measures. As a result, having a high coverage of third booster vaccinations is projected to play a crucial role in decreasing mortality. Our data supports the use of ChAdOx1 nCoV-19, BNT162b2 and mRNA-1273 as booster vaccines, providing much needed flexibility to incorporate different vaccines into schedules according to local supply and logistical considerations.
Contributors:
KI, SC, KC, TW, WK, AT, NC, WT, PK, JW and SI conceptualised the study. SC, KI, and AT led the literature review. KI, SC, KC, TW, WK, AT, NC, WT, PK, JW, and SI contributed towards the methodology. KI did the analysis with the support from SC. SC, KI, AT wrote the initial draft of the manuscript, and all authors critically reviewed the manuscript. All the authors contributed to data collection, curation, validation, and data interpretation. All authors read and approved the final version of the manuscript. KI, SC, and AT had full access of all data in the study. KI and SC contributed equally as joint first authors.
Data sharing statement
All relevant data is available in the paper. Additional requirements, if any will be welcome and de-identified dataset and related codes for analysis will be made available to researchers on request after publication. Requests for data should be addressed to the corresponding author.
Declaration of interests
None.
Supplementary data
Acknowledgments:
This research was supported by the National Research Council of Thailand (NRCT) under The Smart Emergency Care Services Integration (SECSI) project to Faculty of Public Health Chiang Mai University. We are grateful to the Chiang Mai Provincial Health Office and the Department of Disease Control Ministry of Public Health for the collaborative partnerships in managing health information of COVID-19 epidemic.
==== Refs
References:
1 Mathieu E, Ritchie H, Rodés-Guirao L, et al. (2020) - Coronavirus Pandemic (COVID-19). Published online at OurWorldInData.org. Available online at: https://ourworldindata.org/coronavirus [Accessed 9 November 2022].
2 Richards F. Kodjamanova P. Chen X. Economic Burden of COVID-19: A Systematic Review Clinicoecon Outcomes Res 14 2022 293 307 35509962
3 Kaye A.D. Okeagu C.N. Pham A.D. Economic impact of COVID-19 pandemic on healthcare facilities and systems: International perspectives Best Pract Res Clin Anaesthesiol 35 3 2021 293 306 34511220
4 Doroshenko A. The Combined Effect of Vaccination and Nonpharmaceutical Public Health Interventions-Ending the COVID-19 Pandemic JAMA Netw Open 4 6 2021 e2111675
5 Moore S. Hill E.M. Tildesley M.J. Dyson L. Keeling M.J. Vaccination and non-pharmaceutical interventions for COVID-19: a mathematical modelling study Lancet Infect Dis 21 6 2021 793 802 33743847
6 WHO COVID-19 Dashboard. Geneva: World Health Organization, 2022. Available online at https://covid19.who.int/. [Accessed 9 November 2022].
7 WHO COVID-19 Vaccines. Geneva: World Health Organization, 2022. Available online at https://www.who.int/emergencies/diseases/novel-coronavirus-2019/covid-19-vaccines. [Accessed 9 November 2022].
8 Thailand Food and Drug Administration: Medicines Regulation Division, 2022. Available online at https://www.fda.moph.go.th/sites/drug/SitePages/Vaccine_SPC-Name.aspx. [Accessed 9 November 2022].
9 Thailand Department of Disease Control, Available online at https://ddc.moph.go.th/vaccine-covid19/diaryPresentMonth/06/10/2022. [Accessed 9 November 2022].
10 Palacios R. Patiño E.G. de Oliveira Piorelli R. Double-Blind, Randomized, Placebo-Controlled Phase III Clinical Trial to Evaluate the Efficacy and Safety of treating Healthcare Professionals with the Adsorbed COVID-19 (Inactivated) Vaccine Manufactured by Sinovac - PROFISCOV: A structured summary of a study protocol for a randomised controlled trial Trials 21 1 2020 853 33059771
11 Voysey M. Clemens S.A.C. Madhi S.A. Safety and efficacy of the ChAdOx1 nCoV-19 vaccine (AZD1222) against SARS-CoV-2: an interim analysis of four randomised controlled trials in Brazil, South Africa, and the UK Lancet 397 10269 2021 99 111 33306989
12 Polack F.P. Thomas S.J. Kitchin N. Safety and Efficacy of the BNT162b2 mRNA Covid-19 Vaccine N Engl J Med 383 27 2020 2603 2615 33301246
13 Chuenkitmongkol S. Solante R. Burhan E. Expert review on global real-world vaccine effectiveness against SARS-CoV-2 Expert Rev Vaccines 21 9 2022 1255 1268 35748494
14 Atmar R.L. Lyke K.E. Deming M.E. Homologous and Heterologous Covid-19 Booster Vaccinations N Engl J Med 386 11 2022 1046 1057 35081293
15 Barros-Martins J. Hammerschmidt S.I. Cossmann A. Immune responses against SARS-CoV-2 variants after heterologous and homologous ChAdOx1 nCoV-19/BNT162b2 vaccination Nat Med 27 9 2021 1525 1529 34262158
16 Liu X. Shaw R.H. Stuart A.S.V. Safety and immunogenicity of heterologous versus homologous prime-boost schedules with an adenoviral vectored and mRNA COVID-19 vaccine (Com-COV): a single-blind, randomised, non-inferiority trial Lancet 398 10303 2021 856 869 34370971
17 Borobia A.M. Carcas A.J. Pérez-Olmeda M. Immunogenicity and reactogenicity of BNT162b2 booster in ChAdOx1-S-primed participants (CombiVacS): a multicentre, open-label, randomised, controlled, phase 2 trial Lancet 398 10295 2021 121 130 34181880
18 Mayr F.B. Talisa V.B. Shaikh O. Yende S. Butt A.A. Effectiveness of Homologous or Heterologous Covid-19 Boosters in Veterans N Engl J Med 386 14 2022 1375 1377 35139265
19 Accorsi E.K. Britton A. Shang N. Effectiveness of Homologous and Heterologous Covid-19 Boosters against Omicron N Engl J Med 386 25 2022 2433 2435 35613039
20 Intawong K. Olson D. Chariyalertsak S. Application technology to fight the COVID-19 pandemic: Lessons learned in Thailand Biochem Biophys Res Commun 538 2021 231 237 33589143
21 UK Health Security Agency. COVID-19 vaccine surveillance report. Week 24. Available online at https://www.gov.uk/government/publications/covid-19-vaccine-weekly-surveillance-reports. [Accessed 9 November 2022].
22 Ascaso-Del-Rio A. García-Pérez J. Pérez-Olmeda M. Immune response and reactogenicity after immunization with two-doses of an experimental COVID-19 vaccine (CVnCOV) followed by a third-fourth shot with a standard mRNA vaccine (BNT162b2): RescueVacs multicenter cohort study EClinicalMedicine 51 2022 101542
23 García-Pérez J. González-Pérez M. Castillo de la Osa M. Immunogenic dynamics and SARS-CoV-2 variant neutralisation of the heterologous ChAdOx1-S/BNT162b2 vaccination: Secondary analysis of the randomised CombiVacS study EClinicalMedicine 50 2022 101529
24 Sritipsukho P. Khawcharoenporn T. Siribumrungwong B. Comparing real-life effectiveness of various COVID-19 vaccine regimens during the delta variant-dominant pandemic: a test-negative case-control study Emerg Microbes Infect 11 1 2022 585 592 35114893
25 Veneti L. Bøås H. Bråthen Kristoffersen A. Reduced risk of hospitalisation among reported COVID-19 cases infected with the SARS-CoV-2 Omicron BA.1 variant compared with the Delta variant, Norway, December 2021 to January 2022 Euro Surveill 27 4 2022 2200077
26 Nyberg T. Ferguson N.M. Nash S.G. Comparative analysis of the risks of hospitalisation and death associated with SARS-CoV-2 omicron (B.1.1.529) and delta (B.1.617.2) variants in England: a cohort study Lancet 399 10332 2022 1303 1312 35305296
27 Moss P. The T cell immune response against SARS-CoV-2 Nat Immunol 23 2 2022 186 193 35105982
| 36465090 | PMC9705195 | NO-CC CODE | 2022-12-03 23:16:09 | no | Lancet Reg Health Southeast Asia. 2022 Nov 29;:100121 | utf-8 | Lancet Reg Health Southeast Asia | 2,022 | 10.1016/j.lansea.2022.100121 | oa_other |
==== Front
Epilepsy Behav
Epilepsy Behav
Epilepsy & Behavior
1525-5050
1525-5069
Elsevier Inc.
S1525-5050(22)00473-5
10.1016/j.yebeh.2022.109024
109024
Brief Communication
COVID-19 vaccination-related exacerbation of seizures in persons with epilepsy
Pang E.W. ab⁎
Lawn N.D. ab
Chan J. ab
Lee J. a
Dunne J.W. ac
a Western Australian Adult Epilepsy Service, Perth, Western Australia
b Neurology Department, Fiona Stanley Hospital, Murdoch, Western Australia
c Discipline of Internal Medicine, Medical School, The University of Western Australia, Perth, Western Australia
⁎ Corresponding author at: Western Australian Adult Epilepsy Service, Sir Charles Gairdner Hospital, Hospital Avenue, Nedlands, Western Australia 6009, Australia.Presented at the Australian and New Zealand Association of Neurologist Annual Scientific Meeting, Melbourne, 10-13 May 2022.
29 11 2022
29 11 2022
10902411 9 2022
17 11 2022
23 11 2022
© 2022 Elsevier Inc. All rights reserved.
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Although vaccines are generally safe in persons with epilepsy (PWE), seizures can be associated with vaccination, including COVID-19. This study assessed the occurrence of COVID-19 vaccination-related seizure exacerbations in PWE.
Adult PWE who had received a COVID-19 vaccine were consecutively recruited at a tertiary epilepsy clinic between June 2021 and April 2022. Patient demographics, including epilepsy history, vaccination details and reported adverse effects were recorded. Seizure exacerbation, defined as occurring within one week of vaccination, was assessed.
530 PWE received the COVID-19 vaccine. 75% received the Comirnaty (Pfizer) vaccine as their initial dose. Most patients (72%) were taking ≥2 antiseizure medications (ASM) and had focal epilepsy (73%). One third were 12-months seizure free at their first vaccination. 13 patients (2.5%) reported seizure exacerbation following their first vaccination, three of whom required admission. None were seizure-free at baseline. Six of these patients (46%) had a further exacerbation of seizures with their second vaccine. An additional four patients reported increased seizures only with the second vaccine dose.
Seizure exacerbations are infrequently associated with COVID-19 vaccination, mainly in patients with ongoing seizures. The likelihood of COVID-19 infection complications in PWE clearly outweighs the risk of vaccination-related seizure exacerbations.
Keywords
COVID-19
Vaccination
Epilepsy
Seizure exacerbation
==== Body
pmc1 Introduction
Twelve months after the national coronavirus-19 (COVID-19) vaccination rollout in Australia, over 94% of the population over 16 years old were fully vaccinated.[1] Some people have reservations about the vaccination, including patients with epilepsy (PWE) despite the established safety of similar vaccinations in PWE.[2] There are multiple factors that may have fuelled vaccination hesitancy, including safety concerns, lack of trust in the information provided by government and pharmaceutical companies as well as conflicting opinions on social media platforms.[3], [4] For PWE, one of the most common concerns contributing to vaccination hesitancy has been the potential for seizure exacerbation.[5], [6], [7]
Several small reports have assessed the safety of the COVID-19 vaccine in PWE over the past 12 months. Two initial studies from Germany and Kuwait surveyed 52 and 82 patients respectively after their first COVID-19 vaccine within a month of the vaccine becoming available.[5], [6] Neither study demonstrated significant vaccine-related seizure exacerbations. [5], [6] The few patients who had a vaccine-related seizure exacerbation were predominantly female, older and on multiple antiseizure medications (ASM). However, given the small sample, the significance of these factors could not be assessed. [5], [6] A recent study from China assessed COVID-19 vaccine take-up and vaccine side effects, including seizure exacerbation, in PWE compared to healthy controls and patients with neuropsychiatric comorbidities.[7] Only 204 of 491 PWE were vaccinated. Whilst the incidence of adverse effects comparing PWE and controls was no different, 19 patients (9.3%) reported an exacerbation of seizures, but the details provided were limited. [7]
Given ongoing uncertainty and the importance of providing COVID-19 vaccination to PWE we aimed to assess the likelihood of COVID-19 vaccination-related seizure exacerbations in PWE and identify possible predictors.
2 Materials and Methods
PWE aged ≥18 years who had received at least one COVID-19 vaccine were consecutively recruited from a tertiary hospital epilepsy clinic in Perth, Western Australia, between June 2021 and April 2022. All enrolled patients had confirmed epilepsy based on clinical assessment and investigations, including neuroimaging (CT and/or MRI brain), routine EEG and in many cases prolonged video EEG monitoring prior to their first vaccine dose. Patients who did not have epilepsy were excluded. However, patients who had known epilepsy but concurrent psychogenic non-epileptic spells (PNES) clearly distinguisable from their epileptic seizures were included. Patient demographics, including age, gender, epilepsy characteristics, baseline seizure frequency and current antiseizure medications (ASMs) were recorded.
A paper-based survey assessing COVID-19 vaccination status and associated complications, including vaccine-related seizure exacerbations, was conducted during the in-person routine clinic appointments and was completed by the patient and/or their support person (in cases where the patient was unable to complete the survey; i.e. intellectual disability) and the treating clinician. For patients requiring telephone-based appointments, often due to COVID restrictions or remote rural settings, the treating clinician conducted the survey via phone call and completed the paper-based form. Additional information about seizure-related admission was obtained through the computerised state-wide public hospital medical records system. The majority of patients documented their vaccine doses on the initial survey during a single clinic attendance. For patients who had multiple clinic reviews during the data-collection period and obtain sequential vaccine doses after their first survey, a second survey was completed to capture relevant information regarding their subsequent doses.
Seizure exacerbations were documented based on patient’s self-reported seizures (i.e. survey results and clinic reviews) or hospital presentations with seizures, seizure-related injuries or status epilepticus. They were defined as an increase in seizures within one week of vaccination compared to the patient’s baseline seizure frequency by utilising seizure diaries and as determined by the treating clinician. Occurrence of other seizure-like events, including syncope and PNES, which were clarified by clinical assessment and where necessary EEG, was not considered a seizure exacerbation. Patients were also excluded if other external factors unrelated to vaccination were likely responsible for the seizure exacerbation (e.g. ASM non-compliance), based on clinical assessment by the treating epileptologist, at the time the survey was conducted. If a patient had a vaccine-associated seizure exacerbation, the timing and nature of increased seizures, treatment requirement and need for hospital presentation or admission were recorded. The demographics and clinical features of PWE who developed COVID-19 vaccine-related seizure exacerbations were compared to those without seizure excerbation.
Comparisons between groups were conducted using t tests for normally distributed data and Mann-Whitney tests for non-normal data. Chi-square and Fisher exact tests were used for categorical data.
This study was approved by the Fiona Stanley Hospital Human Research Ethics Committee. Verbal consent was obtained from all patients.(Table 1 )Table 2. Table 1 Demographics of patients with and without seizure exacerbation after first COVID vaccination
No seizure exacerbationN=517 Seizure exacerbationN=13 p value
Sex: female n (%) 267 (52) 9 (69) 0.27
Median age at time of first dose, yrs (range) 38 (16-84) 37 (17-62) 0.54
Epilepsy type n (%)- focal
- generalised
- undifferentiated
377 (73)135 (26)5 (1) 9 (69)4 (31)- 0.88
Symptomatic generalised epilepsy n (%) 141 (27) 4 (31) 0.94
12 months seizure free at time of first dose 177 (33) 0 0.006
Seizure frequency at time of first dose- ≥1 seizure/day
- ≥1 seizure/mo
- ≥1 seizure/year
52 (10)161 (31)131 (25) 4 (31)5 (38)4 (31) 0.35
Median no. of ASM at time of first dose 2 3 0.16
2 or more ASM at time of first dose n (%) 372 (72) 11 (85) 0.53
3 or more ASM at time of first dose n (%) 225 (44) 8 (61) 0.26
Brand of first dose- Pfizer Comirnaty
- Astra Zeneca
- Moderna
- Novavax
391 (75)108 (21)17 (3)1 (<1) 10 (77)2 (15)1 (8)- 0.93
Table 2 Vaccine-associated seizure exacerbation after 1st and 2nd vaccine doses
Seizure exacerbation after 1st vaccination Seizure exacerbation after 2nd vaccination
Number of patients 13 patients 10 patients
Median time from vaccine to seizure (range) 1 day (0-3) 1 day (0-7)
Seizure on day of vaccinen (%) 4 (31) 5 (50)
Other vaccine side effectsn (%) 5 (38) 4 (40)
Seizure clustern (%) 6 (46) 4 (40)
Status epilepticusn (%) 1(8) 0
Exacerbation with 2nd vaccination n (%) 6 (46) -
3 Results
3.1 Patient demographics and clinical features
530 PWE who had received the COVID-19 vaccine were identified over the 10-month recruitment period. Most patients (73%) had focal epilepsy and 145 patients (27%) had symptomatic generalised epilepsy (SGE). 177 patients (33%) had been seizure free for 12 months at the time of their first vaccination. 222 of 353 patients (63%) with ongoing seizures at baseline had multiple seizures a month. Median number of ASM was 2, with 72% of patients taking 2 or more ASM.
516 (97%) received two COVID-19 vaccine doses. 14 patients only received a single dose for various reasons, including wait required for second dose at time of survey and patient choice. 400 patients (75%) of patients received the Pfizer Comirnaty vaccine as their initial dose, reflecting vaccine availability in Western Australia. Median age at first vaccine was 38 years (range 16 – 84 years, interquartile range 25).
4 Seizure exacerbation after first COVID-19 vaccine dose
13 patients (2.5%) reported a clear exacerbation of seizures following their first vaccination. None were seizure free at the time of their initial vaccination, compared to 34% of the control cohort (p=0.006). Their demographics were otherwise similar to those without seizure exacerbation, including vaccination brand and number of ASM (Table 1).
Ten of the 13 patients (77%) with a seizure exacerbation reported increasing seizures within 24 hours of their first vaccine dose, and the remainder within 72 hours. Six patients had a cluster of multiple seizures, with five patients reporting several consecutive days of increased recurrent seizures. One patient developed convulsive status epilepticus, which terminated with out-of-hospital midazolam prior to arrival to in the emergency department (ED). No patient required intensive care admission.
Five patients (38%) presented to hospital because of seizure exacerbation post vaccination: two were discharged from ED after a short period of observation and three patients required hospital admission for 1 to 3 days. One patient sustained a seizure-related injury (foot fracture) which was conservatively managed. ASM were escalated in only two patients, with increases of their regular ASM dosing and commencement of a short course of additional clobazam.
Five of the 13 (38%) had concurrent vaccination-related side effects including fatigue, fever and headache. This was similar to the systemic side effect rate of the whole cohort (30.6%) One patient with type 1 diabetes mellitus reported vomiting and diarrhoea leading to metabolic disturbance and hypoglycaemia, contributing to their seizure exacerbation.
5 Seizure exacerbation after second COVID-19 vaccine dose
All 13 patients with seizure exacerbation after the first vaccination received a second vaccination. Six of the 13 patients (46%) reported an increase of seizures after their second vaccination, including the single patient who had received pre-emptive clobazam. Median time from second vaccine to seizure exacerbation was 1 day (range 1-7 days), with 4 of 6 patients having a seizure increase within 24 hours of their second vaccine dose. Only one patient reported an increase in seizures occurring over multiple days. None of these patients presented to hospital or required ASM escalation to manage their seizures.
A further four patients reported a vaccine-related seizure exacerbation only occurring after their second dose. The clinical features of these patients were similar to those who reported seizure exacerbations after the first vaccine; none were seizure-free at baseline. Median time from vaccine to seizure was 3.5 days (range 0 – 7 days) with half reporting seizure exacerbation within 24 hours of their second vaccine. Three of these four patients reported concurrent vaccine side effects, mild in all but one with high fevers and hypotension. Two patients (50%) presented to hospital, with one requiring multiday admission and ASM escalation.
6 Discussion
Of 1,046 COVID-19 vaccine doses administered in 530 PWE (530 initial doses and 516 second doses), only 23 (2%) vaccine-associated seizure exacerbations occurred, exclusively in patients with ongoing seizures. The only predictor of not developing vaccine-associated seizures post COVID vaccine was 12 months seizure freedom at time of first dose. To date, this is the largest study assessing COVID-19 vaccine-associated seizure exacerbation in PWE.
The risk of increased seizures after vaccination is small, compared to the previous studies from Germany, China and Kuwait that reported seizure exacerbation in 1.8 to 16% of patients.[5], [6], [7] These differences may relate to differing sample sizes and study populations. The combined number of patients from these studies is 338 patients compared to our cohort of 530 patients, and we also examined seizure exacerbations after second vaccination.[5], [6], [7] Specific groups of PWE may differ in their vulnerability, for example, a study of 120 patients with Dravet Syndrome found 13% of patients had a self-reported vaccine-associated seizure exacerbation.[8]
This study demonstrates a small risk of seizure exacerbation due to the COVID-19 vaccinations, in contrast to the much higher risks associated with COVID-19 infection itself.[9] PWE are more likely to develop severe complications from COVID-19, including the requirement for mechanical ventilation and ICU admission, and death.[9], [10] Whilst COVID-19 may not worsen seizures in PWE directly, seizures can be triggered by fevers or other systemic factors, as with other infections.[11] Furthermore, COVID-19 may be associated with hypoxia, stroke, systemic inflammatory response syndrome and encephalitis, all capable of precipitating acute symptomatic seizures in already vulnerable patients.[8] Therefore the risk of neurological and systemic complications of COVID infections far outweighs the low risk of vaccine-associated seizure exacerbation.
Our study has limitations. First, most patients had relatively refractory epilepsy, typical of a tertiary hospital epilepsy clinic. The risk of vaccine-associated seizure exacerbations in the general population of PWE may be lower than in our patients. Since most of our patients were not seizure free and required polytherapy with ≥2 ASMs, our data is relevant to drug resistant PWE. Only a small number of our patients had not received a COVID vaccine at the time of the survey and therefore a comparison to assess spontaneous fluctuations in seizure frequency was not possible. We were only able to compare seizure exacerbation post vaccination with each patient’s historical baseline seizure frequencies prior to their vaccination. Nonetheless, if anything, our results, which demonstrates a low risk of seizure COVID-19 vaccine-associated seizure exacerbation risk and highlights vaccine safety in this population, may be an over-estimation of seizure exacerbation risk, providing further reassurance for patients and clinicians when counselling about vaccine safety. The most common reason for lack of vaccination were concerns about vaccine safety and exclusion of these patients may have introduced an element of selection bias. Data regarding previous seizure exacerbations associated with other vaccinations (e.g. influenza vaccination) was not obtained. This may have influenced patients’ decisions to receive the COVID vaccine and information on this may have helped delineate whether the exacerbations were COVID vaccine-specific or a complication of vaccinations in PWE as a whole. In addition, the majority (75%) of our patients received the Pfizer Comirnaty vaccination, limiting generalisability of our findings to other COVID-19 vaccines. As with other studies, we have relied on patient self-reported seizure increases which may be vulnerable to recall bias. Pre-emptive escalation of ASM prior to the first COVID-19 vaccine dose was not explored, nor the potential protective role of this measure subsequently.
7 Conclusions
In summary our findings indicate very low likelihood of COVID-19 vaccine-related seizure exacerbations in PWE, usually occurring in patients with ongoing seizures, and major sequelae are very uncommon. This data can be utilised in counselling PWE regarding the safety of COVID-19 vaccination.
8 Other authors
Nicholas Lawn, Neurology Department, Fiona Stanley Hospital, Murdoch, Western Australia
Judy Lee, Western Australian Adult Epilepsy Service, Sir Charles Gairdner Hospital, Nedlands, Western Australia
Josephine Chan, Neurology Department, Fiona Stanley Hospital, Murdoch, Western Australia
John Dunne, Discipline of Internal Medicine, Medical School, Royal Perth Hospital Unit, The University of Western Australia, Perth, Western Australia
==== Refs
References
1 Department of Health, Australian Government. COVID-19 vaccine rollout update – 28 February 2022 [Internet]. Canberra (AU); 2022 [updated 2022; cited 2022 March 02]. Available from: https://www.health.gov.au/resources/publications/covid-19-vaccine-rollout-update-28-february-2022
2 Top K. Brna P. Ye L. Smith B. Risk of seizures after immunization in children with epilepsy: a risk interval analysis BMC Pediatrics 18 2018 134 10.1186/s12887-018-1112-0 29642863
3 Kumar S. Shah Z. Garfield S. Causes of Vaccine Hesitancy in Adults for the Influenza and COVID-19 Vaccines: A Systematic Literature Review Vaccines 10 2022 1518 10.3390/vaccines10091518 36146596
4 Mascherini M. Nivakoski S. Social media use and vaccine hesitancy in the European Union Vaccine. 40 14 2022 2215 2225 35249775
5 von Wrede R. Pukropski J. Moskau-Hartmann S. Surges R. Baumgartner T. COVID-19 vaccination in patients with epilepsy: First experiences in a German tertiary epilepsy center Epilepsy and Behaviour 122 2021 108160
6 Massoud F. Ahmad S.F. Hassan A.M. Alexander K.J. Al–Hashel J. Arabi M. Safety and tolerability of the novel 2019 coronavirus disease (COVID-19) vaccines among people with epilepsy (PWE): A cross-sectional study Seizure 92 2021 2 9 34391030
7 Lu L.u. Zhang Q.i. Xiao J. Zhang Y. Peng W. Han X. COVID-19 vaccine take-up rate and safety in adults with epilepsy: Data from a multicenter study in China Epilepsia 63 1 2022 244 251 34806164
8 Hood V. Berg A.T. Knupp K.G. Koh S. Laux L. Meskis M.A. COVID-19 vaccine in patients with Dravet syndrome: Observations and real-world experiences Epilepsia. 63 7 2022 1778 1786 35383912
9 Yoo J. Kim J. Jeon J. Kim J. Song T. Risk of COVID-19 Infection and of Severe Complications Among People With Epilepsy: A Nationwide Cohort Study Neurology 98 2022 e1886 e1892 10.1212/WNL.000000000020019 35338078
10 Cabezudo-García P. Ciano-Petersen N.L. Mena-Vázquez N. Pons-Pons G. Castro-Sánchez M.V. Serrano-Castro P.J. Incidence and case fatality rate of COVID-19 in patients with active epilepsy Neurology 95 10 2020 e1417 e1425 32554773
11 Vohora D. Jain S. Tripathi M. Potschka H. COVID-19 and seizures: Is there a link? Epilepsia 61 1840–1853 2020 10.1111/epi.16656
| 36495798 | PMC9705196 | NO-CC CODE | 2022-12-12 23:20:29 | no | Epilepsy Behav. 2023 Jan 29; 138:109024 | utf-8 | Epilepsy Behav | 2,022 | 10.1016/j.yebeh.2022.109024 | oa_other |
==== Front
Clin Chest Med
Clin Chest Med
Clinics in Chest Medicine
0272-5231
1557-8216
Published by Elsevier Inc.
S0272-5231(22)00132-0
10.1016/j.ccm.2022.11.016
Article
The Use of ECMO for COVID-19: Lessons Learned
Parekh Madhavi MD 1#
Abrams Darryl MD 1
Agerstrand Cara MD 1
Badulak Jenelle MD 2
Dzierba Amy PharmD 3
Alexander Peta M.A. MBBS 45
Price Susanna MBBS 67
Fan Eddy MD, PhD 8
Mullin Dana MS, CCP, LP 9
Diaz Rodrigo MD 10
Hodgson Carol PhD 11∗
Brodie Daniel MD 1∗
1 Division of Pulmonary, Allergy, and Critical Care Medicine, Department of Medicine, Columbia University Vagelos College of Physicians and Surgeons, New York, New York, USA
2 Department of Emergency Medicine and Division of Pulmonary, Critical Care and Sleep Medicine, University of Washington, Seattle, Washington, USA
3 Department of Pharmacy, NewYork-Presbyterian Hospital, New York, New York, USA
4 Department of Cardiology, Boston Children's Hospital, Boston, Massachusetts, USA
5 Department of Pediatrics, Harvard Medical School, Boston, Massachusetts, USA
6 Adult Intensive Care Unit, Royal Brompton Hospital, London, UK
7 National Heart and Lung Institute, Imperial College, London, UK
8 Interdepartmental Division of Critical Care Medicine, University of Toronto, Toronto, ON, Canada
9 Department of Clinical Perfusion and Anesthesia Support Services, NewYork-Presbyterian Hospital, New York, New York, USA
10 Clinica Las Condes, Santiago, Chile
11 Australian and New Zealand Intensive Care Research Centre, Department of Epidemiology and Preventive Medicine, Monash University, Melbourne, Australia
# Corresponding author
∗ Co-senior authors
29 11 2022
29 11 2022
© 2022 Published by Elsevier Inc.
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Synopsis
The coronavirus disease 2019 (COVID-19) pandemic has seen an increase in global cases of severe acute respiratory distress syndrome (ARDS), with a concomitant increased demand for extracorporeal membrane oxygenation (ECMO). Outcomes of patients with severe ARDS due to COVID-19 infection receiving ECMO support are evolving. The need for surge capacity, practical and ethical limitations on implementing ECMO, and the prolonged duration of ECMO support in patients with COVID-19-related ARDS, has revealed limitations in organization and resource utilization. Coordination of efforts at multiple levels, from research to implementation, resulted in numerous innovations in the delivery of ECMO.
Key Words
ECMO
acute respiratory distress syndrome
ARDS
COVID-19
Respiratory Failure
==== Body
pmcKey Points
• The COVID-19 pandemic has posed challenges on multiple levels in the implementation of ECMO support for patients with severe ARDS around the world
• Real-time data gathering and new approaches to research have helped evaluate the utility of ECMO during an ongoing pandemic
• Regional, national and international coordination has been crucial in knowledge sharing, research collaboration and development of guidelines
Introduction
The use of extracorporeal membrane oxygenation (ECMO) in the management of severe acute respiratory distress syndrome (ARDS) has been well established in recent years,1, 2, 3, 4 but the ongoing coronavirus disease 2019 (COVID-19) pandemic has presented new limitations in knowledge and has complicated the implementation of ECMO. Severe COVID-19 frequently presents with acute respiratory failure in the form of ARDS, and the pandemic has been characterized by surges in the volume of critically ill patients with ARDS worldwide.
Episodic surges in patients with COVID-19-related ARDS have been accompanied by strain both on healthcare resources (e.g. beds, staffing, medical supplies) and in the ability to provide equitable access to care, all of which is further exaggerated when considering the application of ECMO, a highly resource-intensive and specialized technology.
As the pandemic has evolved, the medical community has learned not only about the role of ECMO for COVID-19 from a clinical standpoint, but also how to gather knowledge in real-time about the use of ECMO for a novel disease, the limitations in our ability to equitably deliver healthcare resources across the globe, and how to devise best strategies for care in light of substantial resource constraints. The use of ECMO for cardiac or circulatory failure, including for extracorporeal cardiopulmonary resuscitation (ECPR) has been relatively limited in the setting of COVID-19; we will focus on the use of ECMO for respiratory failure.
Early Experience
Registry data and large cohort studies from early in the pandemic suggested patients managed with ECMO for COVID-19-related ARDS had a mortality rate comparable to similar patients with ARDS prior to COVID-19 (Table 1 ); mortality rates of 31% at 60 days in a large single-center experience from Paris, France,5 33.2% at 60-days in a cohort study across 60 hospitals in the US,6 and an estimated 37.4% in-hospital mortality at 90-days from the Extracorporeal Life Support Organization (ELSO) Registry, including 213 hospitals across 36 countries.7 A meta-analysis of studies spanning December 1, 2019, through January 10, 2021, tallying 1896 patients, reported an in-hospital mortality rate of 37.1%;8 these analyses were heavily weighted by the aforementioned large cohort studies.Table 1 Early experience of ECMO for COVID-19-related ARDS
Data Source Timeframe Number of Patients Mortality
Single-center observational experience from Paris, France5 Patients admitted between March 8th-May 2nd, 2020 83 31% at 60 days
Cohort study across 60 hospitals in the United States6 Patients admitted between March 1st and July 1st, 2020 190 33.2% at 60-days
Extracorporeal Life Support Organization (ELSO) registry, including 213 hospitals across 36 countries7 Patients in whom ECMO was initiated between Jan 16th and May 1st, 2020 1035 37.4% in-hospital mortality at 90-days
Meta-analysis of 22 studies8 December 1st, 2019, through January 10th, 2021 1896 37.1% in-hospital mortality
Evolving Mortality Over Time
Despite encouraging data early on, additional data gathered as the pandemic continued suggested increasing mortality and duration of ECMO over time. An analysis of a survey from the European chapter of ELSO (EuroELSO) found a mortality of 56% for COVID-19 patients managed with ECMO between September 15th, 2020, and March 8th, 2021, as compared to 47% prior to that time period.9 Similarly, data from 24 centers in Spain and Portugal suggested a higher in-hospital mortality after June 30th, 2020 (60.1%) than before that date (41.1%).10 The Paris-Sorbonne University Hospital Network found a 90-day mortality of 48% in patients after July 1st, 2020, as compared to 36% prior to July (HR 2.27, 95% CI 1.02-5.07).11 Further analysis of the ELSO registry database, encompassing 4,812 patients, also noted a higher 90-day in-hospital mortality for patients after May 1st, 2020 as compared to earlier (51.9% vs 36.9%, RR 0.82, 95% CI 0.7-0.96), with a longer duration of ECMO support in the latter cohort (20 days vs. 14 days).12 Data published from Germany demonstrated a high in-hospital mortality (68%) for all patients supported with ECMO (n=3,397) for COVID-19-related ARDS from the start of the pandemic through May 31st, 2021, despite a lack of resource constraints.13 An updated meta-analysis of 52 studies (18,211 patients) reporting data between December 1st, 2019, to January 26th, 2022, revealed a pooled mortality rate of 48.8% (95% CI 44.8-52.9%) among patients with COVID-19 receiving ECMO, with increasing mortality in the second half of 2020 (46.4%) compared with the first half (41.2%), and an even higher mortality (62%) in the first half of 2021. Predictors of increased mortality included age, later time of enrollment, higher proportion of patients receiving corticosteroids, and reduced duration of ECMO run.14
A number of potential reasons for increasing mortality over time have been speculated, including: greater selection over time for treatment-refractory disease (those who did not respond to COVID-19-directed therapies that were increasingly used over the course of the pandemic, e.g. corticosteroid therapy),9, 10, 11, 12 increased use of non-invasive ventilatory support prior to intubation and ECMO (which may contribute to pre-endotracheal intubation self-inflicted lung injury),10, 11, 12 a rise in superimposed bacterial pneumonia in the setting of immunosuppressive treatments for COVID-19,10,15,16 emergence of SARS-CoV-2 variants with differential effects on prognosis, increased use of ECMO by less experienced centers,10 , 12 and variations in patient selection criteria for ECMO.10
While the mortality of patients managed with ECMO may have increased over the course of the pandemic, ECMO may still benefit selected patients with severe COVID-19-related ARDS.17, 18, 19 A multi-center international emulation trial, which applies principles of randomized controlled trials to the analysis of observational data, including 7,345 patients between January 3rd, 2020, and August 29th, 2021, found a reduction in 60-day in-hospital mortality with a risk ratio of 0.78 (95% CI 0.75-0.82). Adherence adjusted mortality, which accounts for adherence to the treatment assignment, was 26% for patients managed with ECMO as compared to conventional treatment (33.2%). Secondary analyses suggested ECMO was most effective in patients aged <65, those with a PaO2 to FiO2 ratio of <80 mmHg, a driving pressure >15 cmH2O, or during the first 10 days of mechanical ventilation. While this was not a traditional RCT, the emulation design allowed for a more rigorous analysis of real-world effectiveness of ECMO than a traditional observational study.17 Additionally, a study conducted in the UK demonstrated an absolute mortality reduction of 18.2% (44% vs 25.8%; OR 0.44; 95% CI 0.29-0.68, p<0.001) for those receiving ECMO compared to matched controls.18 These data should be interpreted cautiously, as residual confounding may have accounted for some of the benefit, especially as those receiving ECMO were managed at highly specialized centers, whereas those not offered ECMO remained in their original facilities for ongoing care.
The true efficacy of ECMO for severe COVID-19-related ARDS remains uncertain in the absence of high-quality, randomized controlled trials (RCTs). The fact that RCTs could not be implemented despite thousands of ECMO cases highlights the challenges faced in conducting research, especially RCTs, during a constantly changing pandemic with substantial limitations in infrastructure and resources, including staffing and time.20 Ongoing evaluation of emerging data will be necessary to help determine optimal patient selection and management strategies; until then, use of conventional inclusion and exclusion criteria based on pre-COVID ECMO data,1 , 21 modified by factors identified in large registry analyses to be predictive of outcomes, appears to be a reasonable approach.1 , 22
Clinical Care
Patient Selection
Criteria for ECMO initiation in patients with COVID-19-related ARDS remain the same as those recommended prior to the pandemic,21, 22, 23, 24, 25 and fall within the standard approach to ARDS algorithm (Figure 1 ), as there is no good evidence to support deviation from these preestablished guidelines when resources are available. Outcomes with delayed initiation may in fact be worse and can lead to longer duration of ECMO support, which may in turn offset the benefit of an attempt of conservation of resources. However, resource constraints during the pandemic have overwhelmed the ability to provide ECMO at varying times in different regions of the world. As such, selection criteria may need to be more flexible and potentially stringent at any given time, depending on available resources and coordination locally. Criteria may need to evolve as well, as increasing knowledge arises regarding prognostic factors.26 Figure 1 Algorithm flowsheet for management of ARDS including indications for ECMO PEEP=positive end-expiratory pressure. PaO2:FiO2=ratio of partial pressure of oxygen in arterial blood to the fractional concentration of oxygen in inspired air. ECMO=extracorporeal membrane oxygenation. PaCO2=partial pressure of carbon dioxide in arterial blood. ∗With respiratory rate increased to 35 breaths per minute and mechanical ventilation settings adjusted to keep a plateau airway pressure of ≤32 cm of water. †Consider neuromuscular blockade. ‡There are no absolute contraindications that are agreed upon except end-stage respiratory failure when lung transplantation will not be considered; exclusion criteria used in the EOLIA trial1 can be taken as a conservative approach to contraindications to ECMO. §Eg, neuromuscular blockade, high PEEP strategy, inhaled pulmonary vasodilators, recruitment maneuvers, high-frequency oscillatory ventilation. ¶Recommend early ECMO as per EOLIA trial criteria; salvage ECMO, which involves deferral of ECMO initiation until further decompensation (as in the crossovers to ECMO in the EOLIA control group), is not supported by the evidence but might be preferable to not initiating ECMO at all in such patients.From Abrams D, Ferguson ND, Brochard L, et al. ECMO for ARDS: from salvage to standard of care? [published correction appears in Lancet Respir Med. 2019 Feb;7(2):e9]. Lancet Respir Med. 2019;7(2):108-110.
Cannulation and Transport
Conventional cannulation strategies are generally recommended for patients undergoing ECMO initiation for severe COVID-19-related ARDS as there is a paucity of data to support alternative strategies.22 These include two-site or single-site, dual-lumen venovenous cannulation (VV), with additional arterial support depending on whether there is concomitant cardiogenic shock.23 Some centers have utilized a veno-pulmonary artery (V-P) configuration through a single dual-lumen catheter inserted via the internal jugular or subclavian vein in an attempt to provide right ventricular protection,27 as some reports suggest a higher incidence of right ventricular dysfunction in patients with COVID-19-related ARDS,28 but more data is needed to support this strategy either with a dual-lumen cannula or with dual-site cannulation with two separate cannulae. Mobile ECMO or ECMO transport with cannulation at the originating hospital and transfer to another institution appears to be safe from a healthcare exposure standpoint when accompanied by protocols incorporating adequate protective measures.29 , 30 However, it is important to note that the coordination and feasibility of ECMO transport varies across regions and, consequently, rates of transport do as well. This may be due in part to differences in practice, training, certification and regulation.
Ongoing Care while Receiving ECMO
Management strategies of patients requiring ECMO support for COVID-19-related ARDS remain similar to those recommended prior to the pandemic (Figure 1), but specific factors should be considered (Figure 2 ).23 , 31 From a safety and feasibility standpoint, a number of procedures and techniques have been successfully performed during the course of the pandemic. Endotracheal extubation while awake during ECMO has been performed, although data is very limited;27 in contrast, awake ECMO without intubation in selected highly patients has been associated with potentially worse outcomes in one small cohort.32 Prone positioning5 , 33 , 34 and early mobilization27 appear feasible in patients with COVID-19 requiring ECMO support, however data remains too limited to support specific recommendations. Percutaneous tracheostomy appears to be safe in patients with COVID-19.35 From an infection control standpoint, there is no evidence to suggest virions can be expelled through an ECMO circuit.36 Cytokine removal devices, which utilize an absorber to remove excess cytokines from whole blood, have been proposed as adjunctive therapies in patients receiving ECMO for COVID-19-related ARDS. However, clinicians should proceed cautiously and consider only using cytokine removal devices in such patients in the setting of research, given that recent evidence suggests possible harm.37 Figure 2 Specific considerations for ECMO for COVID-19-related ARDS that may differ from ECMO for non-COVID-19 ARDS RV= right ventricular From Brodie D, Abrams D, MacLaren G, et al. Extracorporeal Membrane Oxygenation during Respiratory Pandemics: Past, Present, and Future. Am J Respir Crit Care Med. 2022;205(12):1382-1390.
COVID-19 has been associated with coagulopathy, including an increase in risk of both significant thrombosis and bleeding.38 , 39 Hematological complications – including circuit clotting,40 , 41 pulmonary embolism5 and intracranial hemorrhage42, 43, 44, 45 – have been reported as occurring more frequently in patients with COVID-19 supported with ECMO than in non-COVID ECMO cases, but, when normalized to ECMO run duration, such complication rates appear similar to historical data.12 In the setting of observational data, the similar normalized rates must be interpreted with caution. Multiple centers have adjusted their anticoagulation thresholds, but there are insufficient data to support anticoagulation strategies and monitoring other than usual practices.46 Additionally, there is no evidence to suggest different blood transfusion thresholds for patients with COVID-19 during ECMO support.47
ECMO as a Bridge to Lung Transplantation
COVID-19-related ARDS has often demonstrated a more prolonged recovery process than what is typically seen for ARDS of other etiologies, and the duration of ECMO support for patients with severe disease has increased as the pandemic has evolved.12 Lung transplantation may be considered in select patients who have persistent severe respiratory failure – assuming otherwise appropriate candidacy with preserved extra-pulmonary organ function – and may be particularly relevant for those who are unable to wean from ECMO, with initial reports demonstrating post-transplant outcomes comparable to those with non-COVID end-stage lung disease.48 , 49 However, determination of the potential for recovery of native lung function and optimal timing of transplantation remain areas of uncertainty that warrant further investigation.50 , 51
Surge Capacity and ECMO
Crisis Standards of Care
During the pandemic, surges in case volume often led to the implementation of contingency or crisis standards of care, requiring triage of critical care resources, including intensive care unit beds, medical supplies and staffing. At the same time, there was an increased demand for ECMO – a highly resource-intensive intervention with potential for prolonged use of critical care services52 – due to the high incidence of severe, refractory ARDS among patients with COVID-19. Many institutions, in an effort to provide medical care to the greatest number of patients, became more stringent with patient selection for ECMO – or abandoned the use of ECMO altogether.53, 54, 55, 56 For those centers that continued to perform ECMO despite resource constraints, non-traditional staffing models were helpful in maintaining operations.57 In settings of reduced ECMO capacity, it may be necessary to apply more stringent exclusion criteria based on patient characteristics associated with increased mortality and longer ECMO run duration (Figure 3 ).22 Pre-designed triage systems may be useful in standardizing which patients should receive ECMO at varying levels of capacity, and may also help achieve equitable access to ECMO by establishing allocation policies that avoid discrimination based on age, race, ethnicity, disability or socioeconomic status.55 Figure 3 Patient selection and contingency flowsheet for ECMO during a pandemic Contraindications algorithm for V-A and V-V ECMO use (COVID-19 and non-COVID-19) during a pandemic based on system capacity. ∗The impact of duration on high-flow nasal cannula and/or noninvasive mechanical ventilation in addition to invasive mechanical ventilation is unknown. COVID-19, coronavirus disease 2019; CPR, cardiopulmonary resuscitation; ECMO, extracorporeal membrane oxygenation; ICU, intensive care unit; PaCO2, partial pressure of carbon dioxide in arterial blood; PaO2:FiO2, ratio of partial pressure of oxygen in arterial blood to the fractional concentration of oxygen in inspired air; PEEP, positive end-expiratory pressure; V-A, venoarterial; V-V, venovenous.From Badulak J, Antonini MV, Stead CM, et al. Extracorporeal Membrane Oxygenation for COVID-19: Updated 2021 Guidelines from the Extracorporeal Life Support Organization. ASAIO J. 2021;67(5):485-495.
Regional and National Coordination
Over the course of the pandemic, the coordination of ECMO programs at regional and national levels was leveraged to more effectively standardize ECMO candidacy and allocate resources. In Paris, a regional network coordinated care of 17 hospitals to pool resources, systematize ECMO candidacy evaluations, and expanded mobile ECMO capacity in an effort to improve resource utilization, streamline workflow for clinicians, optimize management of patients prior to ECMO initiation, and facilitate data collection.58 , 59 Chile used a National Advisory Commission to help coordinate ECMO referrals, provide consistent patient selection, optimize capacity, and distribute educational materials.60 The United Kingdom modified a preexisting national system to balance ECMO cases among centers in order to help manage capacity.61 The creation or utilization of existing regional or national ECMO networks has been encouraged by the Extracorporeal Life Support Organization (ELSO), which has an “ECMO Availability Map” to help guide such coordination efforts.22 , 62 The purported effectiveness of these examples also provides a rationale for establishing new networks where none currently exist. However, the ability to coordinate ECMO across a region will depend, in large part, on the established healthcare systems in that area.
Development of New ECMO Centers
Initial guidance early in the pandemic recommended against development of new ECMO centers,24 , 63 given the concerns over the implementation of a resource-heavy intervention in inexperienced centers that may be dealing with simultaneous surge capacity issues. Retrospective data from new ECMO centers developed in the Middle East and India – under the guidance of established centers – did report acceptable survival in these new programs overall compared to established ECMO centers (55% vs. 45%, OR 1.65; 95% CI 0.75-3.67), although, as the authors note, patient selection likely differed between the centers with selection bias favoring the new centers. In light of this relative success, and recognizing the potential need for ECMO in regions that otherwise would not have access to ECMO, ELSO has updated its guidance to recommend that establishment of new ECMO centers may be considered in select cases where regional resources exist to support these programs, there is sufficiently high demand, and there is close collaboration with experienced centers to optimize outcomes.22
Evaluating Effectiveness of ECMO in an Ongoing Pandemic
Surveillance and Study Design Approaches
Early data during a pandemic of a novel disease must be interpreted with caution given the potential for misleading data due to study designs and population characteristics. Determining the effectiveness of ECMO during an evolving pandemic requires ongoing surveillance as well as design and implementation of studies that have the ability to assess both short- and longer-term outcomes, and especially patient-centered outcomes.64 National and international registries are useful in centralizing data and may serve as platforms for analysis and dissemination of information.31 As the course of a pandemic takes shape and data begins to accumulate, especially results from clinical trials, the community must learn to pivot toward practices that are more evidence-based.
While traditional RCTs are considered the “gold standard” for providing credible, unbiased evidence for the efficacy of an intervention, they can be difficult to organize and perform in real-time during a rapidly evolving pandemic, especially with a resource-intensive therapy such as ECMO, substantial limitations in staffing and funding to conduct such trials, and a perceived lack of clinical equipoise for randomization. A number of different study designs have been employed during the COVID-19 pandemic in an effort to approximate an RCT using observational data (Table 2 ), including: emulation trials,6 , 17 registry randomized controlled trials, and matched-pair analyses.18 Adaptive platform trials and weighted lottery systems can also provide a more rapid assessment of ECMO efficacy, the latter of which can provide potentially large sample sizes and a more equitable approach to resource allocation.65 Given that temporal changes during the pandemic – including emergence of viral variants and evolution of management strategies – are likely to impact ECMO efficacy and effectiveness, more adaptive and flexible study designs are likely to have the greatest success in providing high-quality evidence in a timely fashion.20 Table 2 Study Designs in a Pandemic, Pros and Cons
Study design Pros Cons
Randomized Controlled Trial (RCT) •Gold standard design
•Minimizes bias and confounding
•Best at determining efficacy •Time consuming and expensive making it difficult to perform in real time during a pandemic
•Would need clinical equipoise
Emulation of Target Trial or RCT •Can simulate RCT with preexisting observational data
•Less time consuming and expensive
•No ethical concerns •May still have residual bias and confounding
Registry RCT •RCT that is less time consuming and expensive
•No ethical concerns •Registries may lack necessary clinical information
Matched-pair analyses •Allows for evaluation of treatment effect in an observational design •May still have residual bias and confounding
Adaptive Platform Trial •Allows evaluation of multiple interventions which can be added or dropped during the study
•Is useful when mechanism of disease is not well understood
•Requires fewer patients •Requires multiple points of interim analysis
Communication and collaboration
Through the course of the pandemic, regional, national and international coordination has been crucial in knowledge sharing, research collaboration and development of guidelines. Remote learning and communication became a key component of health care provider education throughout the pandemic. Educational webinars and conferences by ELSO and other ECMO networks have been utilized to disseminate new data to both experienced and new ECMO centers and practitioners.22
Pediatric Access to ECMO During the Pandemic
Children were generally less susceptible to severe illness associated with COVID-19 infection, though a new post-viral pathological process, Multisystem Inflammatory Syndrome in Children (MISC), was identified during the pandemic.66, 67, 68 Access to ECMO for children with non-COVID-19 critical illness, congenital anomalies and emergent peri-operative indications, in addition to the infrequent pediatric patients who were critically ill with COVID-19, was preserved in many regions and recommended in guidelines.22 , 69, 70, 71, 72, 73 In addition, many established pediatric ECMO centers expanded their admission criteria to facilitate the care of adult COVID patients who required ECMO and offload regional centers at capacity.69 , 70 , 74 While there are many examples of successful shared resource allocation protocols, particular care was required to address the unique challenges of assessing mortality risk in the neonatal and pediatric populations. The COVID-19 pandemic highlighted that at times of critical care shortages, protocols which ensure equity across the life span should be employed.72 , 75
Health Care Providers and ECMO
The COVID-19 pandemic has affected health care workers in a multitude of ways, notably through occupational stress and provider burnout.76 , 77 Surge situations and crisis standards of care have only amplified the pressure placed on health care providers through unfavorable changes in staffing models and increases in provider responsibilities, ethical dilemmas, and patient mortality, among others. Contingency and crisis standards for ECMO implementation during the COVID-19 pandemic have added to provider stress through a potentially heavy ethical burden of rationing care, potentially resulting in an inability, during surge conditions, to provide ECMO to those who meet standard criteria for initiation and would otherwise have received ECMO under less strained conditions. Prolonged ECMO run times and longer than usual time-to-recovery for these patients with ARDS12 can also contribute to healthcare provider burnout. Triage committees have been proposed to relieve bedside providers of these burdens by objectively and independently setting standards and developing guidelines for rationing decisions. However, operationalizing a triage committee amid a crisis is a complex and fraught undertaking that requires direct input from community leaders in order to address ethical and legal issues, along with the need for equity. The major function of an oversight committee under crisis conditions may simply be to offload the overburdened clinicians on the front lines.54
Summary
The COVID-19 pandemic has led to a marked increase in cases of ARDS globally, leading to a concomitant rise in demand for ECMO support. The utilization of ECMO during the pandemic has been complicated by numerous factors, including limited knowledge of ECMO-related outcomes, challenges in performing real-time, high-quality research in an ongoing evolving pandemic, difficulties in patient selection during episodic severe capacity restraints, the ethical dilemmas of rationing care, and excess strain not only on health care systems but on providers as well. As the pandemic eventually recedes and resources become more readily and consistently available to provide ECMO, both coordinated research and tailoring of guidelines will be necessary to understand the role of ECMO, provide the best care possible to patients with severe COVID-19-related ARDS, and to anticipate needs for future potential pandemics.
Clinics Care Points• The COVID-19 pandemic has led to an increase in severe ARDS cases globally, increasing the demand for ECMO.
• Outcomes of patients with COVID-19 managed with ECMO have evolved during the pandemic, with early data suggesting mortality rates similar to those with non-COVID-19 etiologies of ARDS, but later data suggesting increasing mortality and longer ECMO duration of support over time.
• Eligibility criteria, cannulation and management strategies for patients with severe ARDS requiring ECMO thus far remain largely the same for patients with COVID-19.
• Episodic surge capacity requirements have often led to crisis standards of care, which may include the rationing or tailoring of eligibility criteria of ECMO, particularly with respect to more stringent exclusion criteria.
• Regional, national and international collaboration will continue to help inform ECMO providers, disseminate new information and aid in resource allocation.
• Ongoing surveillance data and new research techniques will continue to inform the role of ECMO in the management of COVID-19-related ARDS.
Uncited reference
15., 16..
622 W 168th St, PH8-101, New York, NY 10023, +19177151534
12Department of Intensive Care, Alfred Health, Melbourne, Australia
Disclosure Statement:
DB receives research support from and consults for LivaNova. He has been on the medical advisory boards for Abiomed, Xenios, Medtronic, Inspira and Cellenkos. He is the President-elect of the Extracorporeal Life Support Organization (ELSO) and the Chair of the Executive Committee of the International ECMO Network (ECMONet).
PMAA reported receiving grant funding from the National Institute of Child Health and Human Development, National Institutes of Health (NIH), Pediatric ECMO Anticoagulation Collaborative (PEACE), outside the submitted work and serving on the Board of the Extracorporeal Life Support Organization as Treasurer.
==== Refs
References
1 Combes A. Hajage D. Capellier G. Extracorporeal Membrane Oxygenation for Severe Acute Respiratory Distress Syndrome N Engl J Med 378 21 2018 1965 1975 29791822
2 Goligher E.C. Tomlinson G. Hajage D. Extracorporeal Membrane Oxygenation for Severe Acute Respiratory Distress Syndrome and Posterior Probability of Mortality Benefit in a Post Hoc Bayesian Analysis of a Randomized Clinical Trial Jama 320 21 2018 2251 2259 30347031
3 Munshi L. Walkey A. Goligher E. Pham T. Uleryk E.M. Fan E. Venovenous extracorporeal membrane oxygenation for acute respiratory distress syndrome: a systematic review and meta-analysis Lancet Respir Med 7 2 2019 163 172 30642776
4 Combes A. Peek G.J. Hajage D. ECMO for severe ARDS: systematic review and individual patient data meta-analysis Intensive Care Med 46 11 2020 2048 2057 33021684
5 Schmidt M. Hajage D. Lebreton G. Extracorporeal membrane oxygenation for severe acute respiratory distress syndrome associated with COVID-19: a retrospective cohort study Lancet Respir Med 8 11 2020 1121 1131 32798468
6 Shaefi S. Brenner S.K. Gupta S. Extracorporeal membrane oxygenation in patients with severe respiratory failure from COVID-19 Intensive Care Med 47 2 2021 208 221 33528595
7 Barbaro R.P. MacLaren G. Boonstra P.S. Extracorporeal membrane oxygenation support in COVID-19: an international cohort study of the Extracorporeal Life Support Organization registry Lancet 396 10257 2020 1071 1078 32987008
8 Ramanathan K. Shekar K. Ling R.R. Extracorporeal membrane oxygenation for COVID-19: a systematic review and meta-analysis Crit Care 25 1 2021 211 34127027
9 Broman L.M. Eksborg S. Lo Coco V. De Piero M.E. Belohlavek J. Lorusso R. Extracorporeal membrane oxygenation for COVID-19 during first and second waves Lancet Respir Med 9 8 2021 e80 e81 34146489
10 Riera J. Roncon-Albuquerque R. Jr. Fuset M.P. Alcántara S. Blanco-Schweizer P. Increased mortality in patients with COVID-19 receiving extracorporeal respiratory support during the second wave of the pandemic Intensive Care Med 47 12 2021 1490 1493 34499190
11 Schmidt M. Langouet E. Hajage D. Evolving outcomes of extracorporeal membrane oxygenation support for severe COVID-19 ARDS in Sorbonne hospitals, Paris Crit Care 25 1 2021 355 34627350
12 Barbaro R.P. MacLaren G. Boonstra P.S. Extracorporeal membrane oxygenation for COVID-19: evolving outcomes from the international Extracorporeal Life Support Organization Registry Lancet 398 10307 2021 1230 1238 34599878
13 Karagiannidis C. Slutsky A.S. Bein T. Windisch W. Weber-Carstens S. Brodie D. Complete countrywide mortality in COVID patients receiving ECMO in Germany throughout the first three waves of the pandemic Crit Care 25 1 2021 413 34844657
14 Ling R, Ramanathan K, J S, et al. Evolving outcomes of extracorporeal membrane oxygenation during the first 2 years of the COVID-19 pandemic: a systematic review and meta-analysis. Critical care. 2022;26(147).
15 Abrams D. Grasselli G. Schmidt M. Mueller T. Brodie D. ECLS-associated infections in adults: what we know and what we don't yet know Intensive Care Med 46 2 2020 182 191 31768569
16 Carbonell R. Urgelés S. Rodríguez A. Mortality comparison between the first and second/third waves among 3,795 critical COVID-19 patients with pneumonia admitted to the ICU: A multicentre retrospective cohort study Lancet Reg Health Eur 11 2021 100243
17 Urner M. Barnett A.G. Bassi G.L. Venovenous extracorporeal membrane oxygenation in patients with acute covid-19 associated respiratory failure: comparative effectiveness study Bmj 377 2022 e068723
18 Whebell S. Zhang J. Lewis R. Survival benefit of extracorporeal membrane oxygenation in severe COVID-19: a multi-centre-matched cohort study Intensive Care Med 48 4 2022 467 478 35238946
19 Gannon W.D. Stokes J.W. Francois S.A. Association Between Availability of ECMO and Mortality in COVID-19 Patients Eligible for ECMO: A Natural Experiment Am J Respir Crit Care Med 2022
20 Granholm A. Alhazzani W. Derde L.P.G. Randomised clinical trials in critical care: past, present and future Intensive Care Med 48 2 2022 164 178 34853905
21 Abrams D. Ferguson N.D. Brochard L. ECMO for ARDS: from salvage to standard of care? Lancet Respir Med 7 2 2019 108 110 30642778
22 Badulak J. Antonini M.V. Stead C.M. Extracorporeal Membrane Oxygenation for COVID-19: Updated 2021 Guidelines from the Extracorporeal Life Support Organization Asaio j 67 5 2021 485 495 33657573
23 Brodie D. Slutsky A.S. Combes A. Extracorporeal Life Support for Adults With Respiratory Failure and Related Indications: A Review Jama 322 6 2019 557 568 31408142
24 Shekar K. Badulak J. Peek G. Extracorporeal Life Support Organization Coronavirus Disease 2019 Interim Guidelines: A Consensus Document from an International Group of Interdisciplinary Extracorporeal Membrane Oxygenation Providers Asaio j 66 7 2020 707 721 32604322
25 Tonna J.E. Abrams D. Brodie D. Management of Adult Patients Supported with Venovenous Extracorporeal Membrane Oxygenation (VV ECMO): Guideline from the Extracorporeal Life Support Organization (ELSO) Asaio j 67 6 2021 601 610 33965970
26 Brodie D. Abrams D. MacLaren G. ECMO During Respiratory Pandemics: Past, Present, and Future Am J Respir Crit Care Med 2022
27 Mustafa A.K. Alexander P.J. Joshi D.J. Extracorporeal Membrane Oxygenation for Patients With COVID-19 in Severe Respiratory Failure JAMA Surg 155 10 2020 990 992 32780089
28 Creel-Bulos C. Hockstein M. Amin N. Melhem S. Truong A. Sharifpour M. Acute Cor Pulmonale in Critically Ill Patients with Covid-19 N Engl J Med 382 21 2020 e70 32374956
29 Salas de Armas I.A. Akkanti B.H. Janowiak L. Inter-hospital COVID ECMO air transportation Perfusion 36 4 2021 358 364 33233987
30 Rafiq M.U. Valchanov K. Vuylsteke A. Regional extracorporeal membrane oxygenation retrieval service during the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pandemic: an interdisciplinary team approach to maintain service provision despite increased demand Eur J Cardiothorac Surg 58 5 2020 875 880 32964930
31 Extracorporeal Life Support Organization (ELSO). Resources > Guidelines. Available at: https://www.elso.org/Resources/Guidelines.aspx. Accessed April 1, 2022.
32 Mang S. Reyher C. Mutlak H. Awake Extracorporeal Membrane Oxygenation for COVID-19-induced Acute Respiratory Distress Syndrome Am J Respir Crit Care Med 205 7 2022 847 851 35044896
33 Garcia B. Cousin N. Bourel C. Jourdain M. Poissy J. Duburcq T. Prone positioning under VV-ECMO in SARS-CoV-2-induced acute respiratory distress syndrome Crit Care 24 1 2020 428 32665007
34 Giani M. Martucci G. Madotto F. Prone Positioning during Venovenous Extracorporeal Membrane Oxygenation in Acute Respiratory Distress Syndrome. A Multicenter Cohort Study and Propensity-matched Analysis Annals of the American Thoracic Society 18 3 2021 495 501 32941739
35 Rosano A. Martinelli E. Fusina F. Early Percutaneous Tracheostomy in Coronavirus Disease 2019: Association With Hospital Mortality and Factors Associated With Removal of Tracheostomy Tube at ICU Discharge. A Cohort Study on 121 Patients Crit Care Med 49 2 2021 261 270 33201005
36 Dres M. Burrel S. Boutolleau D. SARS-CoV-2 Does Not Spread Through Extracorporeal Membrane Oxygenation or Dialysis Membranes Am J Respir Crit Care Med 202 3 2020 458 460 32525400
37 Supady A. Weber E. Rieder M. Cytokine adsorption in patients with severe COVID-19 pneumonia requiring extracorporeal membrane oxygenation (CYCOV): a single centre, open-label, randomised, controlled trial Lancet Respir Med 9 7 2021 755 762 34000236
38 Barnes G.D. Burnett A. Allen A. Thromboembolism and anticoagulant therapy during the COVID-19 pandemic: interim clinical guidance from the anticoagulation forum J Thromb Thrombolysis 50 1 2020 72 81 32440883
39 Yusuff H. Zochios V. Brodie D. Thrombosis and Coagulopathy in COVID-19 Patients Requiring Extracorporeal Membrane Oxygenation Asaio j 66 8 2020 844 846 32740341
40 Bemtgen X. Zotzmann V. Benk C. Thrombotic circuit complications during venovenous extracorporeal membrane oxygenation in COVID-19 J Thromb Thrombolysis 51 2 2021 301 307 32653986
41 Guo Z. Sun L. Li B. Anticoagulation Management in Severe Coronavirus Disease 2019 Patients on Extracorporeal Membrane Oxygenation J Cardiothorac Vasc Anesth 35 2 2021 389 397 32994131
42 Masur J. Freeman C.W. Mohan S. A Double-Edged Sword: Neurologic Complications and Mortality in Extracorporeal Membrane Oxygenation Therapy for COVID-19-Related Severe Acute Respiratory Distress Syndrome at a Tertiary Care Center AJNR Am J Neuroradiol 41 11 2020 2009 2011 32855187
43 Usman A.A. Han J. Acker A. A Case Series of Devastating Intracranial Hemorrhage During Venovenous Extracorporeal Membrane Oxygenation for COVID-19 J Cardiothorac Vasc Anesth 34 11 2020 3006 3012 32828653
44 Zahid M.J. Baig A. Galvez-Jimenez N. Martinez N. Hemorrhagic stroke in setting of severe COVID-19 infection requiring Extracorporeal Membrane Oxygenation (ECMO) J Stroke Cerebrovasc Dis 29 9 2020 105016
45 Heman-Ackah S.M. Su Y.S. Spadola M. Neurologically Devastating Intraparenchymal Hemorrhage in COVID-19 Patients on Extracorporeal Membrane Oxygenation: A Case Series Neurosurgery 87 2 2020 E147 e151 32424429
46 McMichael ABV, Ryerson LM, Ratano D, Fan E, Faraoni D, Annich GM. 2021 ELSO Adult and Pediatric Anticoagulation Guidelines. Asaio j. 2022;68(3):303-310.
47 Ramanathan K. MacLaren G. Combes A. Brodie D. Shekar K. Blood transfusion strategies and ECMO during the COVID-19 pandemic - Authors' reply Lancet Respir Med 8 5 2020 e41 32305077
48 Bharat A. Machuca T.N. Querrey M. Early outcomes after lung transplantation for severe COVID-19: a series of the first consecutive cases from four countries Lancet Respir Med 9 5 2021 487 497 33811829
49 Roach A. Chikwe J. Catarino P. Lung Transplantation for Covid-19-Related Respiratory Failure in the United States N Engl J Med 386 12 2022 1187 1188 35081299
50 Bermudez C.A. Crespo M.M. The Case for Prolonged ECMO for COVID-19 ARDS as a Bridge to Recovery or Lung Transplantation Transplantation 106 4 2022 e198 e199 35135969
51 Cypel M. Keshavjee S. When to consider lung transplantation for COVID-19 Lancet Respir Med 8 10 2020 944 946 32857989
52 Combes A. Brodie D. Bartlett R. Position paper for the organization of extracorporeal membrane oxygenation programs for acute respiratory failure in adult patients Am J Respir Crit Care Med 190 5 2014 488 496 25062496
53 Supady A. Badulak J. Evans L. Curtis J.R. Brodie D. Should we ration extracorporeal membrane oxygenation during the COVID-19 pandemic? Lancet Respir Med 9 4 2021 326 328 33798457
54 Supady A. Brodie D. Curtis J.R. Ten things to consider when implementing rationing guidelines during a pandemic Intensive Care Med 47 5 2021 605 608 33677625
55 Supady A. Curtis J.R. Abrams D. Allocating scarce intensive care resources during the COVID-19 pandemic: practical challenges to theoretical frameworks Lancet Respir Med 9 4 2021 430 434 33450202
56 Abrams D. Lorusso R. Vincent J.L. Brodie D. ECMO during the COVID-19 pandemic: when is it unjustified? Crit Care 24 1 2020 507 32807189
57 Agerstrand C. Dubois R. Takeda K. Extracorporeal Membrane Oxygenation for Coronavirus Disease 2019: Crisis Standards of Care Asaio j 67 3 2021 245 249 33627596
58 Lebreton G. Schmidt M. Ponnaiah M. Extracorporeal membrane oxygenation network organisation and clinical outcomes during the COVID-19 pandemic in Greater Paris, France: a multicentre cohort study Lancet Respir Med 9 8 2021 851 862 33887246
59 Levy D. Lebreton G. Pineton de Chambrun M. Outcomes of Patients Denied Extracorporeal Membrane Oxygenation during the COVID-19 Pandemic in Greater Paris, France Am J Respir Crit Care Med 204 8 2021 994 997 34375170
60 Diaz R.A. Graf J. Zambrano J.M. Extracorporeal Membrane Oxygenation for COVID-19-associated Severe Acute Respiratory Distress Syndrome in Chile: A Nationwide Incidence and Cohort Study Am J Respir Crit Care Med 204 1 2021 34 43 33823118
61 Camporota L. Meadows C. Ledot S. Consensus on the referral and admission of patients with severe respiratory failure to the NHS ECMO service Lancet Respir Med 9 2 2021 e16 e17 33428874
62 Extracorporeal Life Support Organization (ELSO). Membership > ECMO Availability Map. Available at: https://www.elso.org/Membership/ECMOAvailabilityMap.aspx. Accessed April 1, 2022.
63 Bartlett R.H. Ogino M.T. Brodie D. Initial ELSO Guidance Document: ECMO for COVID-19 Patients with Severe Cardiopulmonary Failure Asaio j 66 5 2020 472 474 32243267
64 MacLaren G. Fisher D. Brodie D. Treating the Most Critically Ill Patients With COVID-19: The Evolving Role of Extracorporeal Membrane Oxygenation Jama 327 1 2022 31 32 34919122
65 White D.B. Angus D.C. A Proposed Lottery System to Allocate Scarce COVID-19 Medications: Promoting Fairness and Generating Knowledge Jama 324 4 2020 329 330 32579158
66 Lu X. Zhang L. Du H. SARS-CoV-2 Infection in Children N Engl J Med 382 17 2020 1663 1665 32187458
67 Feldstein LR, Rose EB, Horwitz SM, et al. Multisystem Inflammatory Syndrome in U.S. Children and Adolescents. N Engl J Med. 2020;383(4):334-346.
68 Feldstein L.R. Tenforde M.W. Friedman K.G. Characteristics and Outcomes of US Children and Adolescents With Multisystem Inflammatory Syndrome in Children (MIS-C) Compared With Severe Acute COVID-19 Jama 325 11 2021 1074 1087 33625505
69 Gerall C. Cheung E.W. Klein-Cloud R. Kreines E. Brewer M. Middlesworth W. Allocation of resources and development of guidelines for extracorporeal membrane oxygenation (ECMO): Experience from a pediatric center in the epicenter of the COVID-19 pandemic Journal of pediatric surgery 55 12 2020 2548 2554 32951890
70 DeFazio J.R. Kahan A. Fallon E.M. Development of pediatric surgical decision-making guidelines for COVID-19 in a New York City children's hospital Journal of pediatric surgery 55 8 2020 1427 1430 32553456
71 Cho H.J. Ogino M.T. Jeong I.S. Pediatric intensive care preparedness and ECMO availability in children with COVID-19: An international survey Perfusion 36 6 2021 637 639 33342366
72 Lemmon M.E. Truog R.D. Ubel P.A. Allocating Resources Across the Life Span During COVID-19-Integrating Neonates and Children Into Crisis Standards of Care Protocols JAMA Pediatr 175 4 2021 347 348 33315093
73 Cleary A. Chivers S. Daubeney P.E. Simpson J.M. Impact of COVID-19 on patients with congenital heart disease Cardiol Young 31 1 2021 163 165 33308340
74 Yager P.H. Whalen K.A. Cummings B.M. Repurposing a Pediatric ICU for Adults N Engl J Med 382 22 2020 e80
75 Emanuel E.J. Persad G. Upshur R. Fair Allocation of Scarce Medical Resources in the Time of Covid-19 N Engl J Med 382 21 2020 2049 2055 32202722
76 Myran DT, Cantor N, Rhodes E, et al. Physician Health Care Visits for Mental Health and Substance Use During the COVID-19 Pandemic in Ontario, Canada. JAMA Network Open. 2022;5(1):e2143160-e2143160.
77 Pfefferbaum B. North C.S. Mental Health and the Covid-19 Pandemic N Engl J Med 383 6 2020 510 512 32283003
| 0 | PMC9705197 | NO-CC CODE | 2022-12-03 23:15:43 | no | Clin Chest Med. 2022 Nov 29; doi: 10.1016/j.ccm.2022.11.016 | utf-8 | Clin Chest Med | 2,022 | 10.1016/j.ccm.2022.11.016 | oa_other |
==== Front
J Environ Manage
J Environ Manage
Journal of Environmental Management
0301-4797
1095-8630
Elsevier Ltd.
S0301-4797(22)02491-4
10.1016/j.jenvman.2022.116918
116918
Research Article
Do wildfires exacerbate COVID-19 infections and deaths in vulnerable communities? Evidence from California
Yu Suyang ∗
Hsueh Lily
School of Public Affairs, Arizona State University, 411 N Central Ave Suite 400, Phoenix, AZ, 85004, USA
∗ Corresponding author. School of Public Affairs, Arizona State University. 411 N Central Ave Suite 400, Phoenix, AZ, 85004, USA.
29 11 2022
29 11 2022
11691817 6 2022
25 11 2022
27 11 2022
© 2022 Elsevier Ltd. All rights reserved.
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Understanding whether and how wildfires exacerbate COVID-19 outcomes is important for assessing the efficacy and design of public sector responses in an age of more frequent and simultaneous natural disasters and extreme events. Drawing on the environmental and emergency management literatures, we investigate how wildfire smoke (PM2.5) impacted COVID-19 infections and deaths during California’s 2020 wildfire season and how public housing resources and hospital capacity moderated wildfires' effects on COVID-19 outcomes. We also hypothesize and empirically assess the differential impact of wildfire smoke on COVID-19 infections and deaths in counties exhibiting high and local social vulnerability. To test our hypotheses concerning wildfire severity and its disproportionate impact on COVID-19 outcomes in socially vulnerable communities, we construct a county-by-day panel dataset for the period April 1 to November 30, 2020, in California, drawing on publicly available state and federal data sources. This study's empirical results, based on panel fixed effects models, show that wildfire smoke is significantly associated with increases in COVID-19 infections and deaths. Moreover, wildfires exacerbated COVID-19 outcomes by depleting the already scarce hospital and public housing resources in local communities. Conversely, when wildfire smoke doubled, a one percent increase in the availability of hospital and public housing resources was associated with a 2 to 7 percent decline in COVID-19 infections and deaths. For California communities exhibiting high social vulnerability, the occurrence of wildfires worsened COVID-19 outcomes. Sensitivity analyses based on an alternative sample size and different measures of social vulnerability validate this study's main findings. An implication of this study for policymakers is that communities exhibiting high social vulnerability will greatly benefit from local government policies that promote social equity in housing and healthcare before, during, and after disasters.
Keywords
Wildfires
COVID-19 infections and deaths
Public housing and shelters
Healthcare resources
Social vulnerability
Environmental justice
==== Body
pmc1 Introduction
The United States is facing an increasingly complex, challenging set of scenarios due to the confluence of the two most pressing global health threats — the rapid emergence of the COVID-19 pandemic and the continuously evolving climate crisis (Rodrigues et al., 2020; Salas et al., 2020b; Shultz et al., 2020). Wildfires, storms, flooding, and droughts are among the most immediately apparent sources of displacement and disruption in the context of the pandemic (Phillips et al., 2020; Travaglio et al., 2021).
Many scholars have noted that climate hazards, which are increasing in frequency and intensity in the context of climate change, are likely to intersect with the COVID-19 outbreak and worsen its infections and deaths by impeding public health responses (Conticini et al., 2020; Deek, 2020; Delfino et al., 2009; Henderson, 2020). For example, wildfires occurring globally, such as in Washington state in the U.S., Australia, and England in 2020, have been demonstrated to directly affect the human respiratory system due to the toxicity of wildfire smoke and indirectly squeeze out medical and sheltering resources, thus increasing COVID-19 infections and deaths (Burke et al., 2021; Cortes-Ramirez et al., 2022; Henderson, 2020; Travaglio et al., 2021).
However, for local communities, the need to address wildfires may also impact their already tight emergency budgets, staffing, and health service resources, which could jeopardize COVID-19 infection control (Salas et al., 2020b). In particular, these compound risks exacerbate the unfolding economic crisis and longstanding socioeconomic and racial disparities in ways that can increase the risk to specific populations and impede recovery (Phillips et al., 2020; Watkins and Gerrish, 2018; Wright and Merritt, 2020).
This paper is focused on the state of California because the wildfire season in 2020 was characterized by a record-setting year of burned acres, fire severity, and socioeconomic costs. Occurring at the same time, the COVID-19 pandemic in California was also severe, in part due to its population size and density as well as disparities in local response capacities (Keeley and Syphard, 2021). In this paper, we explore the direct and indirect effects of wildfires on COVID-19 infections and deaths and ask whether their compound effects are likely to disproportionately affect counties with different levels of social vulnerability. Specifically, we answer three questions: First, what are the direct effects of wildfire smoke (represented by PM2.5) on COVID-19 infections and deaths? Second, to what extent have local government policies and responses in the form of hospital and public housing resources, respectively, helped to reduce the impacts of wildfire smoke on COVID-19 outcomes? Third, how have communities with different levels of socioeconomic vulnerabilities been impacted during the 2020 wildfire season?
2 Prior literature and hypotheses
Previous studies have demonstrated that COVID-19 is a disease affecting the respiratory system and that it is impacted by atmospheric factors such as atmospheric pollution and notably fine and coarse particulate matters such as PM2.5 and PM10 (Contini and Costabile, 2020; Travaglio et al., 2021; Wu et al., 2020). Aguilera et al. (2021) report that the toxicity of PM2.5 varies across different sources of emissions, and PM2.5 from wildfire smoke is more toxic than an equal dose of ambient PM2.5. Thus, residents in communities with a history of exposure to frequent wildfires are more likely to possess preconditions that weaken their respiratory system and/or immune system and thus increase their chances of becoming infected by COVID-19 (Prunicki et al., 2019; Isphording and Pestel, 2021; Magazzino et al., 2020; Naqvi et al., 2022). Furthermore, studies have demonstrated that even short-term exposure to wildfire smoke can exacerbate the probability and severity of COVID-19 infections and outcomes (Deek, 2020; Travaglio et al., 2021).
In our paper, we argue that this relationship is intensified and exacerbated when the local community resources that are assigned to deal with both wildfires and COVID-19 are in high demand, but the provisions of such resources lag behind. There are several reasons for this. First, regarding hospital resources, large-scale wildfires (even without a raging pandemic) could force local hospitals to be closed or evacuated due to electricity failure, water shortages in buildings, damaged infrastructure or disrupted transportation systems for the transfer of patients (Hamideh et al., 2022; Melnychuk et al., 2022; Schranz et al., 2010). Wildfires can also rapidly increase demands on health resources such as ventilators, medical services, and extra care for patients with severe COVID-19 symptoms (Ranney et al., 2020; Zangrillo et al., 2020). On the other hand, frontline medical workers are at a higher risk of becoming infected by COVID-19 patients whether due to a lack of personal protective equipment, burnout or extra workload, all of which impose additional challenges in the response to both wildfires and the COVID-19 pandemic (Gupta and Sahoo, 2020).
Second, with respect to public shelters and housing resources, wildfires could leave many people “homeless” (Goodling, 2019; Settembrino, 2017). In previous wildfire seasons, devastating wildfires in California forced more than one million people to evacuate between 2017 and 2019 (Wong et al., 2020). Public shelters and housing resources, whether intended for emergency use or not, are critical after natural disasters (Davlasheridze and Miao, 2021). In the case of wildfires, evacuation and/or sheltering in safe places can help residents decrease their chances of breathing in wildfire smoke (Cova et al., 2009; Whittaker et al., 2017; Wong et al., 2020a, 2021).
Unfortunately, a lack of access to stable shelter, infrastructure, and services means that the homeless community is exposed to a range of environmental hazards. Furthermore, the social vulnerability of local communities, as indicated by a high percentage of elderly people, young children, minorities, low-income households, and people living in nursing homes, among other indicators of vulnerability, can worsen COVID-19 outcomes in these communities (Cabral and Cuevas, 2020; Maxwell et al., 2020). Due to their inability to access adequate medical care, transportation, and nutrition, socially vulnerable populations face increased risks of health challenges during disasters (Karaye and Horney, 2020).
Based on these prior findings in the literature, we propose four hypotheses to answer the aforementioned questions regarding the concurrence of wildfires and the COVID-19 pandemic in California: (1) Wildfire smoke (PM2.5) is positively associated with increases in COVID-19 infections and deaths. As wildfire smoke increases, increases in the availability of (2) hospital capacity and (3) public housing and related shelters, respectively, are associated with a decrease in COVID-19 infections and deaths. (4) Higher social vulnerability status exacerbates COVID-19 infections and deaths compared to lower social vulnerability status, whereas hospital and public housing resources moderate the effects of wildfires (wildfire smoke) on COVID-19 severity.
To test our hypotheses, we analyze data collected from multiple sources and employ a county-by-day panel fixed effects model. The model results show that wildfire smoke (PM2.5) has a significant and positive effect on both COVID-19 infections and deaths. Moreover, by further limiting the number of ICU beds, wildfires in California have presented a challenge to local community health resources and worsened the COVID-19 infections and deaths of those communities. Although local communities have temporarily provided public housing resources – such as contracted hotel rooms to shelter the homeless – and have contained the spread of COVID-19, severe wildfires have also led to property losses. Once these residents have become homeless, they needed to utilize public housing resources. In this way, wildfires have exerted increased pressure on local public housing capacity and consequently exacerbated COVID-19 outcomes. Furthermore, the strain on the local health and housing infrastructures have been especially acute in counties experiencing high levels of social vulnerability.
3 Research methods
This study employs panel fixed effect models to examine the proposed hypotheses. Panel fixed effect models have the advantage of accounting for sources of unobserved heterogeneity, such as unobserved variation that could occur on a specific day in a given county, which our paper's empirical analysis exploits. As such, our unit of analysis is at the county-by-day level.
3.1 Data and variables
Data for the empirical analysis come from the State of California's COVID19.CA.GOV website, the California Department of Forestry and Fire Protection (CAL Fire), and the United States Environmental Protection Agency (EPA). Our research period is from April 1, 2020 to November 30, 2020. During this period, as the COVID-19 pandemic ensued 173 fire incidents were reported across 53 counties in California (with the exceptions of San Francisco, Inyo, Sutter, Del Norte and Alpine counties where no fires were reported), and the affected areas ranged from 69 acres to 1,395,881 acres.
Our sample includes a measure of wildfire smoke (represented by PM2.5), available public shelters and housing units, available staffed ICU beds, and COVID-19 infections and deaths during the research period. These data are available on a daily basis for a majority of counties in the state of California. Of note, not all counties have provided public housing and shelter resources and the air quality monitoring system does not cover all counties nor report data on a daily basis. Excluding county-by-day cases with these missing data means that our final dataset is an unbalanced panel comprises of 5447 county-by-day observations for COVID-19 infections and 5357 county-by-day observations for COVID-19 deaths. Despite this, this study's sample is by and large representative of the state of California as a whole (see Table A in the Appendix).1
This study's outcome or dependent variables (DVs) of interest are COVID-19 infections and deaths. There are four key independent variables: (1) Consistent with other studies (Alman et al., 2016; Henderson, 2020; Reid et al., 2019), we use PM2.5 to measure wildfire smoke and its impacts on human health. Following Miskell et al.’s (2019) approach, we have excluded counties with missing PM2.5 data from our empirical analysis (as noted above). (2) Hospital resource availability is measured by the number of available staffed ICU beds per day at the county level (available ICU beds). (3) The variable for available public shelters and housing units is measured by the number of shelters and temporary emergency housing units provided by local communities (e.g., hotel rooms or RVs). (4) Social vulnerability is measured by the 2018 CDC's social vulnerability index (SVI), and the lower and upper bound score on the SVI is 0 and 1, respectively.
Three variables are included to account for daily time-variant characteristics and county-level resources or policies:2 (1) we employ a fire indicator (binary) to represent whether a county experienced wildfire on a given day (1 for yes, 0 for no fire incidents). (2) Testing is measured by the total number of people who receive COVID-19 tests in a county on a given day. (3) Hospitalization is measured by the number of patients currently hospitalized in an inpatient bed who have suspected or confirmed COVID-19 infections for each county on a given day. We include the latter two variables to preclude alternative explanations that link wildfire smoke to COVID-19 outcomes; they are expected to positively correlate with pandemic related infections and deaths. Recent research shows that testing capacity can mask the COVID-19 infection rate (Omori et al., 2020; Sharfstein et al., 2020). Particularly for counties with limited testing capacity, reported cases may not reflect true epidemic growth.
Given the large data variability, a natural log-transformation was applied to COVID-19 infections, deaths, PM2.5, hospital resource availability, public shelter availability, testing, hospitalization, and their interactions, which are denoted as lninfections, lndeaths, lnpm25, lnICU_beds, lnshelters, lntest, and lnhosp, and lnpm25Xlnshelters and lnpm25XlnICU_beds, respectively.
Table 1 presents summary statistics of the main variables of interest and the control variables for our sample frame. In this table, we also include descriptive statistics for the interaction terms of interest (i.e., lnpm25Xlnshelters and lnpm25XlnICU_beds) in our empirical analysis. Table B in the supplementary materials presents the correlations among the dependent variables, independent variables, and control variables as well as their interaction terms. The results show that lnhosp, lnshelters, and lnICU_beds are highly correlated. To thwart potential multicollinearity concerns (i.e., inflated standard errors), these variables appear in separate models (see model result tables) in our empirical analysis.Table 1 Descriptive statistics (county-by-day dataset).
Table 1Variable Obs Mean Std. Dev. Min Max
infections 14,152 9284.964 30289.320 0 431241
deaths 14,152 183.498 700.726 0 7977
shelters 7108 368.337 763.831 11 4697
pm2.5 10,895 14.516 25.999 −1.3 692.300
hospitalization 14,152 88.343 271.294 0 2907
ICU_beds 12,885 56.770 149.714 −110 1502
fire_or_not 14,152 0.199 0.399 0 1
test 14,152 1769.950 5736.972 0 123257
lnpm25 10,858 2.148 0.927 −2.996 6.540
lninfections 14,002 6.634 2.761 0 12.974
lndeaths 11,183 3.373 2.137 0 8.984
lnhosp 10,792 3.160 1.896 0 7.975
lnICU_beds 11,475 2.828 1.600 0 7.315
lntest 14,152 9.855 2.360 1.609 15.965
lnshelters 7108 4.926 1.287 2.398 8.455
lnpm25Xlnshelters 6052 11.100 5.149 −8.054 39.282
lnpm25XlnICU_beds 9482 6.756 4.499 −12.215 30.058
Overall SVI 14,152 0.624 0.264 0.045 0.998
SVI_theme1 14,152 0.532 0.275 0.033 0.985
SVI_theme2 14,152 0.399 0.301 0.005 0.976
SVI_theme3 14,152 0.821 0.206 0.193 0.998
SVI_theme4 14,152 0.673 0.242 0.063 0.987
ratio_disadvantaged 14,152 0.357 0.254 0 1
3.2 Panel fixed effect modelling
Following a similar approach in other studies (e.g., Schwarz et al., 2022; Zhou et al., 2021), we model the COVID-19 outcomes (infections or deaths) of a county on a given day as follows:COVID−19outcomesi,t=γ0+γ1PM2.5i,t+γ2PM2.5×ICU_bedsi,t+γ3PM2.5×sheltersi,t+γXi,t+αi+fire_or_noti,t+μi,t
where iandt refer to the county and day, respectively. PM2.5i,t represents the wildfire smoke in a given county on a given day and Xi,t represents the county-level controls. Furthermore, γ0 is the county fixed effect and αi is the time (day) fixed effect. fireornoti,t is a binary variable indicating whether wildfire occurred in a county on a given day.
To analyze the indirect effects of wildfire smoke (PM2.5) on COVID-19 infections and deaths through the moderating effects of local hospital and public shelter resources, we introduce, as mentioned above, the cross-terms of PM2.5 with hospital capacity availability and public housing resources, respectively. In this context, PM2.5×ICUbedsi,t represents the cross-term of PM2.5 and hospital capacity of county iondayt, and PM2.5×sheltersi,t represents the cross-term of PM2.5 and public housing resources of county iondayt.
To examine the effect of social vulnerability on the relationship between wildfire smoke and COVID-19 infections and deaths, respectively, we create a dummy variable (svi_high) using the mean + 1 SD (i.e., standard deviation) of the overall SVI as a cutoff point for all California counties, following the approach of Biggs et al. (2021) and Freese et al. (2021). svi_high is recorded as 1 when a county's overall SVI score is higher than 0.8879; otherwise, this value is recorded as 0. There are 14 counties exhibiting a “high” level of social vulnerability. We use this dummy variable to conduct analysis on split samples of counties in California exhibiting high versus low social vulnerability (svi_high = 1, or 0), respectively.
4 Model results
Table 2, Table 3 report the panel fixed effect model results for COVID-19 infections and deaths, respectively. Presented in both tables are the results of three models corresponding to this study's first three hypotheses. Model 1 is a baseline model that includes the key independent variables (PM2.5, ICU beds, and public shelters) and three control variables, namely, fire_or_not, testing, and hospitalization. Model 2 is the baseline model as well as the interaction term of wildfire smoke and available ICU beds in a county on a given day (lnpm25 × lnICU_bed). Model 3 is Model 1 along with an interaction term between wildfire smoke and available public shelters and housing units (lnpm25 × lnshelter).Table 2 Daily panel fixed effect model results.
Table 2DV1 = lninfections Model 1 (baseline) Model 2 (ICU_beds) Model 3 (shelters)
lnpm25 0.038* 0.106** 0.243***
(0.022) (0.049) (0.089)
fire_or_not −0.000 0.002 0.007
(0.036) (0.034) (0.035)
lnICU_beds −0.005 0.041 −0.009
(0.031) (0.041) (0.030)
lnshelters −0.074 −0.067 0.017
(0.054) (0.052) (0.058)
lnhosp 0.248*** 0.245*** 0.241***
(0.046) (0.046) (0.044)
lntest 0.757*** 0.745*** 0.738***
(0.206) (0.205) (0.206)
lnpm25XlnICU_beds −0.024*
(0.012)
lnpm25Xlnshelters −0.041**
(0.016)
constant −1.011 −1.059 −1.257
(1.798) (1.786) (1.804)
day fixed effects YES YES YES
county fixed effects YES YES YES
r2 0.971 0.971 0.971
N 5447 5447 5447
*p < 0.1, **p < 0.05, ***p < 0.01. Robust SEs are included in parentheses.
Table 3 Daily panel fixed effect model results.
Table 3DV2 = lndeaths Model 1 (baseline) Model 2 (ICU_beds) Model 3 (shelters)
lnpm25 0.026 0.158 0.525***
(0.041) (0.097) (0.120)
fire_or_not −0.001 0.003 0.016
(0.069) (0.067) (0.064)
lnICU_beds −0.038 0.050 −0.047
(0.091) (0.120) (0.088)
lnshelters −0.098 −0.087 0.117
(0.093) (0.091) (0.096)
lnhosp 0.119* 0.112* 0.100
(0.064) (0.063) (0.063)
lntest 1.187*** 1.165*** 1.139***
(0.369) (0.364) (0.360)
lnpm25XlnICU_beds −0.045*
(0.026)
lnpm25Xlnshelters −0.099***
(0.024)
constant −7.682** −7.774** −8.248**
(3.345) (3.310) (3.196)
day fixed effects YES YES YES
county fixed effects YES YES YES
r2 0.860 0.862 0.866
N 5357.00 5357.00 5357.00
*p < 0.1, **p < 0.05, ***p < 0.01. Robust SEs are included in parentheses.
Table 2 Model 1, which does not contain the interaction terms, shows that wildfire smoke has a significantly positive effect on COVID-19 infections. By contrast, we do not find a significant relationship between COVID-19 deaths and PM2.5 when the interaction terms are excluded (see Table 3 Model 1). That being said, when the interaction term for wildfire smoke and public shelter resources enters into the model, as shown in Model 3 of Table 3, wildfire smoke exerts a statistically significant positive and direct effect on COVID-19 deaths.
Turning to the results of the interaction terms, Model 2 results in Table 2, Table 3, respectively, show that the interaction of wildfire smoke and the availability of ICU bed has a significant and negative effect on both COVID-19 infections and deaths, at the p < 0.10 level. These results indicate that in areas with a higher density of wildfire smoke, an increase in hospital capacity availability is associated with a decrease in COVID-19 infections and deaths compared to areas that has a lower density of wildfire smoke. Based on our analysis, when PM2.5 doubles, a one percent increase in hospital capacity reduces COVID-19 infections and deaths by 1.7 and 3.1 percent, respectively (see marginal impact calculations in the supplementary materials).
The results of Model 3 show that in the face of increased wildfire smoke, an increase in public shelters and housing resources is associated with a decrease in COVID-19 infections and deaths, at the levels of p < 0.05 and p < 0.01, respectively. A one percent increase in the availability of public shelters and housing resources is associated with a 2.9 and 6.8 percent reduction in COVID-19 infections and deaths, respectively, when PM2.5 increases by two-folds (see marginal impact calculations in the supplementary materials).
Table 4, Table 5 compare the estimated coefficients of Models 1, 2, and 3 across the high and low vulnerability groups of counties. Model 1 results in Table 4 show that wildfire smoke (PM2.5) has statistically significant direct effects on COVID-19 infections for counties exhibiting low social vulnerability. After including its interaction effect with hospital resources in Model 2 of Table 4, the direct effect of wildfire smoke and its interaction effect with healthcare resources are significant for counties with both high and low overall SVI, respectively; there is no statistical difference between the two groups. In Model 3 of Table 4, after including its interaction effect with public housing and shelters, wildfire smoke has a significant direct effect and significant interaction effect on COVID-19 infections only for counties with a low overall SVI, while the effect for counties with a high SVI is nonsignificant.Table 4 Panel fixed effect model results for counties with low and high levels of social vulnerability.
Table 4(DV1 = lninfections) Model 1 (baseline) Model 2 (icu_bed) Model 3 (shelter)
Low SVI High SVI bo-b1 p value Low SVI High SVI bo-b1 p value Low SVI High SVI bo-b1 p value
lnpm25 0.052*** 0.013 0.039 0.150 0.140*** 0.047* 0.094 0.130 0.354*** 0.027 0.327 0.060
(7.010) (0.79) (10.99) (1.91) (15.37) (0.73)
fire_or_not 0.006 0.014 −0.009 0.450 0.004 0.019 −0.015 0.390 0.0104 0.015 −0.004 0.490
(0.500) (0.80) (0.35) (1.05) (0.95) (0.81)
lnICU_beds −0.040*** 0.074*** −0.115 0.020 0.017* 0.102*** −0.085 0.170 −0.036*** 0.073*** −0.109 0.030
(-5.540) (8.47) (1.72) (5.81) (-5.05) (8.23)
lnshelters −0.106*** −0.034** −0.071 0.320 −0.096*** −0.030** −0.066 0.340 0.015 −0.026 0.041 0.450
(-7.740) (-2.58) (-7.03) (-2.20) (0.92) (-1.17)
lnhosp 0.258*** 0.206*** 0.052 0.320 0.252*** 0.205*** 0.046 0.340 0.244*** 0.205*** 0.039 0.260
(32.150) (14.21) (31.44) (14.14) (30.85) (13.96)
lntest 0.704*** 0.791*** −0.087 0.400 0.700*** 0.788*** −0.088 0.470 0.714*** 0.789*** −0.075 0.430
(23.780) (16.81) (23.85) (16.76) (24.70) (16.72)
lnpm25XlnICU_beds −0.030*** −0.012* −0.018 0.240
(-8.46) (-1.83)
lnpm25Xlnshelters −0.059*** −0.003 −0.056 0.070
(-13.82) (-0.43)
constant 4.704*** 3.655*** 0.699 0.440 −0.603** −1.210*** 0.607 0.470 −1.106*** −1.164*** 0.059 0.420
(36.68) (20.50) (-2.26) (-3.33) (-4.15) (-3.19)
day fixed effects YES YES YES YES YES YES
county fixed effects YES YES YES YES YES YES
R2 0.969 0.982 0.970 0.982 0.970 0.982
N 4156 1291 4156 1291 4156 1291
*p < 0.1, **p < 0.05, ***p < 0.01. Robust SEs are included in parentheses.
Table 5 Panel fixed effect model results for counties with low and high levels of social vulnerability.
Table 5(DV2 = lndeaths) Model 1 (baseline) Model 2 (icu_bed) Model 3 (shelter)
Low SVI High SVI bo-b1 p value Low SVI High SVI bo-b1 p value Low SVI High SVI bo-b1 p value
lnpm25 −0.008 −0.011 0.003 0.450 0.066*** 0.269*** −0.203 0.070 0.427*** 0.353*** 0.073 0.430
(-0.53) (-0.42) (2.65) (7.14) (9.48) (6.04)
fire_or_not 0.008 0.034 −0.026 0.470 0.006 0.073*** −0.067 0.360 0.012 0.041 −0.029 0.370
(0.73) (3.68) (0.28) (2.61) (0.59) (1.44)
lnICU_beds −0.132*** 0.055*** −0.187 0.180 −0.085*** 0.282*** −0.368 0.080 −0.127*** 0.036** −0.162 0.190
(-9.52) (3.89) (-4.50) (10.43) (-9.24) (2.54)
lnshelters −0.138*** −0.153*** 0.016 0.470 −0.131*** −0.117*** −0.014 0.460 0.026 0.044 −0.017 0.490
(-5.20) (-7.20) (-4.94) (-5.61) (0.86) (1.24)
Lnhosp 0.174*** −0.090*** 0.264 0.050 0.168*** −0.101*** 0.269 0.010 0.154*** −0.117*** 0.271 0.000
(11.21) (-3.85) (10.83) (-4.47) (9.99) (-5.02)
Lntest 0.315*** 2.135*** −1.820 0.000 0.314*** 2.116*** −1.801 0.000 0.335*** 2.096*** −1.761 0.000
(5.50) (28.24) (5.50) (29.14) (5.93) (28.22)
lnpm25XlnICU_beds −0.025*** −0.102*** 0.077 0.020
(-3.59) (-9.70)
lnpm25Xlnshelters −0.084*** −0.081*** −0.003 0.450
(-10.15) (-6.88)
constant 0.159 −14.06*** 14.217 0.000 0.016 −14.59*** 14.605 0.000 −0.797 −14.50*** 13.699 0.010
(0.30) (-24.13) (0.03) (-25.95) (-1.52) (-25.24)
day fixed effects YES YES YES YES YES YES
county fixed effects YES YES YES YES YES YES
R2 0.840 0.956 0.841 0.959 0.844 0.958
N 4069 1288 4069 1288 4069 1288
*p < 0.1, **p < 0.05, ***p < 0.01. Robust SEs are included in parentheses.
Regarding COVID-19 deaths, the results in Model 1 of Table 5 indicate that wildfire smoke has no direct effect on COVID-19 outcomes for counties of both groups. After including its interaction effects with both hospital and public housing resources (as shown in Models 2 and 3 in Table 5), respectively, the direct and indirect effects of wildfire smoke on COVID-19 deaths are significant and positive for counties with both high and low SVI. That being said, the larger estimated coefficient for the high SVI group indicates a higher direct impact of increased PM2.5 on COVID-19 deaths for areas experiencing higher versus lower social vulnerability (Model 2, Table 5), and the group difference is statistically significant at the 10 percent level. Likewise, the interaction effect between wildfire smoke and ICU beds exhibits a significant group difference with respect to COVID-19 deaths, with the high SVI group's estimated coefficient to be about four times larger than that of the low SVI group.
Taken together, Table 4, Table 5 results suggest that when a wildfire occurs in areas with a higher level of social vulnerability, an increase in the availability of staffed ICU beds is especially critical for reducing COVID-19 deaths. By contrast, hospital resources are critical in both counties with low and high levels of social vulnerability for curtailing COVID-19 infections.
5 Sensitivity analysis
Sensitivity analysis can help identify critical control points, prioritize additional data collection or research, and verify and validate a model (Christopher, Frey and Patil, 2002). We conduct sensitivity analysis in two ways to test the robustness of our findings. First, we construct an alternative measure of social vulnerability using the preliminary Climate and Economic Justice Screening Tool published by the White House's Council on Environmental Quality. This alternative measure is based on census tract-level climate and economic disadvantages.3, 4 Since the state of being disadvantaged is calculated as a binary variable at the census tract level in the Climate and Economic Justice Screening Tool, to match our county-level analysis, we create an index variable of climate and economic disadvantages at the county level using the number of tracts within a county that is disadvantaged divided by the total number of tracts within a county. Therefore, whether a county is identified as disadvantaged or not depends on a ratio variable that varies from 0 (no tract is disadvantaged) to 1 (all tracts are disadvantaged).
We employ the same model specifications as in Table 4, Table 5 and use the same cutoff point strategy (mean + 1 SD) to perform subsample analysis. As such, ratio_disadvantaged_high is recorded as 1 when this value is higher than 0.6101, indicating counties that are more disadvantaged in terms of climate and economy; otherwise, this variable is recorded as 0. In this way, we have identified 13 counties that are highly disadvantaged (denoted as EJ ratio high); among them, there are 7 counties that also exhibit a high overall SVI. Additional analyses conducted that follow the same subsample analysis approach validate our findings in the main analysis. By and large, sensitivity analyses that use the Climate and Economic Justice Screening Tool are consistent with findings in our main analysis (see Tables 6 and 7 in the supplementary materials).
Second, we construct a county-by-week panel dataset to verify whether the results of our county-by-day analysis hold at an aggregated level. An aggregated level of analysis allows for the accumulated effect of local communities’ exposure to PM2.5 from wildfire smoke. Following the approach in Chen et al. (2022), we convert the daily dataset into a weekly dataset using the mean value of a variable (PM2.5) and by summing the values of variables (fire_or_not, available_ICU_beds, public shelters, hospitalization, total tests) within a 7-day frame. We employ the cumulative counts of COVID-19 infection and death counts every seventh day within the research timeframe as our two dependent variables. Extra analysis with weekly dataset shows that while the signs on the estimated coefficients are as expected, statistical significance is not always achieved (see Tables 8 and 9 in the supplementary materials).
6 Discussion and conclusions
This study examines the direct and indirect effects of wildfires on COVID-19 infections and deaths through the moderating impact of hospital resources (as represented by available ICU beds) and public shelters and housing resources in local communities. The main findings of this study are as follows: First, during the 2020 wildfire season in California, when the COVID-19 pandemic was also raging, severe wildfire smoke represented by PM2.5 is significantly associated with the increases in COVID-19 infections and deaths, which might be a result of the compounded impact of wildfire smoke on the human respiratory system. This result is consistent with findings in medical science studies such as Zhou et al. (2021) and Cortes-Ramirez et al. (2022).
Second, wildfires have an indirect effect on COVID-19 infections and deaths by way of the availability of hospital resources. Providing more staffed ICU beds during the wildfire season reduced COVID-19 infections and deaths. Third, similarly, wildfires have an indirect effect on COVID-19 outcomes through the availability of public shelters and housing resources. Providing more shelters and housing resources to the public also reduced COVID-19 infections and deaths. Furthermore, both the direct effect of wildfire smoke and its indirect effects through wildfires’ interactions with hospital resources on COVID-19 outcomes (notably COVID-19 deaths) are more severe for communities that exhibit high social vulnerability or environmental disadvantages.
Our findings complement recent research that shows a link between community-level social vulnerability, policy interventions and capacity, and COVID-19 outcomes (Gupta and Sahoo, 2020; Karaca-Mandic et al., 2020; Qian and Jiang, 2022). Policy interventions, such as social distancing, availability of public shelters and staffed ICU beds are critical for reducing the severity of COVID-19 outcomes, and the provision of social assistance and professional medical aid is particularly urgent and in great demand in counties exhibiting high social vulnerability.
This study faces certain limitations. First, given data availability, we used panel data for only one state rather than for multiple states that have experienced both wildfires and the COVID-19 pandemic to test our hypotheses. As such, our findings may not be generalizable to other states or other emergency contexts, such as floods, droughts, and snowstorms. Second, we are unable to explicitly account for factors such as conditions of local emergency management staffing, budget, assets, and professional or administrative expertise that might affect the emergency management performance of local communities. That being said, assuming that these factors vary by time or county, we have accounted for them through day and county fixed effects. Third, findings on the robust checks indicate that our main model results are somewhat sensitive to sample size, with partial support for our hypotheses. We believe that the county-by-day analysis (our main analysis) provides the most appropriate level of analysis because it allows us to exploit daily variations in our DVs (COVID-19 infections and deaths) and main IVs (wildfire smoke and hospital and shelter resources) for a more statistically powerful estimation.
Despite the research limitations, this paper's findings have significant policy and management implications for local responses to future emergencies, which have become more frequent in the age of climate change, from wildfires to other natural disasters as well as to new variants of the COVID-19 virus and other types of rapid infectious diseases that will challenge local hospitals and housing resources. First, measures can be taken by public organizations to facilitate effective responses to the (co)occurrence or reoccurrence of different types of natural or human-made hazards. Increasing and expanding public shelters and temporary public housing through the use of contracted private hotels may be one direction for identifying appropriate measures to contain the spread of infectious diseases (e.g., mumps) and help wildfire evacuees shelter in safe places. Notably, beyond enhancing provision, answering the question of how to manage public housing resources appropriately to prevent outbreaks of infectious diseases in homeless service sites and shelters is critical to protect already vulnerable social groups.
Second, focusing on the provision and distribution issues related to hospital and healthcare resources could be another direction that health stakeholders and governments can take to prevent and mitigate the severe consequences of compound emergencies. On the one hand, as advocated by some clinical scholars (Dzierba et al., 2020; Melman et al., 2021; Supady et al., 2021), hospitals should develop emergency plans for situations in which medical resources and staffing cannot meet urgent needs. For example, how can scarce hospital resources be distributed in a balanced manner, and how can medical resources and scarce intensive care resources be allocated and managed to accommodate patients with different levels of COVID-19 symptoms and different diseases? How can medical resources and equipment, such as ventilators, be mobilized across different hospitals? To what extent can the variety and availability of health resources meet public needs in an emergency situation? These questions are currently being addressed and deserve more attention from hospitals and healthcare systems.
On the other hand, policymakers have the imperative to intervene when national emergencies and disasters occur when the market system alone may not protect all citizens. Specifically, providing or mobilizing medical resources and equipment across the country, implementing policies to enforce stay-at-home orders and mask mandates, and providing economic assistance to individuals, families, small businesses, and hospitals are critical to reduce the impacts of COVID-19 and other types of infectious diseases (Miao et al., 2021; An et al., 2021; Bel et al., 2021; Dzigbede et al., 2020; Menifield and Clark, 2021).
Third, the differential impacts of wildfire smoke on COVID-19 outcomes across communities with high and low levels of social vulnerability suggest that environmental justice is a critical issue that needs to be addressed (Huang and London, 2012; Sadd et al., 2011; Thomas et al., 2020). In what ways can communities with low emergency response capacity protect their residents from the disproportionate impacts of various disasters, particularly when multiple disasters occur simultaneously? This is an important question that requires further investigation. Although challenges ensue, federal and local governments, in partnership with public and private sector entities, can help to narrow the gap between disaster needs and disaster service provision both nationwide and within specific localities.
Finally, although our paper confirms the findings of many environmental management studies (e.g., Contini and Costabile, 2020; Travaglio et al., 2021; Wu et al., 2020) concerning the direct effect of wildfire smoke on COVID-19 outcomes, previous research has not examined the moderating effects of hospital and housing resources in the local community during the wildfire season. Therefore, our findings also highlight the need for more research on the relationships between wildfires (and wildfire smoke) and COVID-19 outcomes in the context of other types of local responses, such as the disbursement of public and private funding resources to suppress wildfires and collaboration among multiple administrative units and with external private sector partners during emergencies. These investigations provide directions for our future research.
Credit author statement
Suyang Yu: Conceptualization, Methodology, Software, Data curation, Formal analysis, Investigation and Visualization, Project administration, Writing – original draft, Writing – review & editing. Lily Hsueh: Conceptualization, Methodology, Software, Formal analysis, Investigation and Visualization, Resources, Validation, Writing – review & editing; Project administration, Supervision.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Notes
1. Appendix Table A provides summary statistics on several key socioeconomic variables for the state of California (all 58 counties), this study's sample, and counties not in the study. There are 14 counties that did not report PM2.5 nor provided public shelters on a daily basis. These counties are Alpine, Amador, Colusa, El Dorado, Glenn, Inyo, Lassen, Mendocino, Modoc, Sierra, Tehama, Trinity, Tuolumne, and Yuba. These are sparsely populated, rural counties with less diverse populations and experienced on average one to two more fire days in a given month compared to the California average and to the counties in this study's empirical analysis, respectively. Moreover, these counties on average recorded fewer COVID-19 infections and deaths during the 2020 fire season. Notwithstanding, when compared to all of California (58 counties) on key socioeconomic variables (e.g., total population, overall SVI), this study's sample is not statistically different from the entire state as whole.
2. Aside from county fixed effects, we do not separately control for county-level variation in masking because by mid-June 2020 California Governor Gavin Newsom instituted a mask mandate in which he ordered all Californians to wear face covering while in public or high-risk settings (Source: https://twitter.com/GavinNewsom/status/1273696999066353664). California's state-level mandate helps to preclude the possibility that wildfires led to migration into crowded areas, which increased COVID-19 transmission because of the exposure to the virus due to close proximity.
3. The White House's Council and Economic Justice Screen Tool (CEJST) is in its beta version. It is used by President Joseph Biden's Justice40 Initiative and related programs across the federal government to identify disadvantaged communities. That being said, the CEJST is still evolving and likely to undergo substantial changes. Source: https://screeningtool.geoplatform.gov/en/#3/33.47/-97.5 (Retrieved September 12, 2022).
4. We also conduct a separate analysis using the COVID Community Vulnerability Index (CCVI) developed by Surgo Ventures (https://precisionforcovid.org/ccvi). Results based on the CCVI are very similar to our main model results. Analysis is available upon request from the authors.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Appendix A Supplementary data
The following is the Supplementary data to this article:Multimedia component 1
Multimedia component 1
Data availability
Data used in this paper are publicly available.
Acknowledgments
The authors thank Meri Davlasheridze, Yushim Kim, Qing Miao, Ivan Petkov, Aseem Prakash, and participants at the Environmental Justice Policy Initiative seminar at Arizona State University, the 2021 Environmental Politics Workshop at the University of Washington, and the “Disasters, Emergencies, and Implications for Public Responses” panel session at the 2022 APPAM Research Conference for helpful feedback in the development of this project and manuscript.
Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.jenvman.2022.116918.
==== Refs
References
Aguilera R. Corringham T. Gershunov A. Benmarhnia T. Wildfire smoke impacts respiratory health more than fine particles from other sources: observational evidence from Southern California Nat. Commun. 12 2021 1493 10.1038/s41467-021-21708-0 33674571
Alman B.L. Pfister G. Hao H. Stowell J. Hu X. Liu Y. Strickland M.J. The association of wildfire smoke with respiratory and cardiovascular emergency department visits in Colorado in 2012: a case crossover study Environ. Health 15 2016 64 10.1186/s12940-016-0146-8 27259511
Bel G. Gasulla Ó. Mazaira-Font F.A. The effect of health and economic costs on governments' policy responses to COVID-19 crisis under incomplete information Publ. Adm. Rev. 81 2021 1131 1146 10.1111/puar.13394
Biggs E.N. Maloney P.M. Rung A.L. Peters E.S. Robinson W.T. The relationship between social vulnerability and COVID-19 incidence among Louisiana Census Tracts Front. Public Health 8 2021 1048 10.3389/fpubh.2020.617976
Burke Marshall Driscoll Anne Heft-Neal Sam Xue Jiani Burney Jennifer Wara Michael The changing risk and burden of wildfire in the United States Proc. Natl. Acad. Sci. USA 118 2 2021 e2011048118 10.1073/pnas.2011048118
Cabral J. Cuevas A.G. Health inequities among Latinos/Hispanics: documentation status as a determinant of health J. Racial Ethn. Health Disparities 7 2020 874 879 10.1007/s40615-020-00710-0 32026286
Chen B. Liu M. Ye W. Zhang B. Assessing the impact of green nudges on ozone concentration: evidence from China's night refueling policy J. Environ. Manag. 312 2022 114899 10.1016/j.jenvman.2022.114899
Christopher Frey H. Patil S.R. Identification and review of sensitivity analysis methods Risk Anal. 22 2002 553 578 10.1111/0272-4332.00039 12088234
Conticini E. Frediani B. Caro D. Can atmospheric pollution be considered a co-factor in extremely high level of SARS-CoV-2 lethality in Northern Italy? Environ. Pollut. 261 2020 114465 10.1016/j.envpol.2020.114465
Contini D. Costabile F. Does air pollution influence COVID-19 outbreaks? Atmosphere 11 2020 377 10.3390/atmos11040377
Cortes-Ramirez J. Michael R.N. Knibbs L.D. Bambrick H. Haswell M.R. Wraith D. The association of wildfire air pollution with COVID-19 incidence in New South Wales, Australia Sci. Total Environ. 809 2022 151158 10.1016/j.scitotenv.2021.151158
Cova T.J. Drews F.A. Siebeneck L.K. Musters A. Protective actions in wildfires: evacuate or shelter-in-place? Nat. Hazards Rev. 10 2009 151 162 ASCE)1527-6988(2009)10:4(15110.1061/
Davlasheridze M. Miao Q. Natural disasters, public housing, and the role of disaster aid J. Reg. Sci. 61 2021 1113 1135 https://onlinelibrary.wiley.com/doi/abs/10.1111/jors.12534
Deek S.A. Chronic exposure to air pollution implications on COVID-19 severity Med. Hypotheses 145 2020 110303 10.1016/j.mehy.2020.110303
Delfino R.J. Brummel S. Wu J. Stern H. Ostro B. Lipsett M. Winer A. Street D.H. Zhang L. Tjoa T. Gillen D.L. The relationship of respiratory and cardiovascular hospital admissions to the southern California wildfires of 2003 Occup. Environ. Med. 66 2009 189 197 10.1136/oem.2008.041376 19017694
Dzierba A.L. Pedone T. Patel M.K. Ciolek A. Mehta M. Berger K. Ramos L.G. Patel V.D. Littlefield A. Chuich T. May H.B. Muir J. Verkerk B.S. Poon T. Der-Nigoghossian C. Nikolos P. Gunther M. Shah J. Witenko C.J. Rethinking the drug distribution and medication management model: how a New York city hospital pharmacy department responded to COVID-19 J. Am. Coll. Clin. Pharm. 3 2020 1471 1479 10.1002/jac5.1316 32838224
Dzigbede K. Gehl S.B. Willoughby K. Disaster resiliency of U.S. local governments: insights to strengthen local response and recovery from the COVID-19 pandemic Publ. Adm. Rev. 80 2020 634 643 10.1111/puar.13249
Freese K.E. Vega A. Lawrence J.J. Documet P.I. Social vulnerability is associated with risk of COVID-19 related mortality in U.S. counties with confirmed cases J. Health Care Poor Underserved 32 2021 245 257 10.1353/hpu.2021.0022 33678695
Goodling E. Intersecting hazards, intersectional identities: a baseline critical environmental justice analysis of US homelessness Environ. Plan. E Nat. Space 3 2019 833 856 10.1177/2514848619892433
Gupta S. Sahoo S. Pandemic and mental health of the front-line healthcare workers: a review and implications in the Indian context amidst COVID-19 Gen. Psychiatry 33 2020 e100284 10.1136/gpsych-2020-100284
Hamideh S. Sen P. Fischer E. Wildfire impacts on education and healthcare: paradise, California, after the camp fire Nat. Hazards 111 2022 353 387 10.1007/s11069-021-05057-1
Henderson S.B. The COVID-19 pandemic and wildfire smoke: potentially concomitant disasters Am. J. Publ. Health 110 2020 1140 1142 10.2105/ajph.2020.305744
Isphording I.E. Pestel N. Pandemic meets pollution: poor air quality increases deaths by COVID-19 J. Environ. Econ. Manag. 108 2021 102448 10.1016/j.jeem.2021.102448
Karaca-Mandic P. Sen S. Georgiou A. Zhu Y. Basu A. Association of COVID-19-related hospital use and overall COVID-19 mortality in the USA J. Gen. Intern. Med. 2020 1 3 10.1007/s11606-020-06084-7
Magazzino C. Mele M. Schneider N. The relationship between air pollution and COVID-19-related deaths: an application to three French cities Appl. Energy 279 2020 115835 10.1016/j.apenergy.2020.115835
Maxwell A.E. Vargas C. Santifer R. Chang L.C. Crespi C.M. Lucas-Wright A. Facilitators and challenges to health promotion in black and Latino churches J. Racial Ethn. Health Disparities 9 2020 59 67 10.1007/s40615-020-00929-x 33230738
Melman G.J. Parlikad A.K. Cameron E.A.B. Balancing scarce hospital resources during the COVID-19 pandemic using discrete-event simulation Health Care Manag. Sci. 24 2021 356 374 10.1007/s10729-021-09548-2 33835338
Melnychuk E. Sallade T.D. Kraus C.K. Hospitals as disaster victims: lessons not learned? J. Am. Coll. Emerg. Physicians Open 3 2022 e12632 10.1002/emp2.12632
Menifield C.E. Clark C. Pandemic planning in the United States: an examination of COVID-19 data Publ. Adm. Rev. 81 2021 1102 1109 10.1111/puar.13326
Miao Q. Shi Y. Davlasheridze M. Fiscal decentralization and natural disaster mitigation: evidence from the United States Public Budg. Finance 41 2021 26 50 10.1111/pbaf.12273
Naqvi H.R. Mutreja G. Shakeel A. Singh K. Abbas K. Naqvi D.F. Chaudhary A.A. Siddiqui M.A. Gautam A.S. Gautam S. Naqvi A.R. Wildfire-induced pollution and its short-term impact on COVID-19 cases and mortality in California Gondwana Res. 2022 10.1016/j.gr.2022.04.016
Omori R. Mizumoto K. Chowell G. Changes in testing rates could mask the novel coronavirus disease (COVID-19) growth rate Int. J. Infect. Dis. 94 2020 116 118 10.1016/j.ijid.2020.04.021 32320809
Phillips C.A. Caldas A. Cleetus R. Dahl K.A. Declet-Barreto J. Licker R. Merner L.D. Ortiz-Partida J.P. Phelan A.L. Spanger-Siegfried E. Talati S. Trisos C.H. Carlson C.J. Compound climate risks in the COVID-19 pandemic Nat. Clim. Change 10 2020 586 588 10.1038/s41558-020-0804-2
Prunicki Mary Kelsey Rodd Lee Justin Zhou Xiaoying Smith Edward Haddad Francois Wu Joseph Nadeau Kari The impact of prescribed fire versus wildfire on the immune and cardiovascular systems of children Allergy 74 10 2019 1989 1991 10.1111/all.13825 31002401
Qian M. Jiang J. COVID-19 and social distancing J. Public Health 30 2022 259 261 10.1007/s10389-020-01321-z 32837835
Ranney M.L. Griffeth V. Jha A.K. Critical supply shortages - the need for ventilators and personal protective equipment during the COVID-19 pandemic N. Engl. J. Med. 382 2020 e41 10.1056/NEJMp2006141 32212516
Reid C.E. Considine E.M. Watson G.L. Telesca D. Pfister G.G. Jerrett M. Associations between respiratory health and ozone and fine particulate matter during a wildfire event Environ. Int. 129 2019 291 298 10.1016/j.envint.2019.04.033 31146163
Rodrigues M. Gelabert P.J. Ameztegui A. Coll L. Vega-García C. Has COVID-19 halted winter-spring wildfires in the Mediterranean? Insights for wildfire science under a pandemic context Sci. Total Environ. 765 2020 142793 10.1016/j.scitotenv.2020.142793
Sadd J.L. Pastor M. Morello-Frosch R. Scoggins J. Jesdale B. Playing it safe: assessing cumulative impact and social vulnerability through an environmental justice screening method in the South Coast Air Basin, California Int. J. Environ. Res. Publ. Health 8 2011 1441 1459 10.3390/ijerph8051441
Salas R.N. Shultz J.M. Solomon C.G. The climate crisis and COVID-19 - a major threat to the pandemic response N. Engl. J. Med. 383 2020 e70 10.1056/NEJMp2022011 32668132
Schranz C.I. Castillo E.M. Vilke G.M. The 2007 San Diego wildfire impact on the emergency department of the University of California, San Diego hospital system Prehospital Disaster Med. 25 2010 472 476 10.1017/s1049023x0000858x 21053198
Settembrino M.R. Sometimes you can't even sleep at night": social vulnerability to disasters among men experiencing homelessness in Central Florida Int. J. Mass Emergencies Disasters 35 2017 30 48
Sharfstein J.M. Becker S.J. Mello M.M. Diagnostic testing for the novel coronavirus JAMA 323 2020 1437 1438 10.1001/jama.2020.3864 32150622
Shultz J.M. Kossin J.P. Ali A. Borowy V. Fugate C. Espinel Z. Galea S. Superimposed threats to population health from tropical cyclones in the prevaccine era of COVID-19 Lancet Planet. Health 4 2020 e506 e508 10.1016/s2542-5196(20)30250-3 33159875
Supady A. Curtis J.R. Abrams D. Lorusso R. Bein T. Boldt J. Brown C.E. Duerschmied D. Metaxa V. Brodie D. Allocating scarce intensive care resources during the COVID-19 pandemic: practical challenges to theoretical frameworks Lancet Respir. Med. 9 2021 430 434 10.1016/s2213-2600(20)30580-4 33450202
Thomas D.S.K. Jang S. Scandlyn J. The CHASMS conceptual model of cascading disasters and social vulnerability: the COVID-19 case example Int. J. Disaster Risk Reduc. 51 2020 101828 10.1016/j.ijdrr.2020.101828
Travaglio M. Yu Y. Popovic R. Selley L. Leal N.S. Martins L.M. Links between air pollution and COVID-19 in England Environ. Pollut. 268 2021 115859 10.1016/j.envpol.2020.115859
Watkins S.L. Gerrish E. The relationship between urban forests and race: a meta-analysis J. Environ. Manag. 209 2018 152 168 10.1016/j.jenvman.2017.12.021
Whittaker J. Blanchi R. Haynes K. Leonard J. Opie K. Experiences of sheltering during the Black Saturday bushfires: implications for policy and research Int. J. Disaster Risk Reduc. 23 2017 119 127 10.1016/j.ijdrr.2017.05.002
Wong S.D. Broader J.C. Shaheen S.A. Review of California Wildfire Evacuations from 2017 to 2019 2020 UC Office of the President: University of California Institute of Transportation Studies Berkeley, CA
Wong S.D. Walker J.L. Shaheen S.A. Trust and compassion in willingness to share mobility and sheltering resources in evacuations: a case study of the 2017 and 2018 California wildfires Int. J. Disaster Risk Reduc. 52 2021 101900 10.1016/j.ijdrr.2020.101900
Wright J.E. 2nd Merritt C.C. Social equity and COVID-19: the case of African Americans Publ. Adm. Rev. 80 2020 820 826 10.1111/puar.13251
Wu X. Nethery R.C. Sabath B.M. Braun D. Dominici F. Exposure to Air Pollution and COVID-19 Mortality in the United States: a Nationwide Cross-Sectional Study 2020 10.1101/2020.04.05.20054502 medRxiv
Zangrillo A. Beretta L. Silvani P. Colombo S. Scandroglio A.M. Dell'Acqua A. Fominskiy E. Landoni G. Monti G. Azzolini M.L. Monaco F. Oriani A. Belletti A. Sartorelli M. Pallanch O. Saleh O. Sartini C. Nardelli P. Lombardi G. Morselli F. Scquizzato T. Frontera A. Ruggeri A. Scotti R. Assanelli A. Dagna L. Rovere-Querini P. Castagna A. Scarpellini P. Di Napoli D. Ambrosio A. Ciceri F. Tresoldi M. Fast reshaping of intensive care unit facilities in a large metropolitan hospital in Milan, Italy: facing the COVID-19 pandemic emergency Crit. Care Resusc. 22 2020 91 94 10.51893/2020.2.pov1 32227819
Zhou X. Josey K. Kamareddine L. Caine M.C. Liu T. Mickley L.J. Cooper M. Dominici F. Excess of COVID-19 cases and deaths due to fine particulate matter exposure during the 2020 wildfires in the United States Sci. Adv. 7 2021 eabi8789 10.1126/sciadv.abi8789
| 0 | PMC9705198 | NO-CC CODE | 2022-12-16 23:21:36 | no | J Environ Manage. 2023 Feb 15; 328:116918 | utf-8 | J Environ Manage | 2,022 | 10.1016/j.jenvman.2022.116918 | oa_other |
==== Front
Vaccine
Vaccine
Vaccine
0264-410X
1873-2518
Published by Elsevier Ltd.
S0264-410X(22)01487-6
10.1016/j.vaccine.2022.11.062
Article
What Attributes Influence Rural Household’s Willingness to Get Vaccinated for COVID-19? Perspectives from Six Chinese Townships
Sun Yingying a
Huang Shih-Kai b
Arlikatti Sudha c⁎
Lindell Michael K. d
a School of Public Administration and Policy, Renmin University of China, 59 Zhongguancun Street, Beijing 100872, P.R. China
b Jacksonville State University, Jacksonville, Alabama, USA
c Amrita School for Sustainable Development, Amrita Vishwa Vidyapeetham, Amritapuri, India
d University of Washington, Seattle, Washington, USA
⁎ Corresponding author.
29 11 2022
29 11 2022
3 7 2022
22 11 2022
25 11 2022
© 2022 Published by Elsevier Ltd.
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Since the onset of the COVID-19 pandemic, vaccinations have been identified as the most effective mitigation strategy against the deadly virus. This has led developed nations to accelerate research and shorten the licensure process for COVID-19 vaccines, but these changes have caused widespread concerns about vaccine safety. Research literature has long indicated that citizens’ perceptions of protective actions will determine their behaviors, and thus, the relationship between vaccine perception and vaccination intention needs to be assessed. To better understand vaccination willingness, especially in rural populations, this study surveyed 492 households from six townships in the Ya’an region of China’s Sichuan Province in November 2020. The survey followed the Protective Action Decision Model (PADM) framework for collecting perceptions about the influenza and COVID-19 vaccines as protective actions, information sources, emergency preparedness, emotional response, and demographic characteristics. The results showed that influenza vaccine perceptions significantly affected people’s COVID-19 vaccination perceptions and intentions. Unlike previous vaccination willingness and other COVID-19 studies, this study found that perceptions of resource-related attributes and health-related attributes both affected COVID-19 vaccination intentions, but the former were slightly stronger than the latter. Moreover, these effects were strongest among respondents who had the most positive perceptions of their influenza vaccine experience. This study’s findings will benefit local authorities in designing appropriate policies and measures (e.g., hazard education, risk communication, vaccination convenience enhancement) for increasing vaccination compliance for the current and future pandemics.
Keywords
COVID-19
vaccination willingness
protective action perceptions
information sources
Sichuan townships
==== Body
pmc1 Introduction
The novel coronavirus (COVID-19) pandemic has caused unprecedented human and financial losses across the globe. Reflecting back on the progression of the pandemic two and a half years after its onset, it is observed that national governments across the world quickly agreed with the World Health Organization (WHO) that increasing vaccination coverage against this deadly virus was the most effective mitigation strategy [59]. In the first year of the pandemic, many nations accelerated research and shortened their licensure processes for COVID-19 vaccines in order to achieve herd immunity and reduce the pandemic’s deadly impacts [41], [94] . By the end of 2021, the effort had allowed over 70% of the population in high-income countries to receive at least one dose of a COVID-19 vaccine (Our World Data[66]. However, the rise in new variants, vaccine shortages in some countries, and low vaccination coverage in rural areas have made achieving herd immunity difficult to achieve (WHO 2022). Nevertheless, recent studies (e.g., [31] have revealed a significant association between vaccination compliance and patients’ likelihood of overcoming severe symptoms, hospitalization, and death, irrespective of the types of variants. This implies that increasing vaccination uptake continues to be crucial in the battle against COVID-19 or in tackling other emerging contagious diseases of the future. Thus, the present study focuses on vaccine receivers, specifically on rural Chinese populations’ perceptions of the attributes that contribute to their willingness of to get vaccinated, long before the COVID-19 vaccination campaigns were popularized throughout the world. Such research is important for ensuring that health departments can adopt risk communication strategies to increase vaccine uptake by addressing impediments that dissuade populations from getting vaccinated.
Currently, individuals’ vaccination hesitancy (i.e., vaccination delays or vaccination denial) is the greatest obstacle to motivate COVID-19 vaccination [8], [17] . Scholars have published more than 50 articles examining people’s vaccination intentions throughout the world. However, the majority of those articles focused on the relationship between people’s perceptions of the vaccine’s health impacts (efficacy and safety) and their vaccination willingness [64], [82] . Nonetheless, these articles have overlooked other attributes that are likely to affect the Protective Action Decision-Making process (PADM— [52], [54]. For example, previous literature has indicated that people’s confidence in government health agencies and the perceived uncertainty about vaccine safety are important factors associated with vaccine hesitancy [11]. Similarly, the Andrew Wakefield case reveals a growing concern regarding the social influence of fake information and false experience on vaccine hesitancy [2], [42], [44]while tackling these social influences would require a greater amount of energy than determining how it has been shaped[12], [10]. In this regard, the Chinese population have recently witnessed several vaccine-related scandals including the 2018 scandal in which the biomedical company Changchun Changsheng Biotechnology falsified data, bribed officials, and distributed substandard Diphtheria, Pertussis, and Acellular Tetanus (DTaP) vaccines for infants [23]. Other incidents include the illegal sale of Category II vaccines for Rabies and Varicella viruses in the Shandong Province in 2016 (L. [14] and suspicious infant deaths following the use of a Hepatitis B vaccine in 2013 (B. [19]. All of these incidents have undermined Chinese residents’ confidence in their pharmaceutical industry. Furthermore, concerns about COVID-19 vaccine safety rose exponentially since the vaccine’s approvals was fast tracked and emergency use authorizations granted, unlike other time-tested vaccines [22], [47] .
Despite accounting for only 36.1% of the national population, the Chinese rural population’s access to medical resources—including medical staff, operating rooms and surgical equipment—has remained limited with less than 20.0% of the national health budget earmarked for them [30], [91] . Moreover, rural China is also home to a growing aging population from a lower socio-economic background, increasing their pandemic vulnerability in many ways [57], including lower educational attainment, limited health care infrastructure, greater travel distance, inability to take time off from work, and even political party preference. In turn, these factors reduce vaccination compliance in rural areas (Motta et al., 2021; [78]. Previous studies have also speculated that such vulnerabilities amplify the social influence of fake information and false experiences [40]. It underscores the WHO’s call to mitigate the challenges faced in rural areas across the globe as a key to ending the pandemic (Murthy et al., 2021). Hence, to better understand rural households’ willingness to get a COVID-19 vaccine and to examine associated factors influencing protective action decision-making, a China-US collaborative team carried out a field study before any of the COVID-19 vaccines became available.
Empirical data based on the PADM (Lindell, 2018; [54] were collected by conducting face-to-face interviews with 492 respondents’ from November 24-28, 2020 from six rural townships in the Ya’an region of Sichuan Province. The study addressed three broad questions: 1) What do rural residents think about the COVID-19 vaccine in comparison to the familiar influenza vaccine in terms of health-related (i.e., efficacy and safety) and resource-related (i.e., information and availability) attributes? 2) How do the comprehensive protective action assessments (i.e., influenza vaccine confidence and COVID-19 vaccine perceptions) affect residents’ COVID-19 vaccination willingness? 3) Do perceptions of the influenza vaccine interact with their perceptions of the COVID-19 vaccine to influence COVID-19 vaccination willingness? This line of questioning will help identify how rural populations can be reached with targeted interventions to increase uptake of the current COVID-19 vaccine, as well as vaccines against future pandemics.
2 Literature Review
2.1 Theoretical Framework
Factors that influence households’ willingness to adopt protective actions are of great interest to disaster researchers and policy makers. Lindell and Perry[54]drew relevant elements from sociological and psychological communication and behavioral models in proposing the PADM to explain people’s decision-making processes in response to environmental hazards. The PADM describes how people typically process information received from community stakeholders through multiple communication channels to decide on whether to take a protective action. Specifically, factors influencing this decision making process include environmental and social cues; social information sources/channels; prior emergency preparedness actions; perceptions of the threat, protective actions, and stakeholders; and demographic characteristics. This current study focuses on protective action perceptions, information sources, emergency preparedness, emotional response, and demographic variables.
2.1.1 Protective Action Perceptions
When people assess a protective action, they consider two primary groups of attributes—health-related and resource-related attributes. The health-related attributes (more generally known as hazard-related attributes) address whether the protective action will be effective in protecting themselves, their families, and property against a hazard’s impacts. The resource-related attributes are what taking the protective action will require in terms of cost, time/effort, knowledge/skill, specialized tools/equipment, and cooperation with others [53], [54]) . Using the PADM framework, F. Wang et al. [83]) investigated people’s willingness to adopt protective actions during the 2013 Chinese H7N9 outbreak and found that health-related attributes were positively related to people’s behavioral expectations for adopting protective actions, whereas resource-related attributes were negatively related to it. However, other studies have indicated that the association between protective action perceptions and behavioral intentions are inconsistent, and that resource-related attributes have small and non-significant correlations with adoption intentions of protective actions[53], [81], [86].
Recent studies on the uptake of COVID-19 vaccines have shown that the health-related attributes (efficacy and safety) are the most important factors influencing vaccination intentions [17], [64] . A vaccine’s resource-related attributes such as cost, convenience, and timeliness are significant, but have smaller effects than the health-related attributes in Bangladesh [1]), [11] . However, in Bangladesh and China, the rate of vaccination intention decreased if there was a fee to get it versus if it was free [1], [90] . Nevertheless, to date, vaccination intentions and their determinants remain unexamined in the context of rural China.
Scholars have found that new vaccines usually engender more hesitancy than familiar ones [23], [24] . After the outbreak of the COVID-19 in China, the number of influenza vaccinations increased from 2020 to 2021 [33]. Nonetheless, influenza vaccine coverage was quite low in China, with only 0.4% coverage for influenza vaccination, compared to 22.8% coverage for pneumonia vaccination (Y. [85]. Also, influenza vaccination uptake in rural areas was lower than in urban areas (Y. [85]. These data suggest that, to some members of the rural Chinese population, the influenza vaccine was perceived as a new product, just like the COVID-19 vaccine. An online survey in China revealed that people who had higher confidence in domestic vaccine safety were more likely to accept a COVID-19 vaccine (M. [20]. Hence, it is worth exploring whether influenza vaccination perceptions would influence COVID-19 vaccine acceptance.
2.1.2 Information Sources
In the process of protective action decision making against impending threats, information seeking occurs when there is uncertainty and the available information is insufficient to justify immediate actions[6], [54]) . Information seeking refers to active and purposeful searching, obtaining, clarifying and confirming information in order to achieve sufficient information for decision making[39]). The early stage of the COVID-19 outbreak was marked by a lack of knowledge about the disease transmission dynamics and divergent views on the effectiveness of NPIs [4], [77] . Unsurprisingly, this knowledge gap led to information seeking from multiple sources and through multiple channels to assess the certainty, severity, and immediacy of the threat [28], [60]. Consistent with research on seasonal influenza[86], people were concerned about different exposure paths such as close proximity to infected potentially persons and handling potentially contaminated objects (F.[83]. Scholars have found that individuals exposed to health stressors, who therefore experience greater health concerns, search more frequently for information from various governmental, non-profit, and private information sources to help make decisions to protect themselves and their families[77]. Moreover, health information seekers access various information sources such as technology-based sources, print sources, and human sources, depending on channel access and preferences that vary among different groups and likely influence their willingness to get vaccinated[33], [76].
Related to trust in information sources, M. Chen et al. [20] surveyed 3,195 Chinese respondents and reported that those who paid little or no attention to information about COVID-19 vaccine development were less willing (66.3%) to get the vaccine compared to those who did (85.5%). Additionally, those that had heard of the Changchun Changsheng Biotechnology Company scandal in 2018 had significantly higher levels of doubt about the domestic vaccines [23]. Similarly, Lahijani et al. [46] found that persons who mistrusted the US healthcare system and the information provided in a commercial broadcast about the human papillomavirus (HPV) prevention strategies were less likely to get the HPV vaccine. Conversely, Latkin et al. [47] reported that those who trusted in information sources like the Centers for Disease Control and Prevention (CDC), Johns Hopkins University, and the State Health Department were more likely to trust the vaccine. However, trust in mainstream news channels such as CNN, and the White House did not show significant correlations with trust in vaccines. Given these mixed results, studies are needed in rural China to elucidate the linkages, if any, between trust in information sources and COVID-19 vaccine uptake are lacking and needed.
2.1.3 Emergency Preparedness
Emergency preparedness supports active response when an emergency has occurred. A study of the US Department of Veterans Affairs (VA) Home Based Primary Care (HBPC) program indicated that, the early implementation of preparedness procedures by medical and nursing staff (e.g. categorizing patients into a risk category group, hurricane evacuation planning etc.) limited disruptions in patient care and prevented significant hospitalizations of medically vulnerable populations [88]. However, the immediacy of taking actions varied according to the types of emergency preparedness measures. Nakaya et al. [61] found that the rate of tsunami evacuations was significantly higher amongst people who had participated in tsunami disaster drills during normal times than those who had not. However, as was the case in American Samoa [55], other measures like attending a lecture about tsunamis did not influence evacuation behavior.
In the context of infectious disease outbreaks, emergency preparedness includes stockpiling face masks and disinfectants [49] and proper vaccinations and cross-immune protection, especially for people with increased susceptibility to an epidemic such as chemotherapy for cancer, age and underlying illnesses (Priyadarshini et al. 2020). However, even though risk management experts put much effort into encouraging emergency preparedness in the pre-crisis stage, until now, preparedness has usually been targeted for natural disasters. For example, residents in earthquake prone areas have long been advised to have a working battery-powered radio with spare batteries, at least 4 gallons of water in plastic containers, a complete first-aid kit, and a 4-day supply of dehydrated or canned food [58]. However, emergency preparedness is rarely targeted at new and unknown public health threats [83]. Thus, many countries throughout the world experienced personal protective equipment shortages in the early stage of COVID-19, especially among medical staffs [26]. With the development of COVID-19 vaccines, it is crucial to examine whether people’s uptake is also related to people’s prior adoption of these simple emergency preparedness measures.
2.1.4 Emotional Response
In addition to taking protective actions, emergency response is also characterized by emotional reactions that are experienced after observing environmental or social cues, or receiving warnings [54]). Scholars have tried to interpret the different kinds of emotions such as fear, shock, and uncertainty, that emerge in response to a threat [74]) and their positive or negative impacts on adopting protective actions [43], [93] . The COVID-19 pandemic has caused a heightened state of negative emotions such as fear, depression, and anxiety experienced by students, parents, school teachers, healthcare workers, and medical professionals [29], [93] . Interestingly, W. Cao et al. [15] found that 25.0% of Chinese college students who experienced anxiety in the early stages of the pandemic reported that living in urban areas helped reduce this anxiety.
Such negative emotional responses caused a greater willingness to get vaccinated by physicians who were in direct contact with infected patients, compared to administrative healthcare assistants who were not [79]. In an earlier US study, women were found to have higher intentions of getting the influenza vaccine, as they also had greater concerns about contracting the virus than men [34]. Such differences in emotional responses and their impacts on vaccine acceptance by varying population demographics needs to be studied in the rural contexts as well.
2.2 Research Objectives and Hypotheses
The preceding literature review generated three research objectives, five research hypotheses, and one research question.
Objective 1: To better understand how rural respondents assess an unknown COVID-19 vaccine compared to a known influenza vaccine, in terms of health-related and resource-related attributes.
RH1: Respondents’ vaccine ratings will be higher on health-related attributes than on resource-related attributes.
RH2: Respondents’ vaccine ratings on health-related and resource-related attributes will yield higher levels of agreement for the influenza vaccine than for the COVID-19 vaccine.
RH3: Respondents will have lower ratings of health-related attributes and higher perceptions of resource-related attributes for the COVID-19 vaccine compared to the influenza vaccine.
Objective 2: To determine the main effect(s) of respondents’ willingness to take a COVID-19 vaccine in rural contexts.
RH4: Respondents’ willingness to get a COVID-19 vaccine will be significantly correlated with health-related and resource-related attributes, but correlations with health-related attributes will be higher than with resource-related attributes.
RH5: When comprehensively considering the effects of respondents’ demographic characteristics, protective action perceptions, information sources, emergency preparedness, and emotional responses on their willingness to take a COVID-19 vaccine, vaccine attribute variables will have significant effects, but health-related attributes will have the strongest effect.
Objective 3: To explore the possibility that the effect of perceived COVID-19 vaccine attributes on COVID-19 vaccination willingness depends on respondents’ confidence in the influenza vaccine.
RQ1: Does respondents’ confidence in the influenza vaccine moderate the effects of perceived COVID-19 vaccine attributes on COVID-19 vaccination willingness?
3 Method
3.1 Study Area
As indicated in Figure 1 , the study area consists of the rural communities in three administrative districts (i.e., Yucheng District, Lushan County, and Baoxing County) in Ya’an region (which has two districts and six counties), Sichuan, China. As the Sichuan Province Statistical Yearbook revealed, the permanent residential population in the three selected administrative districts in 2020 was 516,700 with approximately 201,100 households yielding an average household size of 2.57 persons. Of these, 261,400 had been identified as rural [71].Figure 1 Location of the Ya’an region in Sichuan Province
As of December 1, 2021, only eight confirmed COVID-19 cases were reported in the Ya’an region. Similar to other rural regions in China, the Ya’an region had fewer COVID-19 cases compared to Wuhan, which accounted for 48.8% of China’s total COVID-19 cases[18]. Hence, the rural population in Ya’an region was targeted as a good representative of a rural region to understand factors affecting its population’s vaccine uptake.
3.2 Survey Design
This survey was part of a larger study conducted in China and the US, led by a team of university faculty and scholars in the two countries. The measures used in the survey questionnaire (e.g., protective action perception, information sources, emergency preparedness, and emotional response) were adopted from the PADM framework and adapted to suit the COVID-19 vaccine study. Only part of the 27-survey questionnaire is analyzed and presented here. These questions measured demographic characteristics, preparedness levels, information source reliance, emotional responses, NPI adoption, and protective action perceptions of the influenza and the COVID-19 vaccines. A 5-point Likert scale was utilized to capture the responses ranging from “not at all =1” to a “very great extent = 5” for all behavioral- and psychological-related questions. As showed in Appendix A, the questionnaire items (except variables of respondents’ and households’ contexts) had been further computed and combined into new scales as a result of factor analyses.
The survey instrument was developed in English, translated into Chinese, and then back translated to English by different researchers. The differences in translations were identified and the questionnaire was revised to ensure that the concepts were not lost in translation. Ethics approval for conducting human subjects research was obtained from both the Chinese and US universities’ ethics committees.
3.3 Sample Instruments and Data Collection
A cluster sampling method, a commonly used strategy when there is no comprehensive list from which respondents could be randomly sampled was adopted (see[70], for a general description, Frankel, 1983 [27], for theoretical details, and[55], for an example). Subsequently, six of the 23 townships from the study areas were randomly selected, which narrowed the sampling pool into an estimated 11,000-14,000 households. Given a confidence level of 95%, an acceptable sampling error of .05, and an expected response rate of 15 percent, the expected sample size was 375 with 2,500 interview attempts. In addition, a quota sampling strategy was applied when picking the respondents to ensure the case pool had sufficient diversity in terms of gender and education levels.
Face-to-face interviews were conducted during November 24-27, 2020 with one member of each household. The interviewing team comprised six Chinese university staff members and four social workers from non-profit organizations (NPOs). All the interviewers attended a one-hour orientation and training session by the lead researcher on how to conduct the surveys following ethical practices, one day before the survey. Each interviewer was assigned a street block and asked to randomly select one of the first five households in that block, conduct the interview, and then repeat the same household selection and interview procedure with the next set of five households, until the block of units was completed on that street. Interviewers then switched to the next street and stopped when all the street blocks from the study areas were covered. Informed consent was obtained from the respondents before commencing questioning. Every attempt was made to ensure that the respondents alternated between a male and female in successive households and educational attainment levels varied from low to high. Participants were given the choice of filling out the questionnaire by themselves or having their responses filled in by the interviewer. On average each interviewer made 50 to 80 attempts to recruit respondents and successfully completed 12-13 interviews per day. This procedure yielded 492 respondents, generating a response rate of roughly 20.0 percent. The goal of the equal distribution of gender and education level of the respondent was partially achieved (χ(3)2= 12.44, p < .01), with the data slightly overrepresenting females with lower educational levels by 29.1 percent compared to the expected 25%.
The summary statistics of the respondents’ demographic characteristics (see Table 1 ) indicate that the respondents tended to be older, with similar levels of education, but lower income compared to the general population of Sichuan Province. Nonetheless, the sample represented the population in Sichuan’s rural areas, despite a slight over-representation of higher educated population[71]. Specifically, 56% of the respondents were female, 80% were married, 97% self-identified as belonging to the Han ethnic group, and 96% were homeowners. The respondents had an average age of 44.2 years, 11.6 years of education, an annual average household income of 67,114 Yuan (translating to 12,093 Yuan per capita), and 5.55 members per household. compared to the national average of 2.7 members per household and GDP per capita of 18, 931 Yuan in rural areas [62].Table 1 Matrix of Means (M), Standard Deviations (SD), and Intercorrelations (rij) among Variables
Variable M SD 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
1 Age 44.18 15.77
2 Female 0.56 0.50 -.09
3 Married 0.80 0.40 .48 .01
4 Majority 0.97 0.17 .08 .03 .09
5 Household Size 5.55 2.31 .05 .01 .15 .01
6 Homeownership 0.96 0.21 .11 .05 .22 .13 .03
7 Education Years 11.60 2.67 -.50 -.02 -.24 -.14 -.07 -.08
8 Income 67.11k 56.14k -.24 -.02 -.12 .00 .04 .01 .39
9 Preparedness 3.86 2.15 -.12 -.06 .00 .00 .07 -.06 .12 .10
10 Infor_Authorities 3.83 0.90 -.08 -.04 .03 -.06 .07 .02 .09 .11 .28
11 Infor_Public 3.86 1.00 -.40 -.01 -.07 -.09 -.01 -.02 .24 .15 .21 .55
12 Infor_Private 3.79 0.93 -.04 .04 .05 -.03 .09 .05 .01 .04 .23 .68 .53
13 Social Cues 28.09 27.41 -.29 -.05 -.17 -.09 .01 -.10 .24 .12 .25 .11 .21 .14
14 Positive Emotion 2.98 1.05 -.17 -.13 -.14 -.03 -.04 -.09 .24 .12 .34 .23 .23 .15 .18
15 Negative Emotion 2.56 1.10 -.11 .01 -.04 .03 .07 -.11 .07 .01 .21 .06 .06 .09 .10 .31
16 Alert Emotion 3.17 1.21 -.10 -.10 -.06 -.07 .08 .00 .07 .06 .19 .20 .14 .18 .10 .35 .64
17 NPIs 3.95 0.80 -.29 .10 -.08 -.08 -.04 -.03 .32 .18 .33 .34 .38 .32 .25 .32 .12 .19
18 FluShot Confidence 3.78 0.71 .05 -.01 .05 .07 .02 -.05 -.01 .08 .24 .31 .22 .32 .17 .24 .17 .18 .27
19 Health 3.76 0.78 .02 -.03 -.02 .08 .00 -.02 -.01 .07 .23 .28 .18 .28 .17 .22 .15 .17 .19 .65
20 RemResource 3.85 0.77 -.09 .00 -.05 .01 .02 -.02 .06 .12 .23 .24 .25 .28 .09 .14 .16 .21 .26 .50 .53
21 Expense 3.51 0.97 -.01 .07 .02 .00 .06 .00 .02 -.04 .15 .11 .05 .09 .08 .28 .34 .31 .14 .22 .12 .29
22 Willingness 3.86 0.92 -.07 .02 .04 .05 .07 .02 .13 .15 .21 .32 .27 .31 .11 .15 .13 .18 .21 .50 .55 .60 .15
* r = .07, p < .05; r = .11, p < .01; r = .14, p < .001
Note: Age = Age; Female = Female; Married = Married; Majority = Han Ethnicity Identification; Household Size = Household Size; Homeownership = Homeownership; Education Years = Years of Education; Income = Income in $1,000 RMB; Preparedness = Level of Preparedness; Infor_Authority = Information Reliance on Authorities; Infor_Public = Information Reliance on Public Intermediate Sources; Infor_Private = Information Reliance on Private Intermediate Sources; Social Cues = Seeing People Wearing Mask; Positive Emotion = Positive Emotion; Negative Emotion = Negative Emotion; Alert Emotion = Alert Emotion; NPIs = Previous Adoptions of Non-Pharmaceutical Interventions; Flushot Confidence = Confidence in Influenza Vaccine; Health = Health-Related Attributes; RemResource = Remaining Resource-Related Attributes; Expense = Expense-Related Attributes; Willingness = Willingness to Get Vaccinated
3.4 Statistical Analyses
3.4.1 Tests for Pseudo-attitudes
Pseudo-attitudes can occur when respondents have no pre-existing beliefs, but attempt to hide their ignorance by answering questions based upon contextual cues. Hence, the responses are termed ‘pseudo’ as they fail to tap into the respondents’ stable attitudes [21], [69]. Scholars have suggested that in cases where test-retest procedures are not available, psychological studies should test pseudo-attitudes in terms of central tendency bias for those respondents who denote biased ratings by choosing the midpoint of the rating scale [16], [39] , and the diversity within subjects [5]. To test whether respondents were affected by central tendency bias, a series of t-tests were conducted, which revealed that all of the 13 vaccine-related variables have ratings significantly different from the mid-point (3) of the 1-5 rating scale, thus indicating a lack of central tendency bias. Furthermore, the results of the interrater agreement rWG tests indicated that the rWG of the 13 psychological ratings (range = .51-.67) were all significantly different from zero but smaller than rWG = .70. The results indicate that the ratings do not have the uniform distribution that would be expected if the responses were random [48], yet there is diversity in the ratings. Hence, it is reasonable to conclude that the data are not significantly affected by pseudo-attitudes.
3.4.2 Analyses
To test the hypotheses, multiple analyses were conducted. Specifically, RH1 and RH3 were examined using descriptive statistics, multivariate analysis of variance (MANOVA), and post-hoc t-tests. RH2 was tested using the Dunlap and colleagues’[25] table of statistical significance for rWG, and Levene’s test of equal variances. RH4 and RH5 were tested using correlation analysis. Additionally, a stepwise regression analysis was conducted by selecting key predictors from each category and finally a two-way ANOVA.
The statistical tests mentioned above generated a total number of 356 p-values in this study, including 61 statistics from mean or variance comparison tests, 256 statistics from correlation analyses, and 39 statistics from regression analyses. To eliminate the concern of an experiment-wise error rate[67]), this study adopted the Benjamini and Hochberg (B&H) approach to determine the critical p-value and its associated false rate[9], [32]. The B&H approach starts with the specification of a false discovery rate (d) for the entire study, followed by sorting the pi significance values for the individual tests in ascending order 1 ≤ i ≤ n, and classifying each pi ≤ d x i/n as statistically significant. These results suggested a critical value of pi = .033 with a false discovery rate of .05. The study rounded up the critical value to p = .05 which yielded a false discovery rate of .073 with the expected number of false-positive test results FP (= α x n) = 26 (=356 x .073).
4 Results
4.1 Perceptions on Vaccine Attributes (Tests of RH1-RH3)
RH1 (Respondents’ vaccine ratings will be higher on health-related attributes than on resource-related attributes) is only partially supported. Results of the MANOVA reveal significant effects for attributes (Wilks Λ = 0.79, F 5,487 = 25.73, p < .001), for within-subjects (F 4,1811 = 27.96, p < .001), and for between-subjects (F 1,491 = 17,606.01, p < .001). As indicated in Figure 2 , one of the resource-related attributes, requiring additional knowledge, surprisingly received ratings (M = 4.01) that were significantly higher than any other attributes. Of the three health-related attributes, defeating disease and improving immunity received the next highest ratings (M = 3.87 and 3.72, respectively). However, only the ratings for defeating disease were significantly different from the rest of the attributes, while the ratings for improving immunity were generally similar to the ratings of availability concerns (M = 3.70) and safety concerns (M = 3.68). Finally, the remaining resource-related attribute, expense concerns, received ratings (M = 3.51) that were significantly lower than all attributes other than safety concerns (Mdiff = 0.17, p < .05).Figure 2 Perceived attributes of COVID-19 vaccine and influenza vaccine, * “Will be available” was not appliable for the influenza vaccine
In response to RH2 (Respondents’ ratings on health-related and resource-related attributes will yield higher levels of agreement for the influenza vaccine than for the COVID-19 vaccine), interrater agreement values for the influenza vaccine attributes were generally similar to those for the COVID-19 vaccine. Specifically, the range of interrater agreement values for the health-related attributes ranged .59-.66 (r¯WG = .62) for the COVID-19 vaccine, while rWG for the influenza vaccine ranged .61-.67 (r¯WG = .64). Similarly, the range of rWG values for the resource-related attributes was .53-.62 (r¯WG = .57) on the COVID-19 vaccine and .51-.53 (r¯WG = .52) on the influenza vaccine. Of these, only the interrater agreement of safety concerns for the influenza vaccine was significantly larger than for the COVID-19 vaccine (F = 10.97, p < .001). On the other hand, the interrater agreement of requiring additional knowledge revealed a reverse tendency; respondents’ ratings on the COVID-19 vaccine were more similar than for the influenza vaccine (F = 5.77, p < .05).
Partially consistent with RH3 (Respondents will have lower perceptions of health-related attributes and higher perceptions of the resource-related attributes for the COVID-19 vaccine compared to the influenza vaccine), the results of the MANOVA revealed significant effects for vaccines (Wilks Λ = 0.97, F 1,491 = 17.62, p < .001), attributes (Wilks Λ = 0.71, F 4,488 = 50.95, p < .001), and interaction (Wilks Λ = 0.89, F 4,488 = 15.11, p < .001). However, the results of the post-hoc pairwise comparisons supported RH3 on the resource-related attributes, but not on the health-related attributes. As indicated in Figure 2, two of the health-related attributes (defeating disease and improving immunity) had nonsignificant differences on the ratings between the COVID-19 and influenza vaccines (t491 = 1.85, 0.34, ns., respectively). As expected, respondents’ perceptions of the COVID-19 vaccine’s safety was slightly lower (M = 3.68) than their confidence on the influenza vaccine’s safety (M = 3.77; t491 = -2.40, p < .05), but the difference was only 2.3% of the response scale. Conversely, concerns of requiring additional knowledge and expense for COVID-19 vaccine were significantly higher than concerns for influenza vaccine (t491 = 3.27, 7.64, p < .001), respectively. However, these differences were also relatively small at, 0.1% and 6.1% of the response scale respectively.
4.2 Main Effect Analyses on Vaccination Intention (Tests of RH4-RH5)
To reduce collinearity effects, two factor analyses were conducted on the vaccine attributes before assessing the effects of those attributes on vaccination willingness. As expected, the factor analysis of the COVID-19 vaccine attributes suggested a two-factor solution in which defeating disease, improving immunity, and safety concerns were assigned to one scale—Health (α = .87, r¯ = .70), while requiring additional knowledge and availability concerns were assigned to another scale—Rem-Resource (remaining resource-related attribute; α = .62, r¯ = .45). However, Expense (expense concerns) was left as an independent variable because it did not load strongly on either factor. The factor analysis of the influenza vaccine attributes yielded a similar result. Hence, defeating disease, improving immunity, and safety concerns of the influenza vaccine were assigned to the scale Flu shot confidence (α = .84, r¯ = .70).
Table 1 displays the means, standard deviations, and intercorrelations among the variables needed to test RH4 (Respondents' willingness to get a COVID-19 vaccine will be significantly correlated with health-related and resource-related attributes, but correlations with health-related attributes will be higher than with resource-related attributes). Consistent with the hypothesis, Health (r = .55) and Rem-Resource (r = .60) have strong correlations with willingness while, despite being much lower, expense (r = .15) also has a significant correlation with COVID-19 vaccination willingness. However, although the effects of Health and Rem-Resource on COVID-19 vaccination willingness were generally similar, Rem-Resource had a slightly stronger correlation than Health. Hence, RH4 was partially supported.
Although not hypothesized, Flu shot confidence (r = .50) also had a strong correlation with COVID-19 vaccination willingness, while NPIs (r = .21), information sources (r= .30), and emergency preparedness (r = .21) also had moderate correlations with COVID-19 vaccination willingness. Besides, age (r = -.07), household size (r = .07), education year (r = .13), income (r = .15), social cues (r = .11), and emotional response (r¯ = .16) also had some weak but significant correlations with COVID-19 vaccination willingness. It is also noteworthy that Health (r = .65) and Rem-Resource (r = .50) were strongly correlated to Flu shot confidence, which suggests that respondents’ perceptions of the COVID-19 vaccine might be based on their experience with the influenza vaccine.
As indicated in Table 2 , the test of RH5 (When comprehensively considering the effects of respondents’ demographic characteristics, protective action perceptions, information sources, emergency preparedness, and emotional responses on their willingness to take a COVID-19 vaccine, vaccine attribute variables will have significant effects, but health-related attributes will have the strongest effect) partially supported the PADM by examining variables in five blocks and adding each block variables stepwise. As expected, contextual, information source, emotional variables—income, emergency preparedness, Info_authorities, and Info_private—have some significant effects in the beginning steps (i.e., Models I, II, and III). However, when previous experience (e.g., NPIs, Flu shot confidence) and emotional response are entered in Model IV, only income (β = .09, p < .05) and Flu shot confidence (β = .54, p < .001) remain significant. When COVID-19 vaccination attributes are added in Model, Income turns nonsignificant (β = .07, ns.) while the effect of Flu shot confidence decreases (β = .11, p < .05). Consistent with RH5, Health and Rem-Resource have significant effects on COVID-19 vaccination willingness. However, Rem-Resource surprisingly has a stronger effect (β = .38, p < .001) than Health (β = .24, p < .001).Table 2 Regression of COVID-19 Vaccine Willingness on Predictor Variables
DV = Willingness Model I Model II Model III Model IV Model V
Predictors B SE β B SE β B SE β B SE β B SE β
Respondents’ and Households’ Contexts
Age .00 .00 -.03
Female .05 .08 .03
Married .17 .12 .07
Majority .28 .24 .05
Household Size .02 .02 .05
Homeownership .07 .20 .02
Education Years .03 .02 .08
Income .00* .00 .10* .00* .00 .10* .00* .00 .11* .00* .00 .09* .00 .00 .07
Preparedness .08*** .02 .19*** .05* .02 .11* .04* .02 .10* .02 .02 .05 .00 .02 .00
Information Sources
Infor_Authorities .13* .06 .13* .15* .06 .14* .09 .05 .09 .09 .05 .09
Infor_Public .07 .05 .07
Infor_Private .15** .06 .16** .17** .06 .17** .10 .05 .10 .04 .05 .04
Social Cues .00 .00 .02
Emotional Variables
Positive Emotion .01 .04 .01
Negative Emotion .04 .05 .04
Alert Emotion .05 .04 .07 .04 .03 .06 .02 .03 .02
Experiences
NPIs .00 .05 .00
FluShot Confidence .54*** .05 .41*** .15* .06 .11*
Vaccine Attributes
Health .28*** .05 .24***
RemResource .44*** .05 .38***
Expense -.03 .03 -.02
(Constant) 2.57 .44 2.22 .19 2.11 .20 0.82 .24 0.02 .21
R2 .08 .15 .15 .30 .46
Adj R2 .06 .14 .14 .29 .45
df (9,482) (6,485) (7,484) (7,484) (11,480)
F 4.46*** 14.20*** 12.71*** 29.57*** 37.89***
* p < .05; ** p < .01; *** p < .001
Note: Age = Age; Female = Female; Married = Married; Majority = Han Ethnicity Identification; Household Size = Household Size; Homeownership = Homeownership; Education Years = Years of Education; Income = Income in $1,000 RMB; Preparedness = Level of Preparedness; Infor_Authority = Information Reliance on Authorities; Infor_Public = Information Reliance on Public Intermediate Sources; Infor_Private = Information Reliance on Private Intermediate Sources; Social Cues = Seeing People Wearing Mask; Positive Emotion = Positive Emotion; Negative Emotion = Negative Emotion; Alert Emotion = Alert Emotion; NPIs = Previous Adoptions of Non-Pharmaceutical Interventions; Flushot Confidence = Confidence in Influenza Vaccine; Health = Health-Related Attributes; RemResource = Remaining Resource-Related Attributes; Expense = Expense-Related Attributes; Willingness = Willingness to Get Vaccinated
4.3 Influenza Vaccine Effects on COVID-19 Vaccination Intention
RQ1 (Does respondents’ confidence in the influenza vaccine moderate the effects of perceived COVID-19 vaccine attributes on COVID-19 vaccination willingness?) was tested by one MANOVA that assessed the interaction effects of Flu shot confidence and Health on COVID-19 vaccination willingness and another MANOVA that assessed the interaction effects of Flu shot confidence and Rem-Resource on COVID-19 vaccination willingness. As indicated in Table 3 , both tests produced significant interactions (F69 = 1.65, p < .01; F55 = 1.59, p < .01, respectively). Specifically, Figure 3 shows that both Health and Rem-Resource have moderately strong positive effects when respondents’ Flu shot confidence is higher than the moderate level but negligible effects when Flu shot confidence is lower than that.Table 3 Tests of Interaction Effects
DV = COVID-19 Vaccination Willingness Model I Model II
Predictors df Mean Square F df Mean Square F
Corrected Model 97 2.35 5.06*** 81 2.86 6.52***
Intercept 1 90.55 194.79*** 1 97.65 222.78***
Covariance Variables
Income 1 3.14 6.76** 1 1.12 2.55
Preparedness 1 0.42 0.90 1 0.62 1.42
Infor_Authorities 1 0.25 0.54 1 1.05 2.38
Infor_Private 1 0.78 1.69 1 0.14 0.32
Alert Emotion 1 0.86 1.85 1 0.04 0.08
Fixed Factors
FluShot Confidence 13 1.16 2.50** 13 1.34 3.07***
Health 10 3.24 6.98***
RemResource 8 4.59 10.48***
FluShot Confidence* Health 69 0.77 1.65**
FluShot Confidence* RemResource 55 0.70 1.59**
Error 394 0.46 410 0.44
Total 492 492
Corrected Total 491 491
R2 .56 .56
Adj R2 .46 .48
Figure 3 Interactions of influenza vaccination confidence with COVID-19 vaccination health-related and resource-related attributes on COVID-19 vaccination willingness, Note: Income, Preparedness, Infor_Authorities, Infor_Private, and Alert emotion are controlled as covariates.
5 Discussion
This study investigated rural households’ perceptions of the already-known influenza vaccine and the unknown COVID-19 vaccine and also explored factors affecting people’s willingness to get the COVID-19 vaccine. While other scholars have studied the general Chinese population’s COVID-19 vaccine willingness, they relied on online surveys for data collection, given the movement restrictions due to pandemic containment protocols (M. [20], [50] ; Y. [85]. Similar survey methods were also used by researchers in the US [45], Australia [11], and Qatar [3] or an extensive bibliographic review in India (Raman et al. 2021), as COVID-19 containment measures prevented face-to-face interviews. Moreover, the respondents of these prior studies were younger, with higher education, and higher income than their national populations [45], [50] , possibly producing variance restriction in the predictor variables that would cause downward bias in the correlation and regression coefficients [65]. This study used face-to-face interviews with a mix of young and elderly, literate and illiterate, male and female members in rural populations adding to its uniqueness.
This study’s findings suggest that rural populations require additional knowledge for both types of vaccines, with more of them requiring additional information about the COVID-19 vaccine. Data on the Expanded Program of Immunization (EPI) in China suggests that vaccine coverage for children has been consistently over 90%, but the coverage of non-EPI vaccines such as influenza vaccines for adults is much lower [23]. For instance, the coverage of influenza vaccine achieved no more than 0.4% in 2019 in China, which was much lower than that in the US (48.4%) (Y. [85]. In addition, there was a higher mortality burden of influenza in the west of China than in the center of China, due to a low influenza vaccination coverage and a shortage of access to health-care services, especially in rural areas [51]. One reason for the influenza vaccine’s low coverage may be people’s underestimates of the vaccine’s health-related attributes and overestimates of its resource-related attributes. During the face-to-face survey, some of the elderly respondents expressed beliefs that if they took the influenza vaccine once, they had beaten the influenza virus forever. Additionally, people in rural China who live a poor life with hardships do not believe the common cold/influenza as something to worry about. Oftentimes the illiterate tend to consider the influenza virus the same as a common cold and simply follow traditional Chinese medicinal practices, drinking more boiled hot water with ginger and brown sugar as remedies. During the time this study was conducted in November 2020, information on vaccine research and development, clinical trials, and manufacturing were being released extensively through multiple media channels like the TV, and social media including TikTok videos (M. [20]; J. [84]. However, most of the information sources were difficult for rural populations to access and the content was difficult to understand.
The fact that amongst the health-related attributes, the safety of influenza vaccine received higher ratings than the COVID-19 vaccine was not surprising. People are usually much more cautious about any new pharmaceutical in the market, especially rural populations who are unfamiliar with new innovations. For instance, during the early stages of the COVID-19 pandemic, due to the fear of the virus, people in rural regions destroyed main roads leading to their communities in order to prevent outsiders from entering their village [80]. Given the big Chinese pharma companies’ vaccine scandals in the past decade, it is understandable that these rural populations were uncertain about the new vaccine’s safety.
The lack of knowledge surrounding the COVID-19 vaccine arose mainly from doctors’, pharmaceutical research and development agencies’, and manufacturers’ very abstract discourse during the pandemic’s initial days about the coronavirus and mitigation measures. These perspectives could be quite different from those of the majority of the population in the Ya’an region, who have little or no education. Due to the limited understanding of COVID-19 vaccines, rural people believed that only the inactivated COVID-19 vaccine is the best for preventing the virus [50]. Moreover concerns about the expense related to this new vaccine was high amongst rural households, whose annual income is very low compared to their urban counterparts. During the survey, rural respondents expressed concerns of affordability because they had heard that the cost of one dose of the COVID-19 vaccine would be higher than that of the influenza vaccine (100-150 RMB), due to limited manufacturing and distribution networks to rural areas.
The strong influence of flu shot confidence on COVID-19 vaccination intentions might be due to the perception of similar infection mechanisms on respiratory system between influenza and COVID-19 [33]; WHO, 2021). Previous studies have indicated that people would be more likely to use an availability heuristic—responding to an emerging threat by using a familiar protective action to a similar previous threat [35], [75] . Hence, given that the influenza vaccine is known to protect individuals from the threats of fever, cough, and throat pain [94], it is not surprising that respondents imagined and linked COVID-19 vaccine with influenza vaccine as they share some similar respiratory symptoms (WHO, 2021). Another reason might be an overconfidence in the influenza vaccine [33], as its production has not been fraught with scandal. Hence, despite the combined effects of low education, low income, and poor accessibility to health care, there could be a strong effect of flu shot experience/perceptions on COVID-19 vaccination intentions. However, further research is needed to assess the reasons for this finding.
Interestingly, perceptions of COVID-19 vaccine health-related and resource-related attributes have the strongest influence on COVID-19 vaccination willingness. These results indicate that it is essential for people to believe that the COVID-19 vaccine will be effective and safe, as well as that information about it and access to it will be readily available. During times of uncertainty, due to lack of understandable information from credible sources, individuals often engage in health information seeking before deciding whether to comply with protective action recommendations [77], [89] . Also, at the time of the survey, there were divergent views and reports about COVID-19 vaccines, such as that COVID-19 vaccine production would be very limited and, therefore, rural people would have to wait for a long time for vaccinations [26]; J. [84]. Additionally, the National Health Commission of China (2022) announced that the distribution of COVID-19 vaccine would be given first to people working for medical services, import-export industries, and domestic and international shipping industries. This message could have increased rural residents’ doubts about their timely access to COVID-19 vaccination.
6 Conclusions
6.1 Theoretical Implications
Compared to previous COVID-19 studies on vaccine uptake, the current study makes a strong contribution by contributing data collected through face-to-face interviews with rural populations before the vaccinations were started. This provides a broader knowledge on how elderly, lower educated, and lower income population segments think about COVID-19 vaccinations. Findings on the significant associated factors with vaccination intentions such as vaccine knowledge and information source, will allow public health authorities to provide effective risk communication that encourages protective health behavior changes in future outbreaks. Specifically, by knowing the effective risk communication channels, societies could avoid the so called “Andrew Wakefield case” that misled people, especially those with lower education and income, to reject vaccines (Motta & Stecula, 2021). Furthermore, because our case study could roughly represent the rural regions in China from the perspective of COVID-19 control and infection, our findings on COVID-19 vaccination intention have strong empirical contributions to identify the general considerations about the COVID-19 vaccine in these places. This is important because at the time of our survey, nearly no information existed and little attention has been paid to rural regions in regard to COVID-19 vaccine in such a country with a large rural population.
The current study applied the PADM framework and related measures to investigate the factors influencing rural Chinese residents’ COVID-19 vaccination intentions. By examining people’s perceptions of the health-related and resource-related attributes of influenza and COVID-19 vaccines, this study utilized some new survey scales to measure these two types of attributes in a pandemic context. Although many previous studies have analyzed the effectiveness of protective action attributes of natural disasters and technological disasters [37], the current study extends previous PADM research on respiratory infectious diseases by F. Wang et al. [83] and Wei et al. (2018) to the COVID-19 pandemic context. Furthermore, based on the data analysis results, the current study statistically categorized the specific items for the health-related and resource-related attributes of COVID-19 vaccines. This method went beyond the simple categorical method of assigning certain items for attributes. For instance, these data show that the expense of COVID-19 vaccine, which the PADM categorizes theoretically as a resource-related attribute, is different from the other resource-related attributes and—contrary to economic theory—to have a low correlation with vaccination intention.
6.2 Practical Implications
The current study has two major practical implications: one at the individual level and the other at the policy level. Getting vaccinated is considered to be the strongest mitigation tool to combat the COVID-19 pandemic. However, many previous studies have reported that overconfidence, inconvenience, and complacency cause vaccine hesitancy [8], [23] . As the vaccine expense was unclear at the time of the survey, one might presume that low-income populations would be hesitant about getting vaccinated but these data show that expense was not a major obstacle to vaccination intention. Instead, the influenza vaccine attributes show strong effect on COVID-19 vaccination intention, which suggests that these respondents tend to extrapolate from their former vaccination experience to the current pandemic. Specifically, those who had positive perceptions of the influenza vaccine also had more positive perceptions of a COVID-19 vaccine. This indicates that, to increase vaccination uptake, local health agencies should consider the geographical and social events that had influenced the local population.
Further, to improve food, nutrition, and livelihood security, these rural communities may be encouraged to grow fruits and vegetables in kitchen gardens as a cheap source of healthy nutritious foods (Suresh et al. 2022) which in turn can make them resilient to future threats.
At the policy level, convenient access to recommended protective actions could affect people’s behavioral intentions. For the COVID-19 vaccine, access to vaccination clinics and the timely availability of the vaccine would ordinarily expected to be rural residents’ two biggest concerns. Nevertheless, the Ya’an data reveal that the COVID-19 vaccine’s resource-related attributes had just as strong an impact on vaccination intention as its health-related attributes. These results indicate that administrative agencies should emphasize new vaccines’ resource-related and health-related attributes through a diverse set of information channels. In addition, administrative agencies in rural regions should consider people’s literacy levels, especially those who cannot obtain information from sources other than TV and peers. Although people’s beliefs about COVID-19 and vaccinations have changed in the Ya’an region since this survey, this does not limit the significant contribution of our results. Because our findings focus on the internal mechanisms of risk communication and coping behaviors during the COVID-19 pandemic, they will be useful for local authorities in managing responses to a future public health emergencies.
One apparent limitation of the current study is the potential for bias due to the unavailability of a random sampling strategy. However, it is important to understand that the effect of geographic/demographic bias depends upon what statistic is being assessed (means and proportions vs. correlation and regression coefficients) and the relationships between demographic characteristics and dependent variables of interest (psychological variables and behaviors). First, estimates of means and proportions are biased by geographic/demographic non-representativeness only to the degree that the relevant dependent variables are correlated with those geographic/demographic variables. However, the literatures on hurricane evacuation [7], [36]and household emergency preparedness (Lindell, 2013; Lindell & Perry, 2000) indicate that the correlations of demographic variables with behavioral variables are small and inconsistent. Consequently, geographic/demographic non-representativeness appears to have little effect on estimates of mean level of behavioral variables (or the proportion of people engaging in a behavior). Even more important is that geographic/demographic non-representativeness will have little effect on correlation and regression coefficients unless it is so severe that there are “ceiling” or “floor” effects that cause these coefficients to be systematically underestimated (Lindell & Perry, 2000;[65]. Since Table 1 shows that the cluster sampling strategy provided a reasonable level of diversity in the sample’s demographic characteristics, it is unlikely that the correlation and regression coefficients were underestimated to any significant degree. Nonetheless, the study was conducted in only one region, so further studies should be conducted in rural communities in other provinces, regions, or even countries.
A more significant limitation is that, the data were only collected once, and due to the vaccination policy changes, could not be repeated. Thus, the cross-sectional data cannot conclusively determine if the factors influencing vaccination intentions determined actual vaccinations when these became available. Following the recommendations of Bubeck et al. [13], Hudson et al. [38] and Siegrist [72], [73], longitudinal cohort studies should be conducted in these communities to determine if their perceptions and decisions about the vaccines have changed and, if so, why. Such studies could identify needed revisions to the survey items for explicating the factors affecting rural people’s vaccination intentions and actual vaccination adoption. Alternative, novel methods such as remote participatory research along with systems thinking design may be envisaged for the future (Bonin et al. 2021).
Ethics approval
Ethics approval for conducting Human Subjects research was obtained from both the US University’s Institutional Review Board (Jacksonville State University), and the Chinese university’s ethics committee (Sichuan University).
Submission declaration
The current manuscript has not been published previous, that it is not under consideration for publication elsewhere, that its publication is approved by all authors and tacitly or explicitly by the responsible authorities where the work has carried out, and that, if accepted, it will not be published elsewhere in the same form, in English or in any other language, including electronically without the written consent of the copyright-holder.
Funding
National Natural Science Foundation of China (No. 72204253) and Sichuan Federation of Social Science Associations, China (No. SC20C012)
Uncited references
Liu and Yang [56], Commission [63], Qin et al. [68], World Health Organization [87 ], Zhang et al. [92].
CRediT authorship contribution statement
Yingying Sun: Conceptualization, Project administration, Investigation. Shih-Kai Huang: Conceptualization, Methodology. Sudha Arlikatti: Conceptualization, Methodology. Michael K. Lindell: Conceptualization.
Declaration of Competing Interest
The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: [Yingying Sun reports financial support was provided by Sichuan Federation of Social Science Associations, China.].
Appendix A . Measurements and Calculations of Survey Items
Item Mean SD
COVID-19 vaccination
Willingness
I will take a COVID-19 vaccination. 3.86 0.92
Health α = .87, r¯ = .70
A COVID-19 vaccination protects against infection. 3.87 0.83
A COVID-19 vaccination improves body immunity. 3.72 0.90
A COVID-19 vaccination is safe. 3.68 0.90
RemResource α = .62, r¯ = .45
I need more information about a COVID-19 vaccination. 4.01 0.94
I can get a COVID-19 vaccination within a year. 3.70 0.87
Expense
A COVID-19 vaccination may be expensive. 3.51 0.97
Influenza vaccination
FluShot Confidence α = .84, r¯ = .57
Vaccination protects against infection. 3.80 0.85
Vaccination improves body immunity. 3.71 0.88
Vaccination is safe. 3.77 0.81
I will take an influenza vaccination 3.83 0.90
Information seeking:To what extent have you accessed to information sources about the COVID-19 and protective actions?
Infor_Authorities α = .93, r¯ = .63
Public health officials 3.74 1.16
Local government (i.e., town, county, district) 3.82 1.01
State government (i.e., city and province) 3.69 1.05
National government 4.07 1.04
Infor_Public α = .96, r¯ = .54
Traditional media (i.e., TV, radio, newspapers) 4.14 0.95
Web-based media (i.e., mobile news, toutiao) 3.78 1.30
Social media (i.e., Weibo, Wechat, TikTok) 3.67 1.32
Infor_Private α = .83, r¯ = .55
Community broadcasting (i.e., village, neighborhood, apartment) 3.83 1.05
Personal chat (i.e., relative, friend, colleague) 3.75 1.06
In the past week, what percentage of people you saw were wearing masks on the street?
Social Cues 28.09 27.41
Emotional responseHow much does the possibility of being infected with COVID-19 make you feel…
Positive Emotion α = .73, r¯ = .58
Optimistic 2.87 1.24
Energetic 3.09 1.12
Negative Emotion α = .94, r¯ = .73
Depressed 2.45 1.22
Angry 2.35 1.27
Nervous 2.86 1.23
Annoyed 2.59 1.22
Fearful 2.66 1.28
Irritated 2.48 1.25
Alert Emotion
Alert 3.17 1.21
NPIs performanceTo what extent have you taken each of the following actions to protect yourself and your loved ones from COVID-19?
NPIs α = .86, r¯ = .46
Reducing outings 4.29 1.01
Social distancing 4.19 0.98
Sanitizing surfaces 4.07 1.02
Hand hygiene 4.33 0.81
Disinfecting hands 4.03 1.08
Taking body temperature 3.48 1.38
Reducng use of public transportation 4.15 1.06
Wearing goggles or disposable gloves 3.06 1.52
Disaster preparednessDid your household have any of the following items before the pandemic?
Preparedness α = .78, r¯ = .33
A working portable radio with spare batteries 0.21 0.41
At least 4 gallons of water per person in plastic containers 0.48 0.50
A two-week supply of non-perishable food for yourself and your family 0.51 0.50
At least a one-week supply of prescription medicines 0.53 0.50
A flashlight with extra batteries 0.74 0.44
Mobile phone charger 0.69 0.46
Face masks 0.70 0.46
Data availability
Data will be made available on request.
==== Refs
References
1 Abedin M. Islam M.A. Rahman F.N. Reza H.M. Hossain M.Z. Hossain M.A. Willingness to vaccinate against COVID-19 among Bangladeshi adults: Understanding the strategies to optimize vaccination coverage PloS One 16 4 2021 e0250495 33905442
2 Adhikari B. Cheah P.Y. Vaccine Hesitancy in the COVID-19 Era The Lancet Infectious Diseases 21 8 2021 1086 34217431
3 Alabdulla M. Reagu S.M. Al‐Khal A. Elzain M. Jones R.M. COVID-19 vaccine hesitancy and attitudes in Qatar: A national cross-sectional survey of a migrant-majority population Influenza and Other Respiratory Viruses 15 3 2021 361 370 33605010
4 Andrew S.A. Arlikatti S. Chatterjee V. Ismayilov O. Ebola crisis response in the USA: Communication management and SOPs International Journal of Disaster Risk Reduction 31 2018 243 250 10.1016/j.ijdrr.2018.04.028 32289013
5 Arlikatti S. Lindell M.K. Prater C.S. Zhang Y. Risk Area Accuracy and Hurricane Evacuation Expectations of Coastal Residents Environment and Behavior 38 2 2006 226 247
6 Arlikatti S. Taibah H.A. Maghelal P. Andrew S.A. Alkhurayyif S.A. Hajj pilgrims’ perceptions of trust and Internet use for (emergency) information International Journal of Religious Tourism and Pilgrimage 10 1 2022
7 Baker E.J. Hurricane evacuation behavior International Journal of Mass Emergencies and Disasters 9 2 1991 287 310
8 Bedford H. Attwell K. Danchin M. Marshall H. Corben P. Leask J. Vaccine hesitancy, refusal and access barriers: The need for clarity in terminology Vaccine 36 44 2018 6556 6558 10.1016/j.vaccine.2017.08.004 28830694
9 Benjamini Y. Hochberg Y. Controlling the false discovery rate-A practical and powerful approach to multiple testing Journal of the Royal Statistical Society Series B-Statistical Methodology 57 1 1995 289 300 10.1111/j.2517-6161.1995.tb02031.x
10 Bergstrom C.T. West J.D. Calling Bullshit: The Art of Skepticism in a Data-Driven World 2021 Randon House New York, NY
11 Borriello A. Master D. Pellegrini A. Rose J.M. Preferences for a COVID-19 vaccine in Australia Vaccine 39 3 2021 473 479 10.1016/j.vaccine.2020.12.032 33358265
12 Brandolini A. (2014). The bullshit asymmetry principle. Lecture delivered at XP2014 in Roma and at ALE2014 in Krakow 2014.
13 Bubeck P. Botzen W.J.W. Aerts J.C.J.H. A review of risk perceptions and other factors that influence flood mitigation behavior Risk Analysis 32 9 2012 1481 1495 10.1111/j.1539-6924.2011.01783.x 22394258
14 Cao L. Zheng J.S. Cao L.S. Cui J. Xiao Q.Y. Evaluation of the impact of Shandong illegal vaccine sales incident on immunizations in China Human Vaccines & Immunotherapeutics 14 7 2018 1672 1678 10.1080/21645515.2018.1473697 29771622
15 Cao W.J. Fang Z.W. Hou G.Q. Han M. Xu X.R. Dong J.X. The psychological impact of the COVID-19 epidemic on college students in China Psychiatry Research 287 2020 5 10.1016/j.psychres.2020.112934
16 Cascio W.F. Aguinis H. Applied psychology in human resource management (6th ed.). 2004 Pearson College Div
17 Caserotti M. Girardi P. Rubaltelli E. Tasso A. Lotto L. Gavaruzzi T. Associations of COVID-19 risk perception with vaccine hesitancy over time for Italian residents Social Science & Medicine 272 2021 113688 10.1016/j.socscimed.2021.113688
18 CDC of China. (2021). The latest situation of the COVID-19 pandemic. (Accessed January 1, 2022) https://www.chinacdc.cn/jkzt/crb/zl/szkb_11803/jszl_11809/
19 Chen B. Zhang J.M. Jiang Z.G. Shao J. Jiang T. Wang Z.T. Media and public reactions toward vaccination during the “hepatitis B vaccine crisis” in China Vaccine 33 15 2015 1780 1785 10.1016/j.vaccine.2015.02.046 25731787
20 Chen M. Li Y. Chen J. Wen Z. Feng F. Zou H. An online survey of the attitude and willingness of Chinese adults to receive COVID-19 vaccination Human Vaccines & Immunotherapeutics 17 7 2021 2279 2288 33522405
21 Converse P.J. Attitudes and non-attitudes: Continuation of a dialogue Tufte E.R. The quantitative analysis of social problems 1970 Addison-Wesley 168 189
22 Dagan N. Barda N. Kepten E. Miron O. Perchik S. Katz M.A. BNT162b2 mRNA Covid-19 vaccine in a nationwide mass vaccination setting New England Journal of Medicine 384 15 2021 1412 1423 10.1056/NEJMoa2101765 33626250
23 Du F.X. Chantler T. Francis M.R. Sun F.Y. Zhang X. Han K.Y. The determinants of vaccine hesitancy in China: A cross-sectional study following the Changchun Changsheng vaccine incident Vaccine 38 47 2020 7464 7471 10.1016/j.vaccine.2020.09.075 33041097
24 Dube E. Laberge C. Guay M. Bramadat P. Roy R. Bettinger J. Vaccine hesitancy An overview Human Vaccines & Immunotherapeutics 9 8 2013 1763 1773 10.4161/hv.24657 23584253
25 Dunlap W. Burke M. Smith-Crowe K. Accurate tests of statistical significance for r(WG) and average deviation interrater agreement indexes Journal of Applied Psychology 88 2 2003 356 362 10.1037/0021-9010.88.2.356 12731720
26 Feng J. Li J. Hu W. Li G. Public interest, risk, trust, and personal protective equipment purchase and usage: Face masks amid the COVID-19 pandemic International Journal of Environmental Research and Public Health 19 9 2022 10.3390/ijerph19095502
27 Frankel M. Sampling theory Rossi P.H. Wright J.D. Anderson A.B. Handbook of Survey Research 1983 Academic Press San Diego 21 67
28 Fridman I. Lucas N. Henke D. Zigler C.K. Association between public knowledge about COVID-19, trust in information sources, and adherence to social distancing: Cross-sectional survey JMIR Public Health and Surveillance 6 3 2020 e22060 32930670
29 Fu W.N. Yan S.J. Zong Q. Anderson-Luxford D. Song X.Y. Lv Z.Y. Mental health of college students during the COVID-19 epidemic in China Journal of Affective Disorders 280 2021 7 10 10.1016/j.jad.2020.11.032
30 Fullman N. Yearwood J. Abay S.M. Abbafati C. Abd-Allah F. Abdela J. Measuring performance on the Healthcare Access and Quality Index for 195 countries and territories and selected subnational locations: A systematic analysis from the Global Burden of Disease Study 2016 Lancet 391 10136 2018 2236 2271 29893224
31 Gilboa M. Regev-Yochay G. Mandelboim M. Indenbaum V. Asraf K. Fluss R. Durability of immune response after COVID-19 booster vaccination and association With COVID-19 Omicron infection JAMA Network Open 5 9 2022 e2231778 36107426
32 Glickman C.D. Gordon S.P. Ross-Gordon J. Supervision and instructional leadership: A developmental approach (10th ed.). 2017 Pearson
33 Gong L. Zhang X. Qu Z. Francis M. Han K. Xu C. Public interest in distribution and determinants of influenza and pneumonia vaccination during the COVID-19 pandemic: An infodemiology and cross-sectional study from China Vaccines 9 11 2021 10.3390/vaccines9111329
34 Hogan C. Atta M. Anderson P. Stead T. Solomon M. Banerjee P. Knowledge and attitudes of us adults regarding COVID-19. International Journal of Emergency Medicine 13 1 2020 10.1186/s12245-020-00309-6
35 Hua C. Huang S.-K. Lindell M.K. Yu C.-H. Rural Households’ Perceptions and Behavior Expectations in Response to Seismic Hazard in Sichuan China. Safety Science. 125 2020 104622
36 Huang S.K. Lindell M.K. Prater C.S. Who leaves and who stays? A review and statistical meta-analysis of hurricane evacuation studies Environment and Behavior 48 8 2016 991 1029
37 Huang S.-K. Lindell M.K. Prater C.S. Multistage model of hurricane evacuation decision: Empirical study of Hurricanes Katrina and Rita Natural Hazards Review 18 3 2017 15 10.1061/(asce)nh.1527-6996.0000237
38 Hudson P. Thieken A.H. Bubeck P. The challenges of longitudinal surveys in the flood risk domain Journal of Risk Research 23 4 2019 1 22
39 Hyman A. Arlikatti S. Huang S.-K. Lindell M.K. Mumpower J. Prater C.S. How do perceptions of risk communicator attributes affect emergency response? An examination of a water contamination emergency in Boston, USA Water Resources Research 58 1 2022 10.1029/2021WR030669
40 Kasperson R.E. Renn O. Slovic P. Brown H.S. Emel J. Goble R. The Social Amplification of Risk: A Conceptual Framework Risk Analysis 8 1988 177 187
41 Keech C. Albert G. Cho I. Robertson A. Reed P. Neal S. Phase 1–2 trial of a SARS-CoV-2 recombinant spike protein nanoparticle vaccine New England Journal of Medicine 383 24 2020 2320 2332 32877576
42 Kennedy J. Vaccine Hesitancy: A Growing Concern Pediatr Drugs 22 2020 105 111
43 Kimhi S. Marciano H. Eshel Y. Adini B. Resilience and demographic characteristics predicting distress during the COVID-19 crisis Social Science & Medicine 265 2020 113389 33039732
44 Koslap-Petraco M. Vaccine Hesitancy: Not a New phenomenon, but a New Threat Journal of the American Association of Nurse Practitioners 31 11 2019 624 626 31688502
45 Kreps S. Prasad S. Brownstein J.S. Hswen Y. Garibaldi B.T. Zhang B. Factors associated with US adults’ likelihood of accepting COVID-19 vaccination JAMA Netw Open 3 10 2020 e2025594 33079199
46 Lahijani A.Y. King A.R. Gullatte M.M. Hennink M. Bednarczyk R.A. HPV vaccine promotion: The church as an agent of change Social Science & Medicine 268 2021 8 10.1016/j.socscimed.2020.113375
47 Latkin C.A. Dayton L. Yi G. Konstantopoulos A. Boodram B. Trust in a COVID-19 vaccine in the US: A social-ecological perspective Social Science & Medicine 270 2021 8 10.1016/j.socscimed.2021.113684
48 LeBreton J. Senter J. Answers to 20 questions about interrater reliability and interrater agreement Organizational Research Methods 11 4 2008 815 852 10.1177/1094428106296642
49 Lehberger M. Kleih A.-K. Sparke K. Panic buying in times of coronavirus (COVID-19): Extending the theory of planned behavior to understand the stockpiling of nonperishable food in Germany Appetite 161 2021 105118 33450299
50 Leng A.L. Maitland E. Wang S.Y. Nicholas S. Liu R.G. Wang J. Individual preferences for COVID-19 vaccination in China Vaccine 39 2 2021 247 254 10.1016/j.vaccine.2020.12.009 33328140
51 Li L.i. Liu Y. Wu P. Peng Z. Wang X. Chen T. Influenza-associated excess respiratory mortality in China, 2010–15: A population-based study The Lancet Public Health 4 9 2019 e473 31493844
52 Lindell M.K. Communicating imminent risk Rodríguez H. Trainor J. Donner W. Handbook of Disaster Research 2nd ed. 2018 Springer New York 449 477
53 Lindell M.K. Arlikatti S. Prater C.S. Why people do what they do to protect against earthquake risk: Perceptions of hazard adjustment attributes Risk Analysis 29 8 2009 1072 1088 10.1111/j.1539-6924.2009.01243.x 19508448
54 Lindell M.K. Perry R.W. The Protective Action Decision Model: Theoretical modifications and additional evidence Risk Analysis 32 4 2012 616 632 10.1111/j.1539-6924.2011.01647.x 21689129
55 Lindell M.K. Prater C.S. Gregg C.E. Apatu E. Huang S.-K. Wu H.-C. Households’ immediate responses to the 2009 Samoa earthquake and tsunami International Journal of Disaster Risk Reduction 12 2015 328 340 10.1016/j.ijdrr.2015.03.003
56 Liu Z. Yang J. In the wake of scandals: How media use and social trust influence risk perception and vaccination intention among Chinese parents Health Communication 36 10 2021 1188 1199 10.1080/10410236.2020.1748834 32264705
57 Luo R.F. Liu C.F. Gao J.J. Wang T.Y. Zhi H.Y. Shi P.F. Impacts of the COVID-19 pandemic on rural poverty and policy responses in China Journal of Integrative Agriculture 19 12 2020 2946 2964 10.1016/s2095-3119(20)63426-8
58 MacPherson-Krutsky C.C. Brand B.D. Lindell M.K. From information to public preparedness in the Cascadia Subduction Zone: Examining risk communication outcomes in Metropolitan Portland OR. Earthquake Spectra 38 2 2022 1223 1253 10.1177/87552930211057582
59 Mathieu E. Ritchie H. Ortiz-Ospina E. Roser M. Hasell J. Appel C. A global database of COVID-19 vaccinations Nature Human Behaviour 5 7 2021 947 953 10.1038/s41562-021-01122-8
60 Naeem S.B. Bhatti R. The Covid-19 ‘infodemic’: A new front for information professionals Health Information & Libraries Journal 37 3 2020 233 239 10.1111/hir.12311 32533803
61 Nakaya N. Nemoto H. Yi C. Sato A. Shingu K. Shoji T. Effect of tsunami drill experience on evacuation behavior after the onset of the Great East Japan Earthquake International Journal of Disaster Risk Reduction 28 2018 206 213
62 National Bureau of Statistics. (2021). Census data. http://www.stats.gov.cn/tjsj/pcsj/.
63 National Health Commission of China. (2022). The latest situation of the COVID-19 pandemic. (Accessed January 1, 2022) http://www.nhc.gov.cn/wjw/xwdt/list.shtml
64 Nehal K.R. Steendam L.M. Campos Ponce M. van der Hoeven M. Smit G.S.A. Worldwide vaccination willingness for COVID-19: A systematic review and meta-analysis Vaccines 9 10 2021 10.3390/vaccines9101071
65 Nunnally J.C. Bernstein I.H. Psychometric Theory 3rd ed 1994 McGraw-Hill New York
66 Our World in Data (2022). Coronavirus (COVID-19) vaccinations. Oct 22, 2022. https://ourworldindata.org/covid-vaccinations
67 Ott R.L. Longnecker M.T. Introduction to statistical methods & data analysis 2015 Cengage Learning
68 Qin H. Sanders C. Prasetyo Y. Syukron M. Prentice E. Exploring the dynamic relationships between risk perception and behavior in response to the Coronavirus Disease 2019 (COVID-19) outbreak Social Science & Medicine 285 2021 114267 34388619
69 Schuman H. Kalton G. Survey methods Lindzey G. Aronson E. Handbook of social psychology Vol. 1 1985 Random House 635 697
70 Schutt R.K. Investigating the Social World: The Process and Practice of Research 2018 Sage publications Los Angeles
71 Sichuan Provincial Bureau of Statistics. (2021). Statistical yearbook of 2021. (Accessed January 1, 2022) http://tjj.sc.gov.cn/scstjj/c105855/nj.shtml
72 Siegrist M. The necessity for longitudinal studies in risk perception research Risk Analysis 33 1 2013 50 51 10.1111/j.1539-6924.2012.01941.x 23231682
73 Siegrist M. Longitudinal studies on risk research Risk Analysis 34 8 2014 1376 1377 10.1111/risa.12249 24975877
74 Siegrist M. Gutscher H. Natural hazards and motivation for mitigation behavior: People cannot predict the affect evoked by a severe flood Risk Analysis 28 3 2008 771 778 10.1111/j.1539-6924.2008.01049.x 18643832
75 Sjoberg L. Engelberg E. Risk perception and movies: a study of availability as a factor in risk perception Risk Anal. 30 1 2010 95 106 20055978
76 Soroya S.H. Farooq A. Mahmood K. Isoaho J. Zara S.E. From information seeking to information avoidance: Understanding the health information behavior during a global health crisis Information Processing & Management 58 2 2021 16 10.1016/j.ipm.2020.102440
77 Sun Y. Hu Q. Grossman S. Basnyat I. Wang P. Comparison of COVID-19 information seeking, trust of information sources, and protective behaviors in China and the US Journal of Health Communication 26 9 2021 657 666 10.1080/10810730.2021.1987590 34702136
78 Sun Y. Monnat S.M. Rural-urban and within-rural differences in COVID-19 vaccination rates The Journal of Rural Health 38 4 2022 916 922 10.1111/jrh.12625 34555222
79 Szmyd B. Karuga F.F. Bartoszek A. Staniecka K. Siwecka N. Bartoszek A. Attitude and behaviors towards SARS-CoV-2 vaccination among healthcare workers: A cross-sectional study from Poland Vaccines 9 3 2021 10.3390/vaccines9030218
80 Tang L. Zou W. Health information consumption under COVID-19 lockdown: An interview study of residents of Hubei Province China. Health Communication 36 1 2021 74 80 10.1080/10410236.2020.1847447 33167736
81 Terpstra T. Lindell M.K. Citizens’ perceptions of flood hazard adjustments: An application of the Protective Action Decision Model Environment and Behavior 45 2013 993 1018 10.1177/0013916512452427
82 Wake A. The willingness to receive COVID-19 vaccine and its associated factors: “Vaccination refusal could prolong the war of this pandemic” A systematic review Risk Management And Healthcare Policy 14 2021 2609 2623 10.2147/RMHP.S311074 34188572
83 Wang F. Wei J. Huang S.-K. Lindell M. Ge Y. Wei H. Public reactions to the 2013 Chinese H7N9 Influenza outbreak: Perceptions of risk, stakeholders, and protective actions Journal of Risk Research 21 7 2018 809 833 10.1080/13669877.2016.1247377
84 Wang J. Yuan B. Lu X. Liu X. Li L. Geng S. Willingness to accept COVID-19 vaccine among the elderly and the chronic disease population in China Human Vaccines & Immunotherapeutics 17 12 2021 4873 4888 10.1080/21645515.2021.2009290 34906026
85 Wang Y. Cheng M. Wang S. Wu F. Yan Q. Yang Q. Vaccination coverage with the pneumococcal and influenza vaccine among persons with chronic diseases in Shanghai, China, 2017 BMC Public Health 20 1 2020
86 Wei H.-L. Lindell M.K. Prater C.S. Wei J. Wang F. Texas households’ expected responses to seasonal influenza Journal of Risk Research 24 11 2021 1405 1425
87 World Health Organization (WHO). (2021). Coronavirus disease (COVID-19): Similarities and differences between COVID-19 and Influenza. Available at https://www.who.int/news-room/questions-and-answers/item/coronavirus-disease-covid-19-similarities-and-differences-with-influenza.
88 Wyte-Lake T. Der-Martirosian C. Chu K.R. Johnson-Koenke R. Dobalian A. Preparedness and response activities of the US Department of Veterans Affairs (VA) home-based primary care program around the fall 2017 hurricane season BMC Public Health 20 1 2020 9 10.1186/s12889-020-09888-8 31906907
89 Ye M. Lyu Z. Trust, risk perception, and COVID-19 infections: Evidence from multilevel analyses of combined original dataset in China Social Science & Medicine 265 2020 10.1016/j.socscimed.2020.113517
90 Yin F.L. Wu Z.L. Xia X.Y. Ji M.Q. Wang Y.Y. Hu Z.W. Unfolding the determinants of COVID-19 vaccine acceptance in China Journal of Medical Internet Research 23 1 2021 17 10.2196/26089
91 Yue L. Sun L. Yang K. Li Z. Research on the status quo and strategy of China’s medical and health system reform (in Chinese) China Health Industry 16 19 2019 185 188 10.16659/j.cnki.1672-5654.2019.19.185
92 Zhang J. Litvinova M. Liang Y. Wang Y. Wang W. Zhao S. Changes in contact patterns shape the dynamics of the COVID-19 outbreak in China Science eabb8001 2020 10.1126/science.abb8001
93 Zheng J. Morstead T. Sin N. Klaiber P. Umberson D. Kamble S. Psychological distress in North America during COVID-19: The role of pandemic-related stressors Social Science & Medicine 270 2021 10.1016/j.socscimed.2021.113687
94 Zhu F.C. Li Y.H. Guan X.H. Hou L.H. Wang W.J. Li J.X. Safety, tolerability, and immunogenicity of a recombinant adenovirus type-5 vectored COVID-19 vaccine: A dose-escalation, open-label, non-randomised, first-in-human trial Lancet 395 10240 2020 1845 1854 10.1016/s0140-6736(20)31208-3 32450106
| 0 | PMC9705199 | NO-CC CODE | 2022-12-04 23:14:57 | no | Vaccine. 2022 Nov 29; doi: 10.1016/j.vaccine.2022.11.062 | utf-8 | Vaccine | 2,022 | 10.1016/j.vaccine.2022.11.062 | oa_other |
==== Front
Cell Rep
Cell Rep
Cell Reports
2211-1247
Cell Press
S2211-1247(22)01723-5
10.1016/j.celrep.2022.111831
111831
Article
Cross-reaction of current available SARS-CoV-2 MAbs against the pangolin-origin coronavirus GX/P2V/2017
Jia Yunfei 12#
Niu Sheng 1#
Hu Yu 24#
Chai Yan 2
Zheng Anqi 25
Su Chao 26
Wu Lili 2
Han Pengcheng 27
Han Pu 2
Lu Dan 23
Liu Zhimin 1
Yan Xinxin 12
Tian Di 8
Chen Zhihai 8
Qi Jianxun 2
Tian Wen-xia 1∗
Wang Qihui 123∗
Gao George Fu 1239∗
1 College of Veterinary Medicine, Shanxi Agricultural University, Jinzhong 030801, China
2 CAS Key Laboratory of Pathogen Microbiology and Immunology, Institute of Microbiology, Chinese Academy of Sciences, Beijing 100101, China
3 Savaid Medical School, University of Chinese Academy of Sciences, Beijing 100049, China
4 School of Life Sciences, University of Science and Technology of China, Hefei, Anhui 230026, China
5 University of the Chinese Academy of Sciences, Beijing 100049, China
6 Department of Biomedical Sciences, City University of Hong Kong, Hong Kong SAR 999077, China
7 School of Medicine, Zhongda Hospital, Southeast University, Nanjing 210009, China
8 Center of Infectious Disease, Beijing Ditan Hospital, Capital Medical University, Beijing 100015, China
∗ Corresponding author (G.F.G.), (Q.W.) and (W.X.T.).
9 Lead contact
# These authors contributed equally.
29 11 2022
29 11 2022
11183116 6 2022
27 10 2022
22 11 2022
© 2022.
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Since the identification of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the etiological agent of COVID-19, multiple SARS-CoV-2-related viruses have been characterized, including pangolin-origin GD/1/2019 and GX/P2V/2017. Our previous study indicated that both viruses have potentials to infect humans. Here, we find that CB6 (commercial name etesevimab), a COVID-19 therapeutic monoclonal antibody (MAb) developed by our group, efficiently inhibits GD/1/2019 but not GX/P2V/2017. A total of 50 SARS-CoV-2 MAbs divided into seven groups based on their receptor-binding domain (RBD) epitopes, together with the COVID-19 convalescent sera, are systematically screened for their cross-binding and cross-neutralizing properties against GX/P2V/2017. We find that GX/P2V/2017 displays substantial immune difference from SARS-CoV-2. Furthermore, we solve two complex structures of the GX/P2V/2017 RBD with MAbs belonging to RBD-1 and RBD-5, providing a structural basis for their different antigenicity. These results highlight the necessity for broad anti-coronavirus countermeasures and shed light on potential therapeutic targets.
Graphical abstract
Pangolin-origin CoVs pose a potential threat to humans. Jia et al. test the cross-reaction of current available SARS-CoV-2 MAbs against the pangolin-origin coronavirus GX/P2V/2017. Structural analysis of two MAbs provides further insight into the immune difference of the pangolin-origin CoV.
Keywords
pangolin CoVs
SARS-CoV-2
monoclonal antibody
immune difference
==== Body
pmc
| 36493785 | PMC9705200 | NO-CC CODE | 2022-12-13 23:17:05 | no | Cell Rep. 2022 Dec 13; 41(11):111831 | utf-8 | Cell Rep | 2,022 | 10.1016/j.celrep.2022.111831 | oa_other |
==== Front
Resour Conserv Recycl
Resour Conserv Recycl
Resources, Conservation, and Recycling
0921-3449
1879-0658
Elsevier B.V.
S0921-3449(22)00632-2
10.1016/j.resconrec.2022.106800
106800
Article
Cascading impacts of global metal mining on climate change and human health caused by COVID-19 pandemic
Wang Yao a
Wang Heming a
Wang Peng b
Zhang Xu a
Zhang Zhihe a
Zhong Qiumeng c
Ma Fengmei b
Yue Qiang a
Chen Wei-Qiang b
Du Tao a
Liang Sai c⁎
a State Environmental Protection Key Laboratory of Eco-Industry, Northeastern University, Shenyang, 110819, People's Republic of China
b Key Lab of Urban Environment and Health, Institute of Urban Environment, Chinese Academy of Sciences, Xiamen, 361021, People's Republic of China
c Key Laboratory for City Cluster Environmental Safety and Green Development of the Ministry of Education, School of Ecology, Environment and Resources, Guangdong University of Technology, Guangzhou, 510006, People's Republic of China
⁎ Corresponding author.
29 11 2022
3 2023
29 11 2022
190 106800106800
2 7 2022
5 11 2022
24 11 2022
© 2022 Elsevier B.V. All rights reserved.
2022
Elsevier B.V.
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
The coronavirus disease 2019 (COVID-19) pandemic has significantly disrupted global metal mining and associated supply chains. Here we analyse the cascading effects of the metal mining disruption associated with the COVID-19 pandemic on the economy, climate change, and human health. We find that the pandemic reduced global metal mining by 10-20% in 2020. This reduction subsequently led to losses in global economic output of approximately 117 billion US dollars, reduced CO2 emissions by approximately 33 million tonnes (exceeding Hungary's emissions in 2015), and reduced human health damage by 78,192 disability-adjusted life years. In particular, copper and iron mining made the most significant contribution to these effects. China and rest-of-the-world America were the most affected. The cascading effects of the metal mining disruption associated with the pandemic on the economy, climate change, and human health should be simultaneously considered in designing green economic stimulus policies.
Keywords
Environmental impacts
Input-output analysis
COVID-19
Metals
Globalization
==== Body
pmc1 Introduction
The COVID-19 pandemic has posed unprecedented challenges to global supply chains and created exceptional socioeconomic hardships (Diffenbaugh et al., 2020; Guan et al., 2020). Among the global supply chains, the metal mining and production sectors are of particular concern because of their high economic value, environmental impacts, and relevance to low-carbon technologies (Hu et al., 2023, Wang et al., 2022). The use of certain metal minerals, such as lithium and cobalt, will need to ramp up by nearly 500% to 2050 in relation to 2018, while the global demand for metals used to deploy wind, solar, and geothermal power, as well as energy storage, will rise to 3 billion tonnes required to transition to a low-carbon economy (Hund et al., 2020). The consumption of critical metals has increased sharply, and as a result, demand and supply were stretched before the COVID-19 pandemic (Eggert, 2011; Hayes and McCullough, 2018). Moving forward, metal supply assurance will be increasingly important worldwide to ensure resource security, which is in line with the Sustainable Development Goals (SDGs) launched by the United Nations (Franks et al., 2022).
As of July 2020, more than 275 metal operations had been disrupted, and the direct loss is estimated to be nearly 9 billion US dollars (at current prices) (S&P, 2020). Latin America and Africa have been the most significantly affected. For example, Peru's lockdown disrupted 12% of the global copper output in 2020 (IEA, 2020), and Chile closed its borders in April 2021 as COVID-19 cases soared, worsening the tight global supply of copper (Writer, 2021). This significant disturbance to the mining sectors can have important cascading impacts on the global economy and environment through the globalized supply chains. Furthermore, finding low-carbon pathways to address the ecological crisis while promoting economic recovery is necessary to achieve the targets specified in the Paris Climate Agreement (IAP, 2020; Shan et al., 2020; Shao et al., 2022). Following the pandemic, global and science-based solutions and green recovery strategies on the supply side of the metal mining sector will also need to be devised and implemented. Thus, characterizing the cascading effects of this disruption on the economies, climate change, and human health of different countries through global supply chains is the prerequisite for solving the above problems.
For the influence on global metal mining caused by the COVID-19 pandemic, most recent studies have focused on the pandemic's effect on metal prices and investments (Habib et al., 2021), the establishment of early warning mechanisms for emergencies (Zhu et al., 2021), and the mitigation of the negative effects of metal mining and production on low-carbon technologies (Akcil et al., 2020; Goldthau and Hughes, 2020). However, few studies have quantitatively analysed the effects of metal mining disturbances on the economic output of nations. The input-output (IO) model has been widely used for disaster impact analysis, which is a tool for measuring the effects of sudden external shocks on the economy (Koks and Thissen, 2016; Okuyama and Santos, 2014). The IO model is used to assess the economic impacts of any sectoral supply change through the linkage across countries embodied in the input-output table (Rocco et al., 2020; Shao et al., 2022). Some scholars have used this method to carry out related studies at the global (Lenzen et al., 2020), national (Cottafava et al., 2022), and regional levels (Dyason et al., 2021). This method emphasizes interactions between producers and consumers and underlines their role in contributing to loss (Rose and Lim, 2002; Zeng and Guan, 2020).
Furthermore, mining and industries related to metal can be a detrimental and intense stressor on the environment and human health (Dialga and Ouoba, 2022; Zhao et al., 2019), whilst also including the impact of spatiotemporal variation (Wang et al., 2020c). Previous research has shown that the environment can benefit from the pandemic, as demonstrated by the abrupt decrease in global CO2 emissions (Le Quéré et al., 2020; Liu et al., 2020) and improvement in air quality that coincided with lockdown measures (Le et al., 2020; Wang et al., 2021). However, the relative contributions of different sectors, particularly the metal mining sector, to these benefits remain unknown. Generally, the failure to consider the cascading gains and losses from supply chains might cause them to be overlooked in the design of epidemic control measures in the future. The global environmental extended multiregional input-output (EE-MRIO) model makes it possible to calculate the climate change and human health impacts caused by sudden changes. The EE-MRIO model is widely used to analyse trade-related environmental impacts, such as carbon emissions (Davis and Caldeira, 2010), material consumption (Wang et al., 2020b; Wiedmann et al., 2015), mercury emissions (Chen et al., 2019; Qi et al., 2019), and the use of scarce water resources (Huang et al., 2015; Qu et al., 2018; Wu et al., 2022). Applying the EE-MRIO model can capture supply-chain-driven impacts across regions for multiple environmental indicators and support measures for international cooperation between primary suppliers and emitters of direct emissions from the supply viewpoint (Li et al., 2022; Liang et al., 2017).
Here, we assess the cascading effect of the disruption of the global metal mining sector caused by the COVID-19 pandemic on the economy, emissions, and human health. First, we evaluate how the pandemic affected metal mining production based on the database of metal production at risk built by S&P Global Market Intelligence (S&P, 2020) and 2020 metal mining statistics (Table S1 in Supplementary information). We then use the EE-MRIO model, linking metal mining production data and EXIOBASE (version 3.6) data, to evaluate the economic losses (measured by production output) and the environmental and human health gains of nations compared with conditions without COVID-19 (see Methods). In addition to exploring the economic, environmental, and health impacts of nations associated with metal supply chains, our analysis makes it possible to quantify the relative contributions that the disruption of the mining of different metals made to the benefits that countries received. This quantification aids in designing green economic stimulus policies that will help achieve the long-term temperature targets of the Paris Agreement and improve human health following the COVID-19 pandemic.
2 Methods and data
2.1 Environmental and human health impacts
In this study, we selected six main categories of emissions (CO2, PM2.5, NMVOC, N2O, NH3, and CH4) to reflect the effects of climate change on various regions. The emission inventories that have not impacted the industry chain due to the epidemic were obtained from the environmental satellite table in the EXIOBASE database (Stadler et al., 2018). Human health damage impacts caused by pollutant emissions were computed by multiplying pollutant emissions with corresponding life cycle impact (LC-IMPACT) characterization factors (Liang et al., 2013). The LC-IMPACT method is one of the Life Cycle Impact Assessment (LCIA) methods that offers characterization factors to quantify the endpoint environmental impacts, such as human health damage (the unit is disability-adjusted life years (DALY)) (Verones, Francesca. et al., 2020). Four impact categories based on selected pollutant emissions include climate change, ozone depletion, photochemical ozone formation, and particulate matter (PM) formation, all of which affect human health (Verones, Francesca et al., 2020).
The quantity of each sector's human health impacts fh, which is a matrix, is then calculated by Eq. (1):(1) eh=ep×CFr
where the m×n matrix eh indicates m categories of human health impacts by n economic sectors. The k×n matrix ep represents the emissions of k categories by n economic sectors. The m×k matrix CFr stands for the LC-IMPACT characterization factors converting k categories of emissions into m categories of human health impacts in region r.
2.2 Global EE-MRIO model
We constructed the EE-MRIO model by using the EXIOBASE database, which included the monetary flows of 49 regions and 200 product groups with relatively high sectorial details and extensive environmental data (Stadler et al., 2018). Also, the database disaggregated “mining and quarrying” from “manufacturing of basic metals” activities, which enabled us to assess the effects of the extraction of each metal and improve the reliability of the results.
Given that lockdown measures have partly or wholly closed metal mines during the pandemic, we calculated income-based emissions and the human health damage of nations to evaluate the impacts on interregional trade on the supply side. The basic equation of the MRIO model is given by(2) x=v(I−B)−1=vG
where x is the 1×n vector of the total input of region sectors (each sector total input equals its total output); v is the 1×n vector of the value-added creation of each regional sector; and B=(bij) is the n×n matrix of the direct-output coefficient. Element bijst equals the direct input from sector i in region s to sector j in region t divided by the total output of region sector i. G=(I−B)−1 is the n×n matrix, which is known as the Ghosh inverse matrix; the element gijst represents coefficients specifying the proportion of the total output of sector i in region s used by each primary input of sector j in region t. Matrix I is an identity matrix.
By defining a 1×n intensity vector f to represent the k-th emissions ep or the m-th human health damage eh of each region sector for its unitary output, as shown by Eq. (3), we can calculate supply-side emissions or human health damage before the pandemic by Eq. (4):(3) f=e(x′^)−1
(4) Er=vrGfT
where e represents the k-th direct emissions ep or the m-th human health damage eh of region sectors (indicated by 1×n row vector e); the column vector x′ represents total outputs of region sectors; Er is income-based emissions or human health damage of each sector of region r; fT denotes the transposition of vector f. The hat ^ means diagonalizing the vector x. The notation ′ means the transposition.
During the COVID-19 outbreak, the metal mining sector has faced the risk of not having a sufficient supply of labour to meet production requirements. The negative impacts of local production can be traded directly or indirectly to the production of external regions through global supply chains (Qu et al., 2018). In this study, a pathway linking reductions in metal extraction to potential output losses for each sector throughout 2020 was conceived. The potential direct input loss (Ulr) from the reduction in metal production in the case of the pandemic is defined as:(5) Ulr=dlr×vlr
where dlr is the vector of the decreasing proportion of metal extraction in the metal mining sector l of region r. Considering that we focus on the economic, environmental, and human health impacts stemming from the reduction in metal mining sectors, we set the other sectors as 0 in the calculation.
All direct and indirect economic losses, decrease in emissions, and decrease in human health damage in region r stemming from the reduction in metal mining sectors can be expressed as:(6) Δxr=Ulr×(I−B)−1=Ulr×G
(7) ΔEr=UlrGfT
The elements of the vector Δxr indicate the economic losses of each region enabled by the primary inputs of region r. The elements of the vector ΔEr represent the environmental or human health impact reduction enabled by the primary inputs of region r.
2.3 Data sources
In this study, eight categories of metals (including copper, lead, nickel, zinc, iron, cobalt, lithium, and molybdenum) were examined. Metal data were collected from S&P Global Market Intelligence (S&P, 2021) and the statistical yearbooks of various regions for 2020. For regions without published metal mining statistics, we used the database of metal production at risk built by S&P Global Market Intelligence (S&P, 2020) to obtain data on the reduction in the entire year's metal mining production. Because this was issued on June 25, 2020, it permitted only the decrease in global metal mining for the first and second quarters to be determined. We collected metal mining data for the third and fourth quarters based on previously published data and report them as a supplement (Table S1). Given that the pandemic began to break out globally in the second quarter, we supplemented missing data by assuming that the stringency of the lockdown measures in the third and fourth quarters was the same as those in the second quarter. For comparison, we use S&P Global Market Intelligence's metal production forecast for 2020 in the pre-pandemic period to estimate the pandemic-induced decline in metal production (S&P, 2021). This study uses the MRIO table in 2015 derived from the EXIOBASE (version 3.6) database (Stadler et al., 2020). The EXIOBASE database provides detailed data for the 49 regions (Table S2) and 200 product classifications (Table S3), for which environmental satellites contain data on industry-specific and final demand air emissions for pollutants. Since the EXIOBASE database's currency unit is the euro, all currencies were converted to 2015 US dollars using the average exchange rate for easy reference (UK, 2021). The life expectancy at birth data was derived from the world bank database (World Bank, 2022).
3 Results
3.1 Effects of the metal mining disruption on the economy during the COVID-19 pandemic
Fig. 1 shows the economic impacts associated with the reductions in metal mining in 2020 caused by COVID-19. Overall, the total global economic losses reached 116.9 billion US dollars, compared with conditions without COVID-19. The economic output losses were concentrated on the American continent (Fig. 1a), as this is the central area of metal ore production and has been affected the most by lockdown measures (Magudulela, 2020). In terms of metal categories, the reduction in copper mining had the greatest effect on the global economy (59.9 billion dollars), followed by iron (27.9 billion dollars) and nickel (13.1 billion dollars) (Table S4). In addition, refined metal production in multiple countries was affected by temporary smelter shutdowns, which affected the entire industrial chain. Thus, the COVID-19 pandemic has also significantly affected the metal refining industry and metal manufacturing, leading to economic losses in various regions, especially the United States, Mexico, and East Asia (China, South Korea, and Japan).Fig. 1 Effects of the metal mining reduction caused by the COVID-19 pandemic on economic output losses in 2020. Notes: a The worldwide economic output losses caused by metal mining reduction. b The top 15 countries or regions ranked by the decrease in economic output stemming from reductions in the mining of various metals. RoW Asia and Pacific is rest-of-the-world Asia and Pacific. RoW Middle East is rest-of-the-world Middle East. China refers to mainland China.
Fig 1
Among countries situated upstream in the industrial metal chains, such as China, rest-of-the-world America (RoW America), and Canada, economic output losses of 32.1, 15.5, and 12.8 billion US dollars were observed, respectively (Fig. 1b). Among those, the impact of copper mining mainly on China and RoW America ranked the highest (20.3 and 11.2 billion US dollars, respectively), whereas the reduction in iron ore mining had the most considerable effect on Brazil's economy (11.9 billion US dollars). RoW America (including Peru and Chile) and Brazil are major mining resource suppliers worldwide, and thus their resource supply disruptions have driven large amounts of upstream economic losses. Significant economic losses were also observed for countries located downstream in the industrial metal chain. The economies of the United States, Germany, and East Asia, whose primary industries are metal processing, machinery manufacturing, and automobile manufacturing, were affected the most. Specifically, in the United States and Germany, the economic losses induced by the reduction in metal mining were as high as 7.2 and 1.4 billion US dollars, respectively. For Japan, South Korea, and Chinese Taiwan, with few domestic resources, these losses were 4.5, 3.9, and 2.8 billion US dollars, respectively. In order to mitigate this influence, countries have been working to maintain metal production to prevent further losses. For example, iron ore production in China was 7% higher in the second quarter of 2020 than over the same period in 2019 (NBS, 2021). Additionally, mining giant Vale announced to resume production at Brazil's Itabira mining complex on June 17, 2020 (Bnamericas, 2020). However, the COVID-19 lockdown still had a non-negligible impact on some countries. For example, in Peru, copper production through July 2020 fell by nearly 23% from that in the same period in 2019 (USGS, 2021).
3.2 Effects of the reduction in metal mining on emissions during the COVID-19 pandemic
The reduction in emissions across countries and regions stemming from the reduction in metal mining in 2020 is shown in Fig. 2 . Global CO2 emissions decreased by 32.6 million tonnes because of the disruption of the metal mining sector, equivalent to more than the annual CO2 emissions of Hungary (33.6 million tonnes), Sweden (33.1 million tonnes), and Switzerland (26.9 million tonnes) (Fig. 2a). In addition, some regions experienced significant reduction in CO2 emissions (Fig. 2b). China experienced the largest reduction in CO2 emissions (12.2 million tonnes), followed by Canada (5.3 million tonnes) and RoW America (4.1 million tonnes). The impact of primary metal inputs dominated the income-based CO2 emission reduction of China's metal industry chain, the largest CO2 emitter in the world. In practical terms, the impacts of the primary inputs of copper, iron, and nickel caused CO2 emissions reduction of 5.6, 4.7, and 0.9 million tonnes, respectively, in China. The metal production-related industries with significant emissions reduction in other regions include iron in Brazil (3.0 million tonnes) and RoW America (1.6 million tonnes), and copper in RoW America (1.9 million tonnes) and Canada (1.6 million tonnes). The abovementioned regions are the primary producers of metal mines. The impacts of mine production and related downstream industries (such as smelters and refineries) owing to restrictions implemented by countries in response to the COVID-19 pandemic promoted reductions in CO2 emissions.Fig. 2 The reduction in emissions stemming from the effect on metal mining during the COVID-19 pandemic. Notes: The six pollutants are in six groups. a, c, e, g, i, and k show the reduction in emissions in different regions for CO2, PM2.5, NMVOC, CH4, NH3, and N2O, respectively. b, d, f, h, j, and l show the top 10 regions experiencing the most significant emissions reduction stemming from the reduction in metal mining. RoW Africa indicates rest-of-the-world Africa. The emissions reduction is provided in Table S5 and Table S6.
Fig 2
In addition to CO2 emissions, our results demonstrate reductions in the emissions of PM2.5, NMVOC, CH4, NH3, and N2O stemming from the reduced production of primary metal minerals. As a large amount of smoke and dust can be generated during the mining and smelting of metals, lockdown policies induced global decreases in PM2.5, NMVOC, CH4, NH3, and N2O emissions of 59.8, 37.3, 36.8, 5.2, and 0.4 thousand tonnes, respectively (Fig. 2c-l and Table S5). Notably, the reduction with the largest emission ratio changes was PM2.5 (Table S5). The two regions that experienced the greatest decrease in PM2.5 emissions because of metal mining were RoW America (22.7 thousand tonnes) and Brazil (16.7 thousand tonnes), which are also the leading metal suppliers in the world.
In 2020, the countries with the largest decrease in other pollutants were China for NMVOC, CH4, and N2O (10.4, 19.1, and 0.2 thousand tonnes, respectively) and RoW America for NH3 (4.2 thousand tonnes). It was found that China was the country that experienced the most significant reduction in emissions, which was attributed to the fact that the effect on copper mining and related industries made the most outstanding contribution to the reduction in CH4 (8.9 thousand tonnes) and NMVOC (6.4 thousand tonnes) emissions. In addition, CH4 emissions from copper mining and the industrial chain overall during the COVID-19 pandemic made the most considerable contribution to the reduction in emissions in other countries (Table S6). It is not surprising because copper mining and related industrial processes have been considered significant sources of greenhouse gas (GHG) emissions (Alvarado et al., 2002; Northey et al., 2013). Elsewhere, copper minerals, iron ores, and lead-zinc minerals were responsible for reducing PM2.5 emissions in RoW America by 19.6, 1.4, and 0.9 thousand tonnes, respectively (Fig. 2d). RoW Asia and Pacific ranked first in reducing NH3 emissions, and the reduction in copper mining was responsible for a reduction in NMVOC emissions of 1.4 thousand tonnes.
3.3 Benefits to human health associated with the reduction in emissions
The long-term effects of worsening air quality are associated with an increased risk of chronic obstructive pulmonary diseases, lung cancer, and stroke (Cohen et al., 2017). Thus, in this regard, the reduction in emissions accompanying the pandemic has decreased healthy years lost due to premature death or disability. Although more than 20.5 million years of life have been estimated to be lost to COVID-19 worldwide in 2020 (Pifarré i Arolas et al., 2021), the reduction in metal mining indirectly provided some health benefits. In this study, we defined life extension as the benefiting population equivalent, which was derived from the reduction in disability-adjusted life years divided by the human life expectancy indicator in various regions to characterize the extent of the human health benefits. The world experienced an increase of 1,024 benefiting population equivalents, or approximately 78,000 DALY were reduced because of the reduction in metal mining (Table S7 and Table S8). Iron and copper made the largest contributions to the reduction in DALY (38.6% and 38.4%, respectively).
The reduction in emissions will lead to less climate change and stratospheric ozone depletion, all of which positively affect public health worldwide. Fig. 3 shows the top twenty regions worldwide that experienced the greatest benefits to human health stemming from emission decreases. The most substantial health benefits were concentrated in regions where more metal ores were extracted, such as Canada, Brazil, and RoW America. The health benefit ratios of these top three regions among the global population were as high as 1.1%, 0.6%, and 0.5%, respectively, compared with the damage before the COVID-19 pandemic in these regions. China had the largest decline in health impacts (17,892 DALY, 233 benefiting population equivalents), and the copper industry had the most significant impact (8,222 DALY, 107 benefiting population equivalents) worldwide (Table S9 and Table S10). Canada (7,513 DALY, 92 benefiting population equivalents) and RoW America (5,898 DALY, 78 benefiting population equivalents) ranked second and third in terms of reducing damage to human health among the total global population (Table S9 and Table S10).Fig. 3 The top twenty regions experienced the most significant health benefits from emissions reduction. Note: The x-axis shows the proportion of the reduction in human health damage compared to before the COVID-19 pandemic in this region. The left part shows the worldwide health benefits from regions that reduced the impacts of climate change and stratospheric ozone depletion. The right part shows the local health benefits from regions that reduced photochemical ozone formation and particulate matter formation. The top twenty regions in the chart are ranked according to the DALY data. The proportions of DALY reduction in different regions are shown in Table S11 and Table S14.
Fig 3
In addition, the impact of photochemical ozone formation and particulate matter formation on human health damage is changed in the DALY of local inhabitants as emissions have decreased in the source region. The human health ratio for all inhabitants in RoW America ranked first (2.3%) (Fig. 3). The rates of benefitting human health in Brazil and RoW Africa were 2.0% and 0.3%, respectively. Notably, China, RoW Asia and Pacific, and RoW America were the three regions where there was the greatest impact on local human health; the reduction in human health damage among these top three regions was as high as 18,075 DALY (236 benefiting population equivalents), 2,846 DALY (38 benefiting population equivalents), and 2,824 DALY (37 benefiting population equivalents), respectively (Table S12 and Table S13). As China has the largest population, serves as the world's factory for metal industrial manufacturers, and generates a large amount of emissions, it experienced substantial health benefits because of emissions reduction.
4 Discussion
Green economic stimulus policies for economic recovery are crucial for preventing global warming (Evans and Gabbatiss, 2020; Forster et al., 2020). The new insights from the cascading effects of the disruptions of metal mining associated with the pandemic on the economy, climate change, and human health can provide implications for developing green economic stimulus policies following the COVID-19 pandemic.
From a global perspective, metal mining sectors with greater economic losses and a lower reduction in environmental impacts and human health damage in specific regions should be given high priority when designing green economic recovery policies through global supply chains (Fig. 4 and Fig. S1-4). For example, the copper mining sector of RoW America (including Peru and Chile) in the red frame experienced more significant economic losses and lower benefits for environmental and human health; the same was the case for the copper and lead-zinc mining sectors of Mexico and the copper mining sectors of China and the United States. In contrast, the iron mining sectors in countries received relatively more minor economic losses but more environmental and health benefits, indicating they should not be given priority when developing green economic stimulus policies. In 2021, the global economic recovery and rising investment in low-emission technologies boosted metal mining production (USGS, 2021). However, few measures for the environmental impact dimension have been identified in metal mining recovery plans (OECD, 2022). Well-designed green recovery plans can generate the double dividend of increased resource security and environmental outcomes. Thus, metal's economic importance and environmental risk should be simultaneously considered in allocating international assistance and emergency aid based on the type of mining sector to optimize the achievement of a green economic recovery.Fig. 4 Composition of the global economic output losses, emissions reduction, and human health benefits associated with the COVID-19-mediated reduction in metal mining among selected countries and regions. Notes: a Comparison of global economic losses and emissions reduction. b Comparison of global economic losses and human health benefits. The black dashed lines in the figures represent the division of economic losses, environmental income, and human health benefits. The area above the black dashed line corresponds to more significant economic losses but lower emissions reduction (Fig. 4a) or human health benefits (Fig. 4b) stemming from the reduction in metal mining associated with the pandemic. The red frames indicate the countries and metal mining sectors that should prioritize restoring production capacity.
Fig 4
From the perspective of international cooperation, countries and regions are linked through metal production chains to a substantial degree. Specifically, we showed how international trade transmits metal mining risks to the global economy and affects the environment and human health (Fig. 5 ). Countries or regions relying on imports took major economic hits during the COVID-19 pandemic. For example, Bulgaria, India, Japan, and China suffered the most considerable economic output losses from RoW America by 0.13%, 0.04%, 0.04%, and 0.04%, respectively. On the other hand, in terms of CO2 emissions reduction, Bulgaria and China benefited from RoW America by 0.05% and 0.04%, respectively. These findings highlight the importance of global cooperation in designing economic stimulus policies.Fig. 5 Impacts propagated through trade for key regions in 2020. Notes: a Economic output losses embodied in trade. b CO2 emissions reduction embodied in trade. Shading indicates the net impacts embodied in trade with net exporters in blue and net importers in red. The arrows start from metal ore producers affected by COVID-19 and end at downstream producers. The numbers in the arrows indicate the volume of embodied impacts.
Fig 5
To effectively boost economic growth, resource-importing countries, such as South Korea, Japan, and Finland, need to strengthen their national cooperation with resource-exporting countries by providing medical supplies to aid pandemic control. Moreover, this study's major international pairs identified from the supply side provide a valuable reference for seeking international partners to control global CO2 emissions and benefit human health (Fig. 5 and Fig. S5). For example, resource-exporting nations (e.g., Canada) could transfer capital and their available technologies to aid the development of end-of-pipe emission reduction technologies in metal industries and to benefit human health in resource-importing countries and regions (e.g., Bulgaria, China, Chinese Taiwan, and Turkey) (Qi et al., 2019). From another perspective, for exporters, metal mining activities can lead to CO2 emissions and threaten biodiversity (Gan and Griffin, 2018; Sonter et al., 2020). Therefore, several developed importing countries (e.g., Japan and Sweden) can consider formulating investment policies to assist with reducing the CO2 emissions and ecological damage caused by the metal mining of resource exporters, such as Peru, Mexico, and Brazil. When these other determinants of comparative advantage are in place, a resource-abundant country tends to export resources to countries with a relative abundance in capital and skilled labour and import capital-intensive goods in return (WTO, 2010). Moreover, the design of a flexible and dynamic border tax adjustment mechanism can generate a double dividend for both resource-exporting and resource-importing nations based on the cascading effects of economic and environmental impacts along global supply chains (Fischer, 2011; Jakob and Marschinski, 2013).
In the long term, for countries, promoting resource-efficient production and reducing dependency on raw metal ores is the key to reducing the risks associated with international metal production chains (Luna-Nemecio et al., 2020; Wang et al., 2020a; Watari et al., 2020). Aside from diversifying metal mining areas, nations should consider pluralistic sources of metal resources when devising national strategies for utilizing such resources. Metal trade diversification can be promoted by identifying key trading partners and a wide range of investment policies targeting metal mineral-rich countries with low environmental impacts on the world. In most cases, metal recycling can provide significant economic and environmental benefits (Kirchherr et al., 2017; Nuss et al., 2019; Wang, H. et al., 2022). Thus, long-term strategies should include redesigning technologies to use diversified or alternative materials and strengthening the recycling efficiency of secondary resources.
Several potential limitations should be considered in this study. To further improve the estimation, the model presented in this study is required to incorporate the mine tailings of the mining of metals into further analysis. In addition, the global economic recovery associated with intensive fiscal stimulus measures may cause the negative environmental impact of uncertain policies. Future studies investigating how different fiscal stimulus packages of metal mining exert heterogeneous impacts on economic output and the environment through global supply chains are warranted by combining the computable general equilibrium (CGE) and IO models.
5 Conclusions
The COVID-19 pandemic has been an unprecedented threat to metal mining through isolated outbreaks and government mandated shutdowns. This study applied EEIO and LC-IMPACT methods to assess the effects of metal mining disturbances on climate change and human health at the global level in 2020. The results show that the COVID-19 pandemic reduced global metal mining by 10-20%. Overall, this reduction subsequently led to losses in global economic output of approximately 117 billion US dollars, reduced CO2 emissions by approximately 33 million tonnes (exceeding the annual emissions of Hungary), and reduced human health damage by 78,192 DALY. Notably, major disruptions such as copper and iron mining mainly occurred in RoW America and Brazil, which had a big impact on other countries. China and RoW America were the most affected by both the overall impact of environmental and human benefits. Thus, sectors with greater economic losses and a lower reduction in environmental and human health damage in specific regions should be given high priority when designing green economic recovery policies from a global perspective, such as the copper mining sector of RoW America. Long-term strategies should include redesigning technologies to use diversified or alternative materials and strengthening the recycling efficiency of secondary resources. On the basis of the cascading effects of mining disruption on the economy, climate change, and human health, this study can provide improved guidance for government policy on designing green economic stimulus policies toward more sustainable metal mining in countries all over the world.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Appendix Supplementary materials
Image, application 1
Image, application 2
Data availability
Data will be made available on request.
Acknowledgements
This study was financially supported by the National Natural Science Foundation of China (72293602, 52070034 and 41871204).
Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.resconrec.2022.106800.
==== Refs
References
Akcil A. Sun Z. Panda S. COVID-19 disruptions to tech-metals supply are a wake-up call Nature 587 2020 365 367 33204017
Alvarado S. Maldonado P. Barrios A. Jaques I. Long term energy-related environmental issues of copper production Energy 27 2 2002 183 196
Bnamericas Vale to resume Itabira iron ore production 2020 https://www.bnamericas.com/en/news/vale-to-resume-itabira-iron-ore-production
Chen L. Liang S. Liu M. Yi Y. Mi Z. Zhang Y. Li Y. Qi J. Meng J. Tang X. Zhang H. Tong Y. Zhang W. Wang X. Shu J. Yang Z. Trans-provincial health impacts of atmospheric mercury emissions in China Nat. Commun. 10 1 2019 1484 30940811
Cohen A.J. Brauer M. Burnett R. Anderson H.R. Frostad J. Estep K. Balakrishnan K. Brunekreef B. Dandona L. Dandona R. Feigin V. Freedman G. Hubbell B. Jobling A. Kan H. Knibbs L. Liu Y. Martin R. Morawska L. Pope C.A. 3rd Shin H. Straif K. Shaddick G. Thomas M. van Dingenen R. van Donkelaar A. Vos T. Murray C.J.L. Forouzanfar M.H. Estimates and 25-year trends of the global burden of disease attributable to ambient air pollution: an analysis of data from the global burden of diseases study 2015 Lancet 389 10082 2017 1907 1918 28408086
Cottafava D. Gastaldo M. Quatraro F. Santhia C. Modeling economic losses and greenhouse gas emissions reduction during the COVID-19 pandemic: Past, present, and future scenarios for Italy Econ. Model. 110 2022 105807
Davis S.J. Caldeira K. Consumption-based accounting of CO2 emissions Proceed. Nat. Acad. Sci. 107 12 2010 5687 5692
Dialga I. Ouoba Y. How do extractive resources affect human development? Evidence from a panel data analysis Resour. Environ. Sustain. 7 2022 100046
Diffenbaugh N.S. Field C.B. Appel E.A. Azevedo I.L. Baldocchi D.D. Burke M. Burney J.A. Ciais P. Davis S.J. Fiore A.M. Fletcher S.M. Hertel T.W. Horton D.E. Hsiang S.M. Jackson R.B. Jin X. Levi M. Lobell D.B. McKinley G.A. Moore F.C. Montgomery A. Nadeau K.C. Pataki D.E. Randerson J.T. Reichstein M. Schnell J.L. Seneviratne S.I. Singh D. Steiner A.L. Wong-Parodi G. The COVID-19 lockdowns: a window into the Earth System Nat. Rev. Earth & Environ. 1 9 2020 470 481
Dyason D. Fieger P. Rossouw R. Covid-19, the effect of lockdowns on retail expenditure and displacement effects on the regional economy The Australasian J. Regional Stud. 27 1 2021 66 87
Eggert R.G. Minerals go critical Nat. Chem. 3 9 2011 688 691 21860456
Evans S. Gabbatiss J. Coronavirus: Tracking How The World's‘Green Recovery Plans Aim To Cut Emissions 2020 https://www.carbonbrief.org/coronavirus-tracking-how-the-worlds-green-recovery-plans-aim-to-cut-emissions
Fischer C. Trade's growing footprint Nat. Clim. Change 1 3 2011 146 147
Forster P.M. Forster H.I. Evans M.J. Gidden M.J. Jones C.D. Keller C.A. Lamboll R.D. Quéré C.L. Rogelj J. Rosen D. Schleussner C.-F. Richardson T.B. Smith C.J. Turnock S.T. Publisher correction: current and future global climate impacts resulting from COVID-19 Nat. Clim. Change 10 10 2020 971 -971
Franks D.M. Keenan J. Hailu D. Mineral security essential to achieving the sustainable development goals Nat. Sustain 2022 10.1038/s41893-022-00967-9
Gan Y. Griffin W.M. Analysis of life-cycle GHG emissions for iron ore mining and processing in China—Uncertainty and trends Resour. Policy 58 2018 90 96
Goldthau A. Hughes L. Protect global supply chains for low-carbon technologies Nature 585 2020 528 530
Guan D. Wang D. Hallegatte S. Davis S.J. Huo J. Li S. Bai Y. Lei T. Xue Q. Coffman D.M. Cheng D. Chen P. Liang X. Xu B. Lu X. Wang S. Hubacek K. Gong P. Global supply-chain effects of COVID-19 control measures Nat. Human Behav. 4 6 2020 577 587 32493967
Habib K. Sprecher B. Young S.B. COVID-19 impacts on metal supply: how does 2020 differ from previous supply chain disruptions? Resour. Conserv. Recycl. 165 2021 105229
Hayes S.M. McCullough E.A. Critical minerals: A review of elemental trends in comprehensive criticality studies Resour. Policy 59 2018 192 199
Hu X. Wang C. Lim M.K. Chen W-Q. Teng L. Wang P. Wang H. Zhang C. Yao C. Ghadimi P. Critical systemic risk sources in global lithium-ion battery supply networks: Static and dynamic network perspectives Renew. Sust. Energ. Rev. 173, 113083 2023
Huang K. Wang Z. Yu Y. Yang S. Assessing the environmental impact of the water footprint in Beijing China Water Policy 17 5 2015 777 790
Hund K. La Porta D. Fabregas T.P. Laing T. Drexhage J. Minerals For Climate Action: The Mineral Intensity of The Clean Energy Transition 2020 World Bank
IAP Global Green Recovery After COVID-19: Using Scientific Advice To Ensure Social Equity, Planetary and Human Health, and Economic Benefits 2020 https://www.interacademies.org/sites/default/files/2020-07/IAP_Green_Recovery.pdf
IEA Clean Energy Progress After The Covid-19 Crisis Will Need Reliable Supplies of Critical Minerals 2020 https://www.iea.org/articles/clean-energy-progress-after-the-covid-19-crisis-will-need-reliable-supplies-of-critical-minerals
Jakob M. Marschinski R. Interpreting trade-related CO2 emission transfers Nat. Clim. Change 3 1 2013 19 23
Kirchherr J. Reike D. Hekkert M. Conceptualizing the circular economy: an analysis of 114 definitions Resour. Conserv. Recycl. 127 2017 221 232
Koks E.E. Thissen M. A multiregional impact assessment model for disaster analysis Econ. Syst. Res. 28 4 2016 429 449
Le Quéré C. Jackson R.B. Jones M.W. Smith A.J.P. Abernethy S. Andrew R.M. De-Gol A.J. Willis D.R. Shan Y. Canadell J.G. Friedlingstein P. Creutzig F. Peters G.P. Temporary reduction in daily global CO2 emissions during the COVID-19 forced confinement Nat. Clim. Change 10 7 2020 647 653
Le T. Wang Y. Liu L. Yang J. Yung Y.L. Li G. Seinfeld J.H. Unexpected air pollution with marked emission reductions during the COVID-19 outbreak in China Science 369 6504 2020 702 706 32554754
Lenzen M. Li M.Y. Malik A. Pomponi F. Sun Y.Y. Wiedmann T. Faturay F. Fry J. Gallego B. Geschke A. Gomez-Paredes J. Kanemoto K. Kenway S. Nansai K. Prokopenko M. Wakiyama T. Wang Y.F. Yousefzadeh M. Global socio-economic losses and environmental gains from the Coronavirus pandemic PLoS One 15 7 2020 13
Li Y. Zhong Q. Wang Y. Jetashree Wang H. Li H. Liang S. Scarcity-weighted metal extraction enabled by primary suppliers through global supply chains J. Clean. Prod. 371 2022 133435
Liang S. Qu S. Zhu Z. Guan D. Xu M. Income-based greenhouse gas emissions of nations Environ. Sci. Technol. 51 1 2017 346 355 27936320
Liang S. Xu M. Suh S. Tan R.R. Unintended environmental consequences and co-benefits of economic restructuring Environ. Sci. Technol. 47 22 2013 12894 12902 24117387
Liu Z. Ciais P. Deng Z. Lei R. Davis S.J. Feng S. Zheng B. Cui D. Dou X. Zhu B. Guo R. Ke P. Sun T. Lu C. He P. Wang Y. Yue X. Wang Y. Lei Y. Zhou H. Cai Z. Wu Y. Guo R. Han T. Xue J. Boucher O. Boucher E. Chevallier F. Tanaka K. Wei Y. Zhong H. Kang C. Zhang N. Chen B. Xi F. Liu M. Bréon F.-M. Lu Y. Zhang Q. Guan D. Gong P. Kammen D.M. He K. Schellnhuber H.J. Near-real-time monitoring of global CO2 emissions reveals the effects of the COVID-19 pandemic Nat. Commun. 11 1 2020 5172 33057164
Luna-Nemecio J. Tobón S. Juárez-Hernández L.G. Sustainability-based on socioformation and complex thought or sustainable social development Resour. Environ. Sustain. 2 2020 100007
Magudulela N. Lockdown Blamed For Contraction in Mining, Manufacturing Sector 2020 https://www.sabcnews.com/sabcnews/lockdown-blamed-for-contraction-in-mining-manufacturing-sector/
NBS National Database 2021 http://data.stats.gov.cn/index.html
Northey S. Haque N. Mudd G. Using sustainability reporting to assess the environmental footprint of copper mining J. Clean Prod. 40 2013 118 128
Nuss P. Ohno H. Chen W.-Q. Graedel T.E. Comparative analysis of metals use in the United States economy Resour. Conserv. Recycl. 145 2019 448 456
OECD Assessing Environmental Impact of Measures in the OECD Green Recovery Database 2022 https://www.oecd.org/coronavirus/policy-responses/assessing-environmental-impact-of-measures-in-the-oecd-green-recovery-database-3f7e2670/
Okuyama Y. Santos J.R. Disaster impact and input–output analysis Econ. Syst. Res. 26 1 2014 1 12
Pifarré i Arolas H. Acosta E. López-Casasnovas G. Lo A. Nicodemo C. Riffe T. Myrskylä M. Years of life lost to COVID-19 in 81 countries Sci. Reports 11 1 2021 3504
Qi J. Wang Y. Liang S. Li Y. Li Y. Feng C. Xu L. Wang S. Chen L. Wang D. Yang Z. Primary suppliers driving atmospheric mercury emissions through global supply chains One Earth 1 2 2019 254 266
Qu S. Liang S. Konar M. Zhu Z. Chiu A.S.F. Jia X. Xu M. Virtual water scarcity risk to the global trade system Environ. Sci. Technol. 52 2 2018 673 683 29231718
Rocco M.V. Guevara Z. Heun M.K. Assessing energy and economic impacts of large-scale policy shocks based on input-output analysis: application to Brexit Appl. Energy 274 2020 115300
Rose A. Lim D. Business interruption losses from natural hazards: conceptual and methodological issues in the case of the Northridge earthquake Glob. Environ. Change Part B: Environ. Hazards 4 1 2002 1 14
S&P COVID-19 Mining Impacts — Mining Projects With At-Risk Production 2020 https://www.spglobal.com/marketintelligence/en/news-insights/blog/covid19-mining-impacts-mining-projects-with-at-risk-production
S&P S&P Global Market Intelligence 2021 https://www.spglobal.com/
Shan Y. Ou J. Wang D. Zeng Z. Zhang S. Guan D. Hubacek K. Impacts of COVID-19 and fiscal stimuli on global emissions and the Paris agreement Nat. Clim. Change 2020
Shao S. Wang C. Feng K. Guo Y. Feng F. Shan Y. Meng J. Chen S. How do China's lockdown and post-COVID-19 stimuli impact carbon emissions and economic output? Retrospective estimates and prospective trajectories iScience 2022 104328
Sonter L.J. Dade M.C. Watson J.E.M. Valenta R.K. Renewable energy production will exacerbate mining threats to biodiversity Nat. Commun. 11 1 2020 4174 32873789
Stadler K. Wood R. Bulavskaya T. Södersten C.-J. Simas M. Schmidt S. Usubiaga A. Acosta-Fernández J. Kuenen J. Bruckner M. Giljum S. Lutter S. Merciai S. Schmidt J.H. Theurl M.C. Plutzar C. Kastner T. Eisenmenger N. Erb K.-H. de Koning A. Tukker A. EXIOBASE 3: developing a time series of detailed environmentally extended multi-regional input-output tables J. Ind. Ecol. 22 3 2018 502 515
Stadler K. Wood R. Bulavskaya T. Södersten C.-J. Simas Moana Schmidt Sarah Tukker Arnold EXIOBASE 3 (Version 3.8) [Data set] 2020 Zenodo
UK E.R. Euro to US Dollar Spot Exchange Rates for 2015 2021 https://www.exchangerates.org.uk/EUR-USD-spot-exchange-rates-history-2015.html
USGS Mineral Commodity Summaries 2021 2021 Mineral Commodity Summaries Reston, VA 200
Verones F. Hellweg S. Antón A. Azevedo L.B. Chaudhary A. Cosme N. Cucurachi S. de Baan L. Dong Y. Fantke P. Golsteijn L. Hauschild M. Heijungs R. Jolliet O. Juraske R. Larsen H. Laurent A. Mutel C.L. Margni M. Núñez M. Owsianiak M. Pfister S. Ponsioen T. Preiss P. Rosenbaum R.K. Roy P.-O. Sala S. Steinmann Z. van Zelm R. Van Dingenen R. Vieira M. Huijbregts M.A.J. LC-IMPACT: a regionalized life cycle damage assessment method J. Ind. Ecol. 24 6 2020 1201 1219
Verones F. Hellweg S. Azevedo L.B. Chaudhary A. Cosme N. Fantke P. Goedkoop M. Hauschild M. Laurent A. Mutel C. Pfister S. Ponsioen T. Steinmann Z. Zelm R.v. Vieira M. Huijbregts M.A.J. LC-Impact Version 0.5: A Spatially Differentiated Life Cycle Impact Assessment 2020 https://lc-impact.eu/doc/LC-Impact_report_SEPT2016_20160927.pdf
Wang H. Li X. Tian X. Ma L. Wang G. Wang X. Wang Z. Wang J. Yue Q. Socioeconomic drivers of China's resource efficiency improvement: A structural decomposition analysis for 1997–2017 Resour. Conserv. Recycl. 178 2022 106028
Wang H. Schandl H. Wang X. Ma F. Yue Q. Wang G. Wang Y. Wei Y. Zhang Z. Zheng R. Measuring progress of China's circular economy Resour. Conserv. Recycl. 2020 163
Wang H. Wang G. Qi J. Schandl H. Li Y. Feng C. Yang X. Wang Y. Wang X. Liang S. Scarcity-weighted fossil fuel footprint of China at the provincial level Appl. Energy 258 2020 114081
Wang H. Wei Y. Zhao S. Liu G.-X. Ma F. Wang G.-Q. Wang Y. Wang X.-Z. Yang D. Liu J.-R. Wang H.-T. Shi F. Chen W.-Q. Temporal and spatial variation in the environmental impacts of China's resource extraction at the provincial scale Ecosyst. Health and Sustain. 6 1 2020
Wang J. Xu X. Wang S. He S. He P. Heterogeneous effects of COVID-19 lockdown measures on air quality in Northern China Appl. Energy 282 2021 116179
Wang P. Wang H. Chen W.-Q. Pauliuk S. Carbon neutrality needs a circular metal-energy nexus Fundam. Res. 2 3 2022 392 395
Watari T. Nansai K. Giurco D. Nakajima K. McLellan B. Helbig C. Global metal use targets in line with climate goals Environ. Sci. Technol. 54 19 2020 12476 12483 32915547
Wiedmann T.O. Schandl H. Lenzen M. Moran D. Suh S. West J. Kanemoto K. The material footprint of nations Proc. Natl. Acad. Sci. U S A 112 20 2015 6271 6276 24003158
World Bank, 2022. https://data.worldbank.org/indicator/SP.DYN.LE00.IN.
Writer M.C.S. Copper Price Surges After Chile Closes Its Borders 2021 https://www.mining.com/copper-price-jumps-as-chile-close-its-borders/
WTO World Trade Report 2010 2010 https://www.wto.org/english/res_e/booksp_e/anrep_e/world_trade_report10_e.pdf
Wu L. Huang K. Ren Y. Yu Y. Huang B. Toward a better understanding of virtual water trade: Comparing the volumetric and impact-oriented virtual water transfers in China Resour. Conserv. Recycl. 186 2022 106573
Zeng Z. Guan D. Methodology and application of flood footprint accounting in a hypothetical multiple two-flood event Philos. Trans. R. Soc. A-Math. Phys. Eng. Sci. 378 2168 2020 24
Zhao S. Wang H.-M. Chen W.-Q. Yang D. Liu J.-R. Shi F. Environmental impacts of domestic resource extraction in China Ecosyst. Health and Sustain. 5 1 2019 67 78
Zhu Y. Ali S.H. Xu D. Cheng J. Mineral supply challenges during the COVID-19 pandemic suggest need for international supply security mechanism Resour. Conserv. Recycl. 165 2021 105231
| 36465718 | PMC9705201 | NO-CC CODE | 2022-12-05 23:15:18 | no | Resour Conserv Recycl. 2023 Mar 29; 190:106800 | utf-8 | Resour Conserv Recycl | 2,022 | 10.1016/j.resconrec.2022.106800 | oa_other |
==== Front
Vaccine
Vaccine
Vaccine
0264-410X
1873-2518
Elsevier Ltd.
S0264-410X(22)01484-0
10.1016/j.vaccine.2022.11.060
Article
Education level modifies parental hesitancy about COVID-19 vaccinations for their children
Tang Shuning a
Liu Xin a
Jia Yingnan ab
Chen Hao a
Zheng Pinpin ab
Fu Hua ab
Xiao Qianyi ab⁎
a Department of Preventive Medicine and Health Education, School of Public Health, Fudan University, China
b School of Public Health, Key Laboratory of Public Health Safety, Fudan University, Ministry of Education, Shanghai, China
⁎ Corresponding author at: Department of Preventive Medicine and Health Education, School of Public Health, Fudan University, Shanghai 200032, China; School of Public Health, Key Laboratory of Public Health Safety, Fudan University, Ministry of Education, Shanghai 200032, China.
29 11 2022
29 11 2022
20 6 2022
19 11 2022
22 11 2022
© 2022 Elsevier Ltd. All rights reserved.
2022
Elsevier Ltd
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
It is important to encourage parental acceptance of children’s vaccination against COVID-19 to ensure population immunity and mitigate morbidity and mortality. This study drew upon protection motivation theory (PMT) to explore the factors of parental hesitancy about vaccinating their children. A national online survey was performed in China. A total of 2054 Chinese parents of children aged 6–12 years were included in this study. They reported on measures that assessed hesitancy about children’s vaccination against COVID-19, PMT constructs (susceptibility, severity, response efficacy, self-efficacy, and response costs) and sociodemographic characteristics. Chinese parents reported a hesitancy rate of 29.4% for children’s vaccination. Parents with higher level education were more likely to hesitate to vaccinate their children against COVID-19. After controlling for parents’ and children’s demographic variables, logistic regression showed that parents’ hesitancy about their children’s vaccination increased if parents had lower levels of susceptibility, response efficacy or self-efficacy, as well as higher levels of response costs. In addition, a high educational level can significantly increase the promotive effect of response cost and the protective effect of response efficacy on vaccine hesitancy. In conclusion, our findings suggested that PMT can explain parents' vaccine hesitancy and that education level can modify the effect of copying appraisal, but not threat appraisal, on parental hesitancy. This study will help public health officials send targeted messages to parents to improve the rate of COVID-19 vaccination in children aged 6–12 years and thus reach a higher level of immunity in the population.
Keywords
COVID-19
Vaccine hesitancy
Protection motivation theory
Children
Parents
==== Body
pmc1 Introduction
The COVID-19 pandemic has caused >638 million infections and >6.5 million deaths as of Nov 2022. The disease affects large numbers of people of all age groups worldwide. Vaccination is one of the most cost-effective public health intervention strategies in limiting the spread of the infectious disease [1], [2]. It is important for government and public health officials to encourage acceptance and uptake of the vaccine to ensure population immunity and mitigation of morbidity and mortality. In this context, after the COVID-19 vaccine was approved for use in children, overcoming barriers to vaccinating children became crucial.
Vaccination reduces infection risk of infectious disease in healthy children from 30% to 11% [3] and controls virus transmission [4], [5]. With the emergence of new variants, the risk of disease transmission and outcomes in children requires close surveillance [6], [7]. Despite the importance of vaccination, in our previous study, the prevalence of vaccine hesitancy in Chinese adults was 44.3% after the emergency use authorization of COVID-19 vaccine for adults [8]. Vaccine hesitancy refers to a delay in the acceptance or refusal of vaccination despite adequate access and availability [9]. Since parents are often key decision-makers for whether their children will receive vaccinations, it is important to measure vaccine confidence among parents of young children and to investigate the factors of parental hesitancy about children’s vaccination.
In this study, parental hesitancy about vaccinating children can be regarded as a health-related behaviour. People usually use information to evaluate the threat of diseases and the efficacy of responses before making behavioural decisions and taking action. Protection motivation theory (PMT) is an important theoretical framework to predict an individual’s health behaviour, including threat appraisal and coping appraisal. Threat appraisal includes perceived susceptibility and perceived severity of the threat [10]. The coping appraisal process includes efficacy appraisal (response efficacy and self-efficacy) and response costs [11]. The hypothesis of this theory is that a high threat appraisal and high efficacy appraisal will increase the probability of health action and reduce the probability of undesirable behaviour, while a high response cost will reduce the probability of health behaviour. The PMT constructs have been successfully applied to understand and predict changes in the health behaviours associated with severe acute respiratory syndrome (SARS) [12], influenza A H1N1 [13], [14], hepatitis B [15] and COVID-19 [16]. In the context of parental hesitancy about vaccinating their children, susceptibility refers to parents’ perception of whether their children are vulnerable to COVID-19 infection, and severity refers to the damage that COVID-19 may cause to their children’s health. Response efficacy is the belief that the COVID-19 vaccine will be beneficial to their children. Self-efficacy is the belief that parents themselves have the ability to have their children vaccinated. Response costs refer to the ineffectiveness and the side effects of the COVID-19 vaccine, such as pain and swelling of the injection site and headache.
PMT is an important theoretical framework to explain parental hesitancy about vaccinating their children. Previous studies have used the Health Belief Model (HBM) and Theory of Planned Behavior (TPB) to explain parents’ COVID-19 vaccine hesitancy or intention [17], but with no relevant evidence for PMT. In addition, some demographic factors were found to be associated with vaccine hesitancy and may influence their threat appraisal and coping appraisal [18], [19], [20]. This study, conducted during a period in which COVID-19 vaccines were not authorized for nationwide use in children aged 6–12 in China, examined the factors related to parents’ hesitancy based on the PMT model and explored the influence of key demographic factors on the effect of the PMT construct on parental hesitancy about children’s COVID-19 vaccination. This study will help public health officials send targeted messages to parents to improve the COVID-19 vaccination rate in children aged 6–12 years and thus reach a higher level of immunity in the population.
2 Method
2.1 Participants and procedures
A cross-sectional, anonymous online survey was performed among Chinese residents. The survey was made available on the Wenjuanxing platform from October 19 to October 28, 2021, the period in which COVID-19 vaccines were not yet authorized for nationwide use in children aged 6–12 in China. Parents who had children aged 6–12 were invited to participate in the survey. If the parents had more than one eligible child, information was requested for the child whose birthday was closest to the survey date, to avoid confusion and inconsistency on the survey. A convenience sampling strategy was utilized. The online questionnaire link was disseminated via websites and WeChat, which were public websites that could be shared with family members, friends, and colleagues and forwarded to others. In addition, the online questionnaire link was sent to several primary schools that cooperated with us, including four schools in North China and two schools in South China. These schools were invited to share the questionnaire link with the parents of their students. Informed consent was provided on the first page of the survey. The survey took approximately 3–5 min on average, and questionnaires that were completed in <100 s were considered invalid, reflecting a careless response. A quality control item with a required answer was also set to avoid the return of invalid questionnaires. Prior approval by the Human Research Ethics Committee at Fudan University was obtained.
2.2 Measures
The questionnaires requested sociodemographic characteristics of parents and their children, parental hesitancy about vaccinating their children against COVID-19, and PMT constructs.
Outcome measures: Parental hesitancy about vaccinating their children against COVID-19 was measured with a single item, “How willing would you be to vaccinate your child aged 6–12 y against COVID-19?”, which was measured on a seven-point scale (from “1 = refuse all” to “7 = accept all”). In this study, parental hesitancy about vaccinating their children was regarded as any response on the scale except for “accept all” or “accept but unsure.”.
Sociodemographic characteristics: This study recorded the sociodemographic characteristics of parents (e.g., age, gender, region, marital status, educational level, occupation and annual family income per capita) and children’s gender. In addition, participants were asked to rate their children's overall health on five semantic differential scales (from “1 = very good” to “5 = bad”). Participants with answers of “very good” and “good” were categorized as overall “healthy”. Proportion of COVID-19 vaccination among family members over 12 years old was assessed on five semantic differential scales (from “1 = all” to “5 = no one”).
PMT constructs: Five PMT constructs, namely, susceptibility, severity, response efficacy, response cost, and self-efficacy, were each measured using several self-devised items. A five-point Likert scale was used to measure all of the items related to the PMT constructs, with a range from “1 = strongly disagree” to “5 = strongly agree”. The questions related to PMT constructs as well as the reliability and validity evaluation results of the questionnaire are shown in Table 1 . The Cronbach’s alpha was above 0.800, indicating good reliability. The factor loading of most questions was higher than 0.700, and the validity of the questions was good. The KMO value was equal to 0.892, and the structural validity of the questionnaire was good.Table 1 Items used to assess the constructs and the factor analysis results of the PMT factors of COVID-19 vaccine hesitancy.
Constructs Assignment and Variable Processing Questions Factor Lading Cronbach's α
Perceived susceptibility 1 = Strongly disagree, 2 = Disagree, 3 = Neutral, 4 = Agree, 5 = Strongly agree. The median of respondents’ averaged index (median = 3.00) was used for binary categorical classification (high/low level). a. I think my children are more likely to be infected with the COVID-19 virus. 0.827 0.814
b. I think COVID-19 infection is likely in children 6–12 years old because of a lack of health consciousness and protective behaviour. 0.666 0.808
Perceived severity 1 = Strongly disagree, 2 = Disagree, 3 = Neutral, 4 = Agree, 5 = Strongly agree. The median of respondents’ averaged index (median = 5.00) was used for binary categorical classification (high/low level). a. If the child is infected with COVID-19, it will have a serious impact on the family. 0.866 0.801
b. If the child is infected with COVID-19, it will have a serious impact on the child's school and life. 0.867 0.801
c. If a child is infected with COVID-19, there is a risk of severe illness. 0.800 0.801
d. Infection with the novel coronavirus pneumonia is harmful to children's health. 0.756 0.804
Response efficacy 1 = Strongly disagree, 2 = Disagree, 3 = Neutral, 4 = Agree, 5 = Strongly agree. The median of respondents’ averaged index (median = 4.33) was used for binary categorical classification (high/low level). a. The COVID-19 vaccine for children helps to establish a universal immune barrier. 0.668 0.803
b. Vaccine can effectively prevent COVID-19 infection. 0.585 0.805
c. After children are vaccinated for COVID-19, their daily lives and travel will be more convenient. 0.581 0.804
Response cost 1 = Strongly disagree, 2 = Disagree, 3 = Neutral, 4 = Agree, 5 = Strongly agree. The median of respondents’ averaged index (median = 3.00) was used for binary categorical classification (high/low level). a. I am worried about COVID-19 vaccine adverse reactions. 0.836 0.814
b. I am worried about the long-term adverse effects of the vaccine on the body. 0.807 0.811
c. I'm worried that the effective time of prevention is not long enough. 0.768 0.807
d. If children receive the COVID-19 vaccine, the adverse reactions could be serious. 0.758 0.814
e. I think the vaccine is not effective in preventing COVID-19. 0.768 0.820
f. I don't think the COVID-19 vaccine can protect against a mutant strain. 0.738 0.819
Self-efficacy 1 = Strongly disagree, 2 = Disagree, 3 = Neutral, 4 = Agree, 5 = Strongly agree. The median of respondents’ averaged index (median = 3.50) was used for binary categorical classification (high/low level). a. Even if there are adverse reactions, I believe it will not cause long-term damage to health. 0.755 0.811
b. If adverse reactions occur after vaccination, I believe they can be handled in time. 0.803 0.806
c. I think the possibility of adverse reactions after vaccination is low. 0.790 0.806
d. If I want to vaccinate my child, I know where and how to get vaccinated successfully. 0.697 0.808
2.3 Statistical analyses
First, the frequencies and proportions were calculated for the sociodemographic characteristics to capture the tendencies of parental hesitancy about vaccinating their children. Only significant sociodemographic factors, with p values of <0.05, were selected for the subsequent multivariate logistic regression analysis. Second, Kaiser–Meyer–Olkin (KMO) was calculated to assess validity and overall construct validity for each question. The factor loading of each item in PMT was assessed by using exploratory factor analysis (EFA). EFA utilized a principal component analysis framework with varimax rotation, which was conducted for each item in PMT, using Cronbach’s alpha to estimate the internal reliability of the items for PMT construct measures. The median scores of the five constructs in PMT were calculated separately. If the score was less than the median, it was defined as “low level”; otherwise, it was defined as “high level” and converted to binary variables. Next, logistic regression analyses were applied to identify the predictors of parental hesitancy based on PMT constructs, adjusting for age, education level, occupation, annual family income per capita, parent-rated children’s health, proportion of vaccinated family members, and PMT constructs. Odds ratios (ORs) and 95% confidence intervals (95% CIs) were used to quantify the effects. Finally, Pearson’s bivariate correlations and Cochran–Mantel–Haenszel (CMH) analysis were applied to study the correlation between PMT and vaccine hesitancy after considering the modification effect of educational level. All analyses were carried out using SPSS 25.0. All tests were two-tailed with a significance level of P < 0.05.
3 Results
3.1 Sociodemographic characteristics
A total of 2199 participants from 29 provinces and autonomous regions (China consists of a total of 34 provinces and autonomous regions) completed the survey. A total of 145 questionnaires were excluded from the analysis due to the following exclusion criteria: (1) children with contraindications to vaccines (n = 47), (2) participants who answered quality control questions incorrectly (n = 75), (3) participants who were abroad (n = 3), (4) participants whose completion time was <100 s (n = 13), and (5) participants who have the same answer to all questions (n = 7). Finally, 2054 (93.4%) valid questionnaires were included in the following analyses.
As shown in Table 2 , the results of descriptive analyses showed that participants reported a hesitancy rate of 29.4% (n = 603). Most respondents were 31–40 years old (60.4%), women (70.4%). In a multivariate logistic regression analysis, the factors associated with hesitancy to vaccinate children were parents who had a bachelor’s degree or above, front-line workers in health care, had a family income per capita of more than ¥120,000, had a lower proportion of vaccinated family members, or rated their children’s health as poor (P < 0.05). Parents over 40 years old were less likely to be hesitant to vaccinate their children (P < 0.05).Table 2 Distribution of vaccine hesitancy by participant demographics and health-related characteristics.
Demographics Overall (n = 2054) Vaccine hesitancy rate (n = 603, 29.4 %) aOR (95%CI) Pvalue
n (%) n (%)
Parents' information
Age (years)
<30 385 (18.7) 117 (30.4) 1
31–40 1240 (60.4) 375 (30.2) 0.83 (0.63, 1.08) 0.160
>40 429 (20.9) 111 (25.9) 0.71 (0.51, 0.99) 0.042
Gender
Men 607 (29.6) 165 (27.2) 1
Women 1447 (70.4) 438 (30.3) 1.04 (0.83, 1.30) 0.743
Ethnic
Han 1927 (93.8) 572 (29.7) 1
Other 127 (6.2) 31 (24.4) 0.86 (0.55, 1.35) 0.522
Marital status
Married 1892 (92.1) 563 (29.8) 1
Not married 162 (7.9) 40 (24.7) 0.85 (0.58, 1.26) 0.427
Region
Urban 1357 (66.1) 426 (31.4) 1
Town 492 (23.0) 129 (26.2) 0.98 (0.76, 1.26) 0.876
Rural 205 (10.0) 48 (23.4) 0.91 (0.62, 1.31) 0.601
Educational level
Middle school degree and below 722 (35.2) 156 (21.6) 1
High school degree 602 (29.3) 153 (25.4) 1.23 (0.94, 1.61) 0.135
Bachelor degree and above 730 (35.5) 294 (40.3) 1.98 (1.48, 2.64) <0.001
Risk of infection in occupation
High risk 188 (9.2) 58 (30.9) 1
Low risk 1866 (90.8) 545 (29.2) 1.14 (0.78, 1.64) 0.504
Occupation
Non-medical related workers 1908 (92.9) 543 (28.5) 1
Front-line workers in health care 64 (3.1) 32 (50.0) 2.20 (1.25, 3.88) 0.006
Other relevant workers in health care 82 (4.0) 28 (34.1) 1.05 (0.64, 1.75) 0.839
Family income per capita (RMB)
<30,000 662 (32.2) 158 (23.9) 1
30,000–59,999 503 (24.5) 122 (24.3) 0.87 (0.65, 1.16) 0.346
60,000–120,000 480 (23.4) 143 (29.8) 1.02 (0.77, 1.37) 0.872
>120,000 409 (19.9) 180 (44.0) 1.48 (1.07, 2.06) 0.02
Proportion of COVID −19 vaccination among family members > 12 years old
All 1739 (84.7) 469 (27.0) 1
Most 201 (9.8) 85 (42.3) 1.78 (1.30, 2.43) <0.001
Half and below 114 (5.5) 49 (43.0) 1.97 (1.31, 2.95) 0.001
Children's information
Gender
Men 1112 (54.1) 316 (28.4) 1
Women 942 (45.9) 287 (30.5) 1.08 (0.88, 1.32) 0.463
Parents' self-assessment of children's overall health
Health 1953 (95.1) 551 (28.2) 1
Poor/bad 101 (4.9) 52 (51.5) 2.86 (1.87, 4.36) <0.001
a OR: odds ratio.
3.2 Predictors of parents’ hesitancy about children’s vaccination based on PMT constructs and vaccination experiences
Logistic regressions were run with “vaccine hesitant” or “non-vaccine hesitant” as the outcome to explore the predictors of parents’ hesitancy. Adjusted variables in parental hesitancy logistic regression included age, education level, occupation, family income per capita, parent-rated children’s health, proportion of vaccinated family members, and each PMT construct. As shown in Table 3 , a high level of perceived susceptibility, response efficacy and self-efficacy (P < 0.001) significantly decreased parents’ hesitancy about vaccinating their children. However, parents with a high level of response cost were three times more likely to be hesitant about vaccinating their children than those with a low level of response cost (P < 0.001).Table 3 Logistic regression analysis of PMT constructs.
Constructs Vaccine hesitancy n (%) a OR (95%CI) P value
Perceived susceptibility
Low 310 (43.2) 1.00
High 293 (21.9) 0.55 (0.43, 0.69) <0.001
Perceived severity
Low 399 (39.2) 1.00
High 204 (19.7) 0.84 (0.65, 1.08) 0.180
Response efficacy
Low 481 (50.2) 1.00
High 122 (11.1) 0.25 (0.19, 0.33) <0.001
Self-efficacy
Low 404 (43.3) 1.00
High 199 (17.8) 0.58 (0.45, 0.75) <0.001
Response cost
Low 149 (16.2) 1.00
High 454 (40.1) 2.94 (2.30, 3.76) <0.001
Multivariate logistic regression for psychosocial factors predicting the hesitancy to receive the COVID-19 vaccine.
a OR was adjusted for age, education, occupation, family income per capita, proportion of vaccinated family members >12 years old, parents' self-assessment of children's overall health, and other PMT constructs.
3.3 Association between PMT constructs and parental hesitancy among different education attainment layers
Considering that education level is a significant demographic predictor for parental hesitancy, we further explored the influence of level of education on the association between PMT constructs and parental hesitancy. Table 4 presents the correlations between COVID-19 vaccine hesitancy and parents’ education level and each PMT construct. A high education level was related to a low level of response efficacy and self-efficacy and a high level of response cost. As shown in Fig. 1 and Supplementary Table 1, the impact of education level on the relationship between response cost and parental hesitancy (P = 0.001), as well as between response efficacy and parental hesitancy (P = 0.010), is different. As parents' education level increased, their response cost played a greater role in promoting parental hesitancy (Fig. 1 A, Supplementary Table 1) with an OR of 1.87 (95% CI 1.30–2.68) for individuals with a middle school education or less, an OR of 3.54 (95% CI 2.34–5.35) for those with a high school diploma, and an OR of 5.05 (95% CI 3.52–7.23) for those with a bachelor’s degree or more. With the increase in parents' education level, their response efficacy will have a less inhibitory effect on parental hesitancy (Fig. 1 B, Supplementary Table 1), with an OR of 0.22 (95% CI 0.15–0.32) in the middle school education or less group, an OR of 0.10 (95% CI 0.07–0.16) in the high school diploma group, and an OR of 0.10 (95% CI 0.06–0.15) in the bachelor’s degree or more group. However, the impact of self-efficacy on parental hesitancy between education levels was homogeneous (P > 0.05, Supplementary Table 1). These findings suggest that education levels modify the impact of coping appraisal (response cost and response efficacy) on parental hesitancy.Table 4 Correlations between COVID-19 vaccine hesitancy and parental education level and PMT constructs.
1 2 3 4 5 6
1. vaccine hesitancy –
2. education level 0.172** –
PMT constructs
3. perceived susceptibility −0.222** 0.020 –
4. perceived severity −0.215** −0.010 0.232** –
5. response efficacy −0.428** −0.121** 0.274** 0.433** –
6. self-efficacy −0.279** −0.075** 0.208** 0.302** 0.483** –
7. response cost 0.262** 0.133** −0.013 0.022 −0.175** 0.023
* P < 0.05.
** P < 0.01.
Fig. 1 Association between vaccine hesitancy and PMT constructs in CMH analyses.
4 Discussion
Our findings indicate that a high education level and family income, a low proportion of COVID-19 among family members, and parent-assessment of their children's poor health were associated with high vaccine hesitancy. Furthermore, a low level of perceived susceptibility, response efficacy and self-efficacy and a high level of response cost were associated with high COVID-19 vaccine hesitancy. Moreover, education levels modified the impact of response cost and response efficacy on parental hesitancy.
Although there are many studies on parents’ attitudes towards childhood vaccination [21], [22], we first reported that education level could modify the effect of coping appraisal, but not threat appraisal, on parental hesitancy. Based on our findings, parents who have higher levels of education are more likely to hesitate to vaccinate their children against COVID-19, and a high level of education can significantly increase the promotive effect of response cost and the protective effect of response efficacy on vaccine hesitancy. The educational level and the degree of reluctance to vaccination continues to be controversial. Previous studies have demonstrated the general argument that COVID-19 vaccine hesitancy rates are higher among parents with lower educational attainment [23], [24], [25], [26]. For example, US parents who had a bachelor's degree or higher education had already received or were likely to receive a COVID-19 vaccine for their children [25], [26]. In addition, the same phenomenon has been found among Italy parents/guardians of children aged <18 years old [23] and Canada parents of children aged ≤12 years old [24]. However, in other studies carried out in Saudi Arabia and Turkey, one of the factors associated with lower intention to vaccinate children was parents with higher education levels [27], [28]. Here our team also identified a positive association between education level and vaccine hesitancy both among Chinese parents for their children’s vaccination (this study) and among Chinese adults [8], [16]. The possible explanation for our findings may be that parents with higher educational levels were more likely to have higher social status and therefore may have more channels to learn about the effects and side effects of vaccines. As we are discovering, the higher the education level of parents was, the greater the promotion effect of response cost assessment on vaccine hesitancy and the stronger the protective effect of response efficacy assessment on vaccine hesitancy. These findings point out that interventions targeting response cost and efficacy are especially important for parents with high education levels to enhance the children’s vaccination against to COVID-19.
As expected, coping appraisal, including response efficacy, response cost and self-efficacy, was crucial for predicting parental hesitancy about children’s vaccination. In our study, response efficacy includes agreements about the contribution of children’s vaccination to the prevention of COVID-19 infection, to the convenience of children’s daily lives and travel, and to establishing immunity. To a certain extent, response efficacy can reflect parents' evaluation of the effectiveness of the vaccine, their hope to restore normal life and social life, and their perception of the externality and altruism of the vaccine. Consistent with our results, response efficacy was also identified as an important predictor of adults’ willingness to vaccinate themselves against COVID-19 [29] and seasonal influenza [30] or parents’ willingness to vaccine their children against measles [31] and HPV [32]. When parents perceive higher response costs, such as concerns about side effects after vaccination and long-term adverse reactions, they are more likely to be vaccine hesitant, which is detrimental to the implementation of vaccine programs. Studies from China [33], Turkey [27] and a global study from six countries [34] all reported that concern about the safety and side effects of vaccination is one of the reasons why parents do not vaccinate their children against COVID-19. Except for awareness of vaccine safety and side effects, response cost also included concerns about the effectiveness of the vaccine. If parents believe that the vaccine cannot protect against the mutant strain, they believe it is useless to vaccinate their children. Lack of knowledge about vaccine effectiveness is one of the most common reasons for parents in Shanghai, China to refuse COVID-19 vaccination [35]. Vaccine effectiveness and safety have been reported to be important factors predicting parental COVID-19 vaccine hesitancy or vaccine willingness in children under 18 years of age [21], [36], [37], [38], and here, we provide evidence about parental attitudes towards the COVID-19 vaccine for their children aged 6–12.
We also found that low perceived susceptibility was related to parental vaccine hesitancy, probably because perception of the risk of COVID-19 disease may affect parents' decision-making [5]. Moreover, low self-efficacy indicates an increased possibility of vaccine hesitancy. Several studies have shown that self-efficacy is a key factor affecting COVID-19 vaccination willingness and predicting adults’ or parents’ vaccination behaviour in China [8], [16], [17]. Studies of H1N1 vaccines also suggest that the public's self-efficacy of H1N1 vaccination can be boosted by increasing the benefits of vaccination [39]. In this study, severity was considered to have little to do with vaccine hesitancy, which is similar to the results of some studies [16], [40], probably because parents will first consider whether their child is susceptible to an illness before considering how severe the illness may be [40].
Other demographic factors should also be noted. Young parents showed more COVID-19 vaccine hesitancy, which is consistent with another study in which the highest vaccine hesitancy rates were detected in parents ≤29 years old [23]. Parents who assess their children as unhealthy are more likely to hesitate to vaccinate their children against COVID-19, and the same situation was also found among adults [16]. In our study, health care workers showed higher vaccine hesitancy about children, which is consistent with previous findings that health care workers showed low acceptability of COVID-19 vaccination despite their important roles in vaccination promotion [33]. According to previous studies, there are three possible reasons. First, concerns about the expedited development of COVID-19 vaccines have led to vaccination hesitancy [41]. Second, they might be more aware that the risk of death caused by COVID-19 is low among children, and most infected children would not be symptomatic [42]. Third, health care workers have growing scepticism about the safety and effectiveness of COVID-19 vaccines [43], [44]. In addition, the lower the proportion of vaccinated family members was, the stronger the parental vaccine hesitancy for children. The lower proportion of vaccinated family members may reflect a history of unwillingness to receive vaccines. A study from China Hong Kong found that parents’ histories of receiving COVID-19 vaccines themselves were significantly related to parents’ intentions [17]. It was also suggested that previous vaccination experience might have an impact on current willingness to vaccinate [5]. Considering the hysteresis for vaccination among children [45], the important role of children in transmitting COVID-19, and the side effects and efficacy of vaccines [46], health educators have been encouraging child vaccination to reduce school and community transmission. We suggest that health educators should focus on the knowledge about safety and efficacy of vaccines among parents with medium to high levels of education to reduce their concerns about vaccinating their children.
5 Limitations
There are several limitations of the present study that should be noted in interpreting the results. First, the sampling process of the online survey may result in selection bias. The survey link disseminated by WeChat can only reach network members, which limits the generalizability of the results. Second, the assessment of parental hesitancy about children’s vaccination occurred during a short time; thus, the results may not reflect the long-term effect of these identified influencing factors because the hesitancy to obtain the vaccine will decrease with the increase in the number of vaccinated children in the future. Third, the cross-sectional design conducted to measure the exposure and outcome of parental vaccine hesitancy simultaneously only measured the situation of a certain population at a certain point in time. Thus, causality cannot be proved; only possible factors for causality can be provided. As mentioned above, parental hesitancy about children’s vaccination against COVID-19 may change as the pandemic evolves and more information about vaccines is released. Despite these limitations, our findings contribute insights into targeted interventions aimed at reducing parental hesitancy for children’s vaccination.
6 Conclusion
Health education to parents is warranted when the COVID-19 vaccine is available and authorized for use in children. This study’s findings suggested that the PMT model can be used to develop strategies for reducing parental hesitancy about children’s vaccination. Interventions targeting response costs, i.e., worries about COVID-19 vaccine adverse reactions, short effective time of prevention and useless resistance to the mutant strain, are crucial. Similarly, interventions targeting response efficacy, i.e., agreements in the contribution of children’s vaccination to the prevention of COVID-19 infection, convenience of children’s daily travel and life and establishing children’s immunity, are also important. Importantly, all of these interventions targeting response cost and efficacy are especially important for parents with high education levels. Perceived susceptibility could decrease parental hesitancy, which suggests that the government should publicize children's high risk of infection, emphasizing the current popularization of vaccination among adults and the lack of prevention awareness of children. Self-efficacy was negatively related to parental hesitancy, providing guidance for policy recommendations for enhancing education about vaccination knowledge for parents, such as vaccination locations and treatments of side effects.
Funding sources
This research was funded by Healthy Shanghai action special project of Shanghai Municipal Commission (JKSHZX-2022-01), and did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Data Availability Statement: The data that support the findings of this study are available from school of public health, Fudan University but restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly available. Data are however available from the authors upon reasonable request and with permission of school of public health, Fudan University.
Ethics approval and consent to participate: This study was approved by the Ethics Committee of Department of Public Health in Fudan University, Shanghai, China (IRB#2021-10-0932). All participants or their legally acceptable representatives provided written informed consent.
CRediT authorship contribution statement
Shuning Tang: Investigation, Data curation, Writing - original draft. Xin Liu: Investigation. Yingnan Jia: Investigation. Hao Chen: Investigation. Pinpin Zheng: Investigation, Writing - review & editing. Hua Fu: Writing - review & editing. Qianyi Xiao: Conceptualization, Investigation, Data curation, Writing - review & editing.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Appendix A Supplementary material
The following are the Supplementary data to this article:Supplementary data 1
Data availability
Data will be made available on request.
Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.vaccine.2022.11.060.
==== Refs
References
1 Dubé E. Laberge C. Guay M. Bramadat P. Roy R. Bettinger J. Vaccine hesitancy: an overview Hum Vaccin Immunother 9 2013 1763 1773 23584253
2 Rappuoli R. Mandl C.W. Black S. De Gregorio E. Vaccines for the twenty-first century society Nat Rev Immunol 11 2011 865 872 22051890
3 Jefferson T. Rivetti A. Di Pietrantonj C. Demicheli V. Vaccines for preventing influenza in healthy children Cochrane Database Syst Rev 2 2018 Cd004879 29388195
4 Grech V. Borg M. Influenza vaccination in the COVID-19 era Early Hum Dev 148 2020 105116
5 Goldman R.D. McGregor S. Marneni S.R. Katsuta T. Griffiths M.A. Hall J.E. Willingness to vaccinate children against influenza after the coronavirus disease 2019 pandemic J Pediatr 228 87–93 2021 e2
6 Bhopal S.S. Bagaria J. Olabi B. Bhopal R. Children and young people remain at low risk of COVID-19 mortality Lancet Child Adolesc Health 5 2021 e12 e13 33713603
7 Brookman S. Cook J. Zucherman M. Broughton S. Harman K. Gupta A. Effect of the new SARS-CoV-2 variant B.1.1.7 on children and young people. lancet Child Adolesc Health 5 2021 e9 e10
8 Chen H. Li X. Gao J. Liu X. Mao Y. Wang R. Health belief model perspective on the control of COVID-19 vaccine hesitancy and the promotion of vaccination in China: web-based cross-sectional study J Med Int Res 23 2021 e29329
9 MacDonald N.E. Vaccine hesitancy: Definition, scope and determinants Vaccine 33 2015 4161 4164 25896383
10 Milne SS, P., Orbell, S. Prediction and intervention in health-related behavior: a meta-analytic review of protection motivation theory, J. Appl. Psychol. 2000;30:106-43.
11 Floyd D.-L.-S.-P.-D. Rogers R. A meta-analysis of research on protection motivation theory J Appl Soc Psychol 30 2 2000 407 429
12 Jiang X. Elam G. Yuen C. Voeten H. de Zwart O. Veldhuijzen I. The perceived threat of SARS and its impact on precautionary actions and adverse consequences: a qualitative study among Chinese communities in the United Kingdom and the Netherlands Int J Behav Med 16 2009 58 67 19277874
13 de Zwart O. Veldhuijzen I.K. Richardus J.H. Brug J. Monitoring of risk perceptions and correlates of precautionary behaviour related to human avian influenza during 2006–2007 in the Netherlands: results of seven consecutive surveys BMC Infect Dis 10 2010 114 20462419
14 Sharifirad G. Yarmohammadi P. Sharifabad M.A. Rahaei Z. Determination of preventive behaviors for pandemic influenza A/H1N1 based on protection motivation theory among female high school students in Isfahan Iran J Educ Health Promot 3 2014 7 24741647
15 Liu R. Li Y. Wangen K.R. Maitland E. Nicholas S. Wang J. Analysis of hepatitis B vaccination behavior and vaccination willingness among migrant workers from rural China based on protection motivation theory Hum Vaccin Immunother 12 2016 1155 1163 27078191
16 Xiao Q. Liu X. Wang R. Mao Y. Chen H. Li X. Predictors of willingness to receive the COVID-19 vaccine after emergency use authorization: the role of coping appraisal Vaccines (Basel) 2021 9 35062670
17 Li J.B. Lau E.Y.H. Chan D.K.C. Why do Hong Kong parents have low intention to vaccinate their children against COVID-19? testing health belief model and theory of planned behavior in a large-scale survey Vaccine 40 2022 2772 2780 35339306
18 Bish A. Yardley L. Nicoll A. Michie S. Factors associated with uptake of vaccination against pandemic influenza: a systematic review Vaccine 29 2011 6472 6484 21756960
19 Lahiri A. Jha S.S. Chakraborty A. Dobe M. Dey A. Role of threat and coping appraisal in protection motivation for adoption of preventive behavior during COVID-19 pandemic Front Public Health 9 2021 678566
20 Nguyen T.P.L. Pattanrsi S. The influence of ethnic, social, and demographic factors on urban slum Dwellers' threat appraisal, awareness, and protective practices against COVID-19 in Thailand Am J Trop Med Hyg 2022
21 Skjefte M. Ngirbabul M. Akeju O. Escudero D. Hernandez-Diaz S. Wyszynski D.F. COVID-19 vaccine acceptance among pregnant women and mothers of young children: results of a survey in 16 countries Eur J Epidemiol 36 2021 197 211 33649879
22 Kempe A. Saville A.W. Albertin C. Zimet G. Breck A. Helmkamp L. Parental hesitancy about routine childhood and influenza vaccinations: a national survey Pediatrics 146 2020
23 Montalti M, Rallo F, Guaraldi F, Bartoli L, Po G, Stillo M, et al. Would parents get their children vaccinated against SARS-CoV-2? Rate and predictors of vaccine hesitancy according to a survey over 5000 families from Bologna, Italy. Vaccines (Basel). 2021;9.
24 Hetherington E. Edwards S.A. MacDonald S.E. Racine N. Madigan S. McDonald S. SARS-CoV-2 vaccination intentions among mothers of children aged 9 to 12 years: a survey of the All Our Families cohort CMAJ Open 9 2021 E548 E555
25 Teasdale C.A. Borrell L.N. Kimball S. Rinke M.L. Rane M. Fleary S.A. Plans to vaccinate children for coronavirus disease 2019: a survey of United States parents J Pediatr 237 2021 292 297 34284035
26 Szilagyi P.G. Shah M.D. Delgado J.R. Thomas K. Vizueta N. Cui Y. Parents' intentions and perceptions about COVID-19 vaccination for their children: results from a national survey Pediatrics 148 2021
27 Yigit M. Ozkaya-Parlakay A. Senel E. Evaluation of COVID-19 vaccine refusal in parents Pediatr Infect Dis J 40 2021 e134 e136 33410650
28 Temsah M.H. Alhuzaimi A.N. Aljamaan F. Bahkali F. Al-Eyadhy A. Alrabiaah A. Parental attitudes and hesitancy about COVID-19 vs. Routine childhood vaccinations: a national survey. front Public Health 9 2021 752323
29 Li L. Wang J. Nicholas S. Maitland E. Leng A. Liu R. The intention to receive the COVID-19 vaccine in china: insights from protection motivation theory Vaccines (Basel) 2021 9 35062670
30 Ling M. Kothe E.J. Mullan B.A. Predicting intention to receive a seasonal influenza vaccination using protection motivation theory Soc Sci Med 233 2019 87 92 31195194
31 Camerini AL, Diviani N, Fadda M, Schulz PJ. Using protection motivation theory to predict intention to adhere to official MMR vaccination recommendations in Switzerland. SSM Popul Health. 2019;7:005-5.
32 Gainforth H.L. Cao W. Latimer-Cheung A.E. Determinants of human papillomavirus (HPV) vaccination intent among three Canadian target groups J Cancer Educ 27 2012 717 724 22865323
33 Wang Z. She R. Chen X. Li L. Li L. Huang Z. Parental acceptability of COVID-19 vaccination for children under the age of 18 years among Chinese doctors and nurses: a cross-sectional online survey Hum Vaccin Immunother 17 2021 3322 3332 34137670
34 Goldman R.D. Yan T.D. Seiler M. Parra Cotanda C. Brown J.C. Klein E.J. Caregiver willingness to vaccinate their children against COVID-19: cross sectional survey Vaccine 38 2020 7668 7673 33071002
35 Lu J, Wen X, Guo Q, Ji M, Zhang F, Wagner AL, et al. Sensitivity to COVID-19 vaccine effectiveness and safety in Shanghai, China. Vaccines (Basel). 2021;9.
36 Ruggiero K.M. Wong J. Sweeney C.F. Avola A. Auger A. Macaluso M. Parents' intentions to vaccinate their children against COVID-19 J Pediatr Health Care 35 2021 509 517 34217553
37 Scherer A.M. Gedlinske A.M. Parker A.M. Gidengil C.A. Askelson N.M. Petersen C.A. Acceptability of adolescent COVID-19 vaccination among adolescents and parents of adolescents – United States, April 15–23, 2021 MMWR Morb Mortal Wkly Rep 70 2021 997 1003 34264908
38 Faye S.L.B. Krumkamp R. Doumbia S. Tounkara M. Strauss R. Ouedraogo H.G. Factors influencing hesitancy towards adult and child COVID-19 vaccines in rural and urban West Africa: a cross-sectional study BMJ Open 12 2022 e059138
39 Hilyard K.M. Quinn S.C. Kim K.H. Musa D. Freimuth V.S. Determinants of parental acceptance of the H1N1 vaccine Health Educ Behav 41 2014 307 314 24369176
40 Smith L.E. Amlot R. Weinman J. Yiend J. Rubin G.J. A systematic review of factors affecting vaccine uptake in young children Vaccine 35 2017 6059 6069 28974409
41 Gadoth A. Halbrook M. Martin-Blais R. Gray A. Tobin N.H. Ferbas K.G. Cross-sectional assessment of COVID-19 vaccine acceptance among health care workers in Los Angeles Ann Intern Med 174 2021 882 885 33556267
42 Anderson E.J. Campbell J.D. Creech C.B. Frenck R. Kamidani S. Munoz F.M. Warp speed for coronavirus disease 2019 (COVID-19) vaccines: why are children stuck in neutral? Clin Infect Dis 73 2021 336 340 32945335
43 Raude J. Fressard L. Gautier A. Pulcini C. Peretti-Watel P. Verger P. Opening the 'Vaccine Hesitancy' black box: how trust in institutions affects French GPs' vaccination practices Exp Rev Vacc 15 2016 937 948
44 Rey D. Fressard L. Cortaredona S. Bocquier A. Gautier A. Peretti-Watel P. Vaccine hesitancy in the French population in 2016, and its association with vaccine uptake and perceived vaccine risk-benefit balance Euro Surveill 23 2018
45 Xue F.X. Shen K.L. COVID-19 in children and the importance of COVID-19 vaccination World J Pediatr 17 2021 462 466 34581958
46 Gerber J.S. Offit P.A. COVID-19 vaccines for children Science 374 2021 913 34793207
| 36494253 | PMC9705202 | NO-CC CODE | 2022-12-07 23:15:49 | no | Vaccine. 2022 Nov 29; doi: 10.1016/j.vaccine.2022.11.060 | utf-8 | Vaccine | 2,022 | 10.1016/j.vaccine.2022.11.060 | oa_other |
==== Front
J Autoimmun
J Autoimmun
Journal of Autoimmunity
0896-8411
1095-9157
The Authors. Published by Elsevier Ltd.
S0896-8411(22)00167-6
10.1016/j.jaut.2022.102959
102959
Article
Post-mRNA vaccine flares in autoimmune inflammatory rheumatic diseases: Results from the COronavirus National Vaccine registry for ImmuNe diseases SINGapore (CONVIN-SING)
Ma Margaret ab
Santosa Amelia ab
Fong Warren bcd
Chew Li-Ching bcd
Low Andrea HL bcd
Law Annie cd
Poh Yih Jia c
Yeo Siaw Ing c
Leung Ying Ying cd
Ng Victoria WW e
Koh Joshua ZE e
Tay Sen Hee ab
Mak Anselm ab
Teng Gim Gee abf
Xu Chuanhui g
Tang Johnston GX g
Kong Kok Ooi g
Angkodjojo Stanley h
Goh Wei-Rui h
Chuah Tyng Yu h
Roslan Nur Emillia h
Arkachaisri Thaschawee di
Teh Kai Liang i
Sriranganathan Melonie j
Tan Teck Choon k
Phang Kee Fong bf
Yap Qai Ven l
Chan Yiong Huak l
Cheung Peter PM ab
Lahiri Manjari ab∗
a Division of Rheumatology, Department of Medicine, National University Hospital, Singapore
b Department of Medicine, Yong Loo Lin School of Medicine, National University of Singapore, Singapore
c Department of Rheumatology and Immunology, Singapore General Hospital, Singapore
d Duke-NUS Medical School, Singapore
e Yong Loo Lin School of Medicine, National University of Singapore, Singapore
f Chronic Programme, Alexandra Hospital, Singapore
g Department of Rheumatology, Allergy and Immunology, Tan Tock Seng Hospital, Singapore
h Rheumatology Service, Department of General Medicine, Sengkang General Hospital, Singapore
i Rheumatology and Immunology Service, Department of Paediatric Subspecialties, KK Women's and Children's Hospital, Singapore
j Division of Rheumatology, Department of Medicine, Changi General Hospital, Singapore
k Division of Rheumatology, Department of Medicine, Khoo Teck Puat Hospital, Singapore
l Biostatistics Unit, Yong Loo Lin School of Medicine, National University of Singapore, Singapore
∗ Corresponding author. Department of Medicine National University Hospital 1E Kent Ridge Road Singapore 119228.
29 11 2022
1 2023
29 11 2022
134 102959102959
27 8 2022
21 11 2022
23 11 2022
© 2022 The Authors
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Background
Studies of flares of autoimmune inflammatory rheumatic diseases (AIIRD) after COVID-19 mRNA vaccination are limited by small sample size, short follow up or at risk of selection bias.
Methods
A national retrospective cohort study of consecutive AIIRD patients ≥12 years old, across 8 hospitals who received at least one dose of a COVID-19 mRNA vaccine. Patients were included from the date of 1st vaccine dose and censored at the time of flare or on the date of the clinic visit at least 3 months from cohort entry, whichever came first. Predictors of flare were determined by Cox proportional hazards analysis.
Findings
4627 patients (73% Chinese, 71% female) of median (IQR) age 61 (48, 70) years were included; 42% Rheumatoid arthritis, 14% Systemic lupus erythematosus and 11% Psoriatic arthritis. 47% were in remission, 41% low disease activity, 10% moderate disease activity and 1% in high disease activity. 18% patients flared, of which 11.7% were within the 3-month period of interest. 11.8% patients improved. Median (IQR) time-to-flare was 60 (30, 114) days. 25% flares were self-limiting, 61% mild-moderate and 14% severe. Older patients (53–65 years and >66 years) had a lower risk of flare [HR 0.6 (95% CI 0.5–0.8) and 0.7 (0.6–0.8) respectively]. Patients with inflammatory arthritis and with active disease had a higher risk of flare [HR 1.5 (1.2–2.0) and 1.4 (1.2–1.6), respectively]. Treatment with conventional synthetic disease-modifying anti-rheumatic drugs (csDMARDs), immunosuppression and prednisolone was also associated with an increased risk of flare [HR 1.5 (1.1–2), 1.2 (1.1–1.4) and 1.5 (1.2–1.8) for prednisolone ≤7.5 mg respectively].
Interpretation
There was a moderately high rate of AIIRD flares after mRNA vaccination but also improvement in several patients. Severe flares and hospitalisation were rare. Thus, vaccination remains safe and highly recommended.
Keywords
Autoimmune inflammatory rheumatic diseases
COVID vaccines
Registry
==== Body
pmc1 Introduction
Patients with autoimmune inflammatory rheumatic diseases (AIIRD) have poorer outcomes when infected with SARS-CoV2, including higher rates of hospitalisation, oxygen support, intensive care unit (ICU) admission and mortality [1]. Vaccination is currently the most effective way to reducing mortality, with international and national rheumatology societies recommending COVID-19 vaccination for all patients with AIIRD [[2], [3], [4]]. Although initial clinical trials of COVID-19 vaccines have largely excluded patients with AIIRD [5,6], there is a growing body of evidence that COVID-19 vaccines are safe and effective for this group of patients [[7], [8], [9]]. Universal uptake of vaccination is an essential part of controlling COVID-19 at both the patient and population level. However, vaccine hesitancy has been a major barrier, with fear of side effects being a significant factor [10]. In addition, up to 44% of AIIRD patients have a fear of potentially triggering flares of their rheumatic disease [9].
Rates of flares of AIIRD after COVID-19 vaccination in various studies have been wide-ranging from no flares to up to 18% [[7], [8], [9],[11], [12], [13], [14], [15], [16], [17], [18]]. Contributing factors to this variability include different study designs, varying sample size, varying length of follow-up and definition of “at-risk” period and differences in flare definitions. In addition, most of these studies are at risk of selection bias, as they are based on voluntary physician reports or patient surveys. Higher rates of flares have been described in patients who have history of self-reported flares within the past 6 months, patients with inflammatory arthritides and those who have had past COVID infection [7,8].
Comprehensive inclusion of all vaccinated patients is an important factor when evaluating incidence of flares. In Singapore, COVID-19 vaccination began in January 2021, almost exclusively with the two mRNA vaccines (BNT162b2 Pfizer-BioNTech® COVID-19 vaccine and the mRNA-1273 Moderna® COVID-19 vaccine). Vaccination was offered to all adults, and children 12 years and older by July 2021. By November 2021, 82% of the Singapore population had completed 2 doses of COVID-19 vaccination [19]. As such, we set up the COronavirus National Vaccine registry for ImmuNe diseases SINGapore (CONVIN-SING), to accurately assess flares, and the risk factors for flares, from direct medical record review of all patients with AIIRD attending public sector hospitals in Singapore and who had received an mRNA vaccine against SARS-COV2.
We aimed to determine the proportion of patients with rheumatic disease who flared within 3 months of the 1st dose of an mRNA vaccine, assess time to flare and predictors of flare.
2 Methods
Using a retrospective cohort study design and electronic health record (EHR) review, we included adults and children 12 years of age or older, who had received at least one dose of an mRNA vaccine and who had been diagnosed with an AIIRD prior to vaccination. The AIIRD of interest were pre-defined and are listed in the Supplementary Appendix Table S1. Eight of the nine public hospitals in Singapore participated in this study. A list of patients with AIIRD was extracted using SNOMED diagnosis codes and matched against the National Immunisation Register (NIR) for type and dates of vaccine doses. Matching with the NIR was done in August 2021 and refreshed once in November 2021. We included patients consecutively, in order of the date of the 1st vaccine dose.
Patients entered the cohort from the date of 1st vaccine dose and were censored at the time of flare or on the date of the clinic visit which was at least 3 months from cohort entry, whichever came first. Data were abstracted manually onto a secure web-based portal via manual chart review and included demographics, confirmation of diagnosis, physician-defined baseline disease activity (remission, low, moderate or high), baseline treatment, previous diagnosis of COVID-19, severity and date of flare if any, and hospitalisation for flare. A training session was conducted to standardise data entry across sites. Practice cases were used to harmonise definitions of pre-vaccine disease activity, and adjudication of flare. The primary outcome was a flare of AIIRD between 0 and 3 months after the 1st dose of anti-SARS-CoV mRNA vaccine.
The severity of flare was classified as follows: A mild, self-limiting flare was defined as one that did not require an escalation of treatment or early consult with the specialist, and resolved before the index visit, though the patient may have taken analgesics or non-steroidal anti-inflammatory drugs (NSAID) or self-increased glucocorticoids (GC) for a few days. A mild to moderate, non-self-limiting flare was one that necessitated an earlier specialist consult and/or increase in treatment, but not exceeding prednisolone 20 mg per day and not requiring intramuscular (IM) or intra-articular (IA) GC injection or new biologic or cytotoxic drug (e.g. cyclophosphamide) initiation. A severe flare was one that required hospitalisation or GC more than prednisolone equivalent of 20 mg per day or IM or IA GC, or new initiation of a biologic or cytotoxic agent.
Ethics approval was obtained from the National Healthcare Group (NHG) Domain Specific Review Board (DSRB) Ref: 2021/0043. The requirement for patient consent was waived as the data were collected via EHR review only, without any direct patient contact.
Statistical analyses: Sample size was calculated based on a background flare rate of 8% over 3 months in a rheumatoid arthritis (RA) cohort and power of 90% to detect a 25% relative increase in flares [20]. All analyses were performed using Stata version 16 (StataCorp). Continuous variables are presented as median (interquartile range, IQR) while categorical variables are presented as frequencies and percentages. Predictors of flare and time to flare were assessed using a mixed effect stepwise backward-selection Cox regression model to account for clustering effect of site, adjusting for demographics and relevant covariates. Age (as tertiles), race, AIIRD diagnosis (grouped as inflammatory arthritis (IA), connective tissue disease (CTD), vasculitis and others) treatment [grouped as no disease modifying antirheumatic drugs (DMARDs), hydroxychloroquine only, conventional synthetic (cs)DMARDs (methotrexate, sulfasalazine or leflunomide), immunosuppressants (cyclosporine A, mycophenolate mofetil, mycophenolate sodium, tacrolimus, azathioprine, cyclophosphamide), and biological (b)DMARDs or Janus Kinase inhibitors (JAKi)] and GC dose (grouped as no GC, prednisolone≤7.5 mg, prednisolone >7.5 mg) were entered into the model as categorical variables. Gender, vaccine type and disease activity (grouped as inactive disease for patients in remission or low disease activity and active disease for those with moderate or high disease activity) were used as binary variables. The model was adjusted for number of vaccine doses received. Kaplan Meier curve was generated for time to flare after first dose. The proportion of patients who flared within 3 weeks of the first versus second dose were compared using Chi squared test. For subjects who flared, line graphs by disease group were plotted for time to flare.
3 Results
4627 patients (73.3% Chinese, 71.1% female) of median (IQR) age 61 (48, 70) years were included in the analysis (Table 1 ). 4058 (87.7%) received the BNT162b2 (Pfizer/BioNTech) COVID-19 vaccine and 523 (11.3%) received the mRNA-1273 (Moderna) vaccine, with a median (IQR) interval of 21 (21, 28) days between the two doses. 127 patients (2.7%) had a single vaccine dose. The most common AIIRD diagnoses were RA (1966, 42.4%), systemic lupus erythematosus (SLE) (641, 13.8%) and psoriatic arthritis (486, 10.5%). At the pre-vaccination clinic visit, 2188 (47.3%) patients were in remission, 1897 (41%) had low disease activity, 480 (10.4%) had moderate disease activity and 62 (1.3%) had high baseline disease activity. 1389 (30%) were treated with hydroxychloroquine, 1684 (36%) with methotrexate and 848 (18%) treated with sulfasalazine. 431 (9.3%) were treated with bDMARD/JAKi. The median prednisolone dose was 0 mg (IQR 0,4). Treatment was interrupted for vaccination in only 63 (1.4%) patients. Only 45 (1%) patients had previous COVID-19 infection.Table 1 Patient characteristics.
Table 1Baseline characteristics Whole Cohort (n = 4627) Stable (n = 3228,
69.8%) Flares within 0–3 months of 1st vaccine dose (n = 542, 11.7%) Flares outside of 0–3 months after 1st vaccine dose (n = 312, 6.8%) Improved (n = 545, 11.8%)
Age 61 61 58 62 60
(median years, IQR) (48, 70) (47, 70) (46, 67) (48, 70) (51, 67)
Ethnicity
Chinese 3394 (73) 2388 (74) 395 (73) 226 (72) 385 (71)
Malay 486 (11) 322 (10) 60 (11) 31 (10) 73 (13)
Indian 473 (10) 321 (10) 61 (11) 43 (14) 48 (9)
Others 274 (6) 197 (6) 26 (5) 12 (4) 39 (7)
Gender
Female 3293 (71) 2313 (72) 390 (72) 205 (66) 385 (71)
Vaccine type
Pfizer/BioNTech 4058 (88) 2862 (89) 459 (86) 271 (87) 466 (87)
Moderna 523 (11) 339 (11) 76 (14) 39 (13) 69 (13)
Diagnosis
Rheumatoid Arthritis 1966 (42) 1314 (41) 264 (49) 149 (49) 239 (44)
Systemic Lupus Erythematosus 641 (14) 496 (15) 51 (9) 21 (7) 73 (13)
Psoriatic Arthritis 486 (11) 302 (9) 79 (15) 41 (13) 64 (12)
Spondyloarthropathies 484 (10) 325 (10) 60 (11) 46 (15) 53 (10)
Sjogren's Syndrome 261 (6) 208 (6) 22 (4) 11 (4) 20 (4)
Systemic sclerosis 159 (3) 128 (4) 8 (1) 7 (2) 16 (3)
Treatment
Methotrexate 1684 (36) 1088 (34) 233 (43) 124 (40) 239 (44)
Sulfasalazine 848 (18) 553 (17) 125 (23) 73 (23) 97 (18)
Hydroxychloroquine 1389 (30) 969 (30) 154 (28) 77 (25) 189 (35)
Mycophenolate Mofetil 296 (6) 210 (7) 32 (6) 13 (4) 41 (8)
Biological DMARDs 383 (8) 249 (8) 57 (11) 24 (8) 53 (10)
JAK inhibitors 46 (1) 24 (1) 6 (1) 8 (3) 8 (1)
Prednisolone dose (mg, median, IQR) 0 (0,4) 0 (0, 2.5) 0 (0,5) 0 (0,5) 0 (0,5)
Baseline Physician Disease Activity
Remission 2188 (47) 1761 (55) 179 (33) 115 (36) 133 (24)
Low Disease Activity 1897 (41) 1235 (39) 257 (47) 157 (50) 248 (46)
Moderate Disease Activity 480 (10) 213 (7) 92 (17) 37 (12) 138 (25)
High Disease Activity 62 (1) 19 (1) 14 (3) 3 (0.96) 26 (5)
Data are shown in number (percentages) unless specified.
854 (18%) flares were recorded during 20,287 patient-months of follow-up (4.5/100 patient-months, median (IQR) follow up duration 4.3 (3.4, 5.5) months), of which 542 (11.7%) patients flared in the 3-month period of interest (Fig. 1 ). Median (IQR) time to flare was 60 (30–114) days. When examining by disease groups, a biphasic peak was seen for patients with IA (Fig. 2 A). Separating the flares after the 1st dose from those after the 2nd dose, the median time to flare for patients with IA was 10 (5–17) days after the 1st dose (Figs. 2B) and 60 (21–108) days after the 2nd dose (Fig. 2C)].Fig. 1 Kaplan-Meier graph - Time to Flare after the First Dose of mRNA Vaccine.
Fig. 1
Fig. 2 Time to flare by disease group. A = All Flares, B = Flares after dose 1 and before dose 2, C = Flares after dose 2 (patients who flared after dose 1 are excluded).
Fig. 2
To compare flare rate after the 1st versus the 2nd dose, we limited the period of observation to within 3 weeks of the vaccine, as the 2nd dose was given 3 weeks after the 1st dose in most patients. 144 patients (3.2%) flared after the 1st dose. These patients were censored at the date of flare and hence not included in the susceptible population for flare after the 2nd dose. 159 patients (3.6%) flared after the 2nd dose (chi2 for difference in proportions = 0.23).
134 (24.7%) of all flares were mild and self-limiting, 333 (61.4%) were mild-moderate and 75 (13.8%) were severe. 393 (72.5%) of those who flared required escalation of treatment and 30 (5.5%) required hospital admission. Conversely, 545 patients (11.8%) had improved disease activity after the vaccine.
On Cox regression analysis, patients in the older age tertiles (53–65 and > 66 years) had a lower risk of flare, hazard ratio (HR) 0.6, 95% CI 0.5–0.8, p < 0.001 and 0.7, 95% CI 0.6–0.8 p < 0.001 respectively (Table 2 ). Patients of the minority non-Chinese/non- Malay/non-Indian ethnicity had a lower risk of flare (HR 0.7 95% CI 0.5–0.9). Compared to CTD, patients with IA had a HR of flare of 1.5, 95% CI 1.2–2.0, p = 0.006. Patients with baseline active disease had a 40% higher risk of flare compared to those with inactive disease (HR 1.4, 95% CI 1.2–1.6, p < 0.001). Treatment with csDMARDs, immunosuppressants and prednisolone was also associated with an increased risk of flare [HR 1.5 (1.1–2), 1.2 (1.1–1.4) and 1.5 (1.2–1.8) for prednisolone dose ≤7.5 mg respectively].Table 2 Predictors of time to flare.
Table 2Variables Unadjusted Adjusted
HR (95% CI) p-value HR (95% CI) p-value
Vaccine type
BNT162b2 Pfizer-BioNTech 1.0 0.06
mRNA-1273 Moderna 1.2 (1–1.5)
Age
≤52 1.0 1.0
53-65 0.8 (0.7–0.97) 0.02 0.6 (0.5–0.8) <0.001
>66 0.8 (0.7–0.9) 0.002 0.7 (0.6–0.8) <0.001
Ethnicity
Chinese 1.0 1.0
Malay 1.0 (0.8–1.3) 0.78 0.7 (0.4–1.1) 0.14
Indian 1.2 (1–1.5) 0.08 1.0 (0.8–1.3) 0.75
Others 0.8 (0.5–1.1) 0.10 0.7 (0.5–0.9) 0.006
Gender
Male 1.0
Female 0.9 (0.8–1.1) 0.53
Disease group
Connective Tissue Disease 1.0 1.0
Vasculitis 1.4 (0.9–2.2) 0.18 1.2 (0.7–1.8) 0.67
Inflammatory arthritis 1.9 (1.6–2.2) <0.001 1.5 (1.2–2.0) 0.001
Others 1.3 (0.8–2.2) 0.35 1.4 (0.8–2.5) 0.28
Active disease
No 1.0 1.0
Yes 1.7 (1.4–2.1) <0.001 1.4 (1.2–1.6) <0.001
Glucocorticoids
None 1.0 1.0
Pred≤7.5 mga 1.4 (1.2–1.6) <0.001 1.5 (1.2–1.8) <0.001
Pred >7.5 mga 1.7 (1.2–2.3) 0.002 1.7 (1–3.0) 0.05
Treatment
None 1.0 1.0
Hydroxychloroquine only 0.9 (0.7–1.2) 0.48 1.1 (0.7–1.7) 0.55
csDMARDsb 1.9 (1.5–2.3) <0.001 1.5 (1.1–2.0) 0.005
Immunosuppressantsc 1.2 (0.9–1.6) 0.16 1.2 (1.1–1.4) 0.004
bDMARDs or JAKid 1.9 (1.4–2.5) <0.001 1.3 (0.9–1.7) 0.11
Multivariate analysis has been adjusted for number of vaccine doses. In the adjusted model, 96.8% of subjects (4327 out of 4627 subjects) were used.
Significant p-value in bold.
a Or equivalent dose in prednisolone.
b csDMARDs = conventional synthetic disease modifying anti-rheumatic drugs which included Methotrexate, Sulfasalazine and Leflunomide.
c Immunosuppressants include Cyclosporin A, Cyclophosphamide, Azathioprine, Mycophenolate Mofetil, Mycophenolate Sodium, Tacrolimus.
d bDMARDs or JAKi = biological disease modifying anti-rheumatic drugs or Janus Kinase inhibitors.
4 Discussion
We have described post-vaccination flares in an inception cohort of consecutively vaccinated AIIRD patients across Singapore. Flares were seen in 18% of our patients, of which 11.7% were within 3 months of the first vaccine dose, with a median time to flare of 60 days. Conversely, 11.8% had improved disease activity after the vaccine. Only 75 (1.6%) patients had a severe flare, while 30 (0.6% of those vaccinated) required hospitalisation. Given the high morbidity and mortality of COVID-19 in patients with AIIRD, our study strongly supports a favourable risk-benefit ratio for vaccination of these vulnerable patients [7,21].
Our study has several strengths. We included all AIIRD patients from eight of the nine public hospitals in Singapore, sequentially, in order of the date of vaccination. Vaccination data were obtained centrally through the NIR and thus represents a near-complete capture of all eligible AIIRD patients in Singapore. Demographic and clinical data were abstracted manually through detailed EHR review, by trained personnel. Flares were adjudicated by the attending physician, and were graded for severity.
The rate of flares in our cohort is comparable to other studies that evaluated the rate of flare after COVID-19 vaccination, although, as mentioned, these studies were mostly based on patient self-reported symptom questionnaires. These previous studies have reported flare rates between 2.2 and 18.8% [7,13,15,16]. Studies which employed physician assessment were based on voluntary physician reporting of patients at varying intervals post-vaccination. A recent study using linked government vaccine records noted no increased risk of flare with COVID-19 vaccine in patients with RA [11]. However, this study used administrative data on outpatient/hospital visits and prescription patterns as proxies to assess disease flares.
Only 1% of our cohort had previous COVID-19 infection at the time of our study. This was in keeping with the low population prevalence of COVID in Singapore in 2021 (3%, 198,374) [22] due to stringent COVID restrictions mandated by the Singapore government. As in other cohorts, the predominant AIIRDs in our patients were IA (67%) and SLE (13.8%). In studies based on patient-surveys, symptoms such as joint pain, stiffness, myalgia and fatigue were taken to indicate disease flare. While these are important symptoms to elicit from patients, objective assessment by a physician is important to discern whether these symptoms truly represent disease flare. To this end, 75% of our patients who flared were assessed by their rheumatologist to require escalation of treatment, suggesting that they were true flares.
We found that patients with IA were more likely to experience a flare of disease. Our observed 3-month flare rate does appear to be higher than previously reported normalized background yearly flare-rates among patients with stable RA (approximately 7.9% flares per 3-months) [20] and SLE (approximately 3.2% per 3-months) [23]. Common post-vaccination effects such as arthralgias or myalgias leading to tender joints on examination and elevated inflammatory markers may contribute toward a higher perceived disease activity, potentially overestimating the observed flare rates. It is also reassuring to know that over 10% of patients continue to report improvement in their disease activity, as part of their natural disease history, highlighting that patients are still able to have better control of their disease over the 3-months despite receiving vaccination.
It is interesting to observe that treatment with csDMARDs, immunosuppression and a higher dose of prednisolone were associated with flares. Firstly, while only 1.4% patients were documented to have withheld their treatment prior to vaccination, this may be an under-estimation, leading to more flares in patients pausing higher dose treatment. Secondly, patients on such therapy are likely on closer follow up than those who are on minimal treatment, and thus this finding may be due to ascertainment bias.
Vaccine recommendations in people with rheumatic disease, both for COVID-19 vaccination and other non-COVID vaccines, are largely based on vaccine studies conducted in patients with quiescent disease [4,24]. Suboptimal disease control at the time of vaccination has been reported to be associated with reduced vaccine immunogenicity, likely due to the associated therapy [25]. In addition, similar to others using patient-reported flares, we found that suboptimal disease control at the time of vaccination was found to predict post-vaccination flare [7,15].
Older patients had a lower risk of AIIRD flare in our study. This may be partially attributed to reduced vaccine immunogenicity in the elderly population [26]. Other factors, such as disease duration, depth and/or length of remission at the time of vaccination and immunosuppressive medications in this population may have also played a part.
Our study has certain limitations. While the Singapore national guidelines do not recommend interruption of treatment for vaccination [4], suboptimal medication adherence and dose changes by the patients themselves may not have been captured in our retrospective study. Similarly, flares were only captured if reported to, and documented by, the treating physician in the medical record, and the time of flare relative to vaccine administration was usually approximated.
To our knowledge this is the largest and most comprehensive report of post-COVID-19 vaccination flares in AIIRD to date. While the flare rate appears higher than the previously reported background flare rate among stable AIIRD patients, causality cannot be determined from this retrospective observational study. Moreover, we reported a large number of patients with improved disease activity after vaccination, suggesting that both the flares and improvement may just be a part of the natural history of disease, and warrants further study. Patients with IA and suboptimal disease control were at higher risk of post-vaccination flares and may require closer follow up after vaccination. Severe flares and hospitalisation were rare; thus, vaccination remains safe and highly recommended for our vulnerable patients with AIIRD.
Author statement
Margaret MA: Conceptualization, Data curation, Methodology, Supervision, Validation, Writing – original draft; Amelia SANTOSA: Conceptualization, Methodology, Writing – review & editing; Warren FONG: Conceptualization, Data curation, Methodology, Supervision, Validation, Writing – review & editing; Li-Ching CHEW: Writing – review & editing; Andrea HL LOW; Conceptualization, Writing – review & editing; Annie LAW: Writing – review & editing, Yih Jia POH: Writing – review & editing; Siaw Ing YEO: Writing – review & editing; Ying Ying LEUNG: Writing – review & editing; Victoria WW NG: Data curation; Joshua ZE KOH: Data curation; Sen Hee TAY: Conceptualization, Methodology, Writing – review & editing; Anselm MAK: Conceptualization, Methodology, Writing – review & editing; Gim Gee TENG: Writing – review & editing; Chuanhui XU: Data curation; Johnston GX TANG: Data curation: Kok Ooi KONG: Writing – review & editing; Stanley ANGKODJOJO: Data curation, Supervision, Validation, Writing – review & editing; Wei-Rui GOH: Data curation; Tyng Yu CHUAH: Data curation; Nur Emillia ROSLAN: Data curation; Thaschawee ARKACHAISRI: Writing – review & editing; Kai Liang THE: Data curation; Melonie SRIRANGANATHAN: Writing – review & editing; Teck Choon TAN: Data curation; Kee Fong PHANG: Data curation; Qai Ven YAP: Formal analysis; Yiong Huak CHAN: Formal analysis, Supervision; Peter PM CHEUNG: Conceptualization, Methodology, Writing – review & editing; Manjari LAHIRI: Data curation, Formal analysis, Methodology, Project administration, Supervision, Validation, Visualization, Writing – review & editing.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Appendix A Supplementary data
The following is the Supplementary data to this article:Multimedia component 1
Multimedia component 1
Acknowledgement
The authors would like to acknowledge the Chapter of Rheumatologists, College of Physicians, Academy of Medicine, Singapore.
Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.jaut.2022.102959.
==== Refs
References
1 Xu C. Yi Z. Cai R. Clinical outcomes of COVID-19 in patients with rheumatic diseases: a systematic review and meta-analysis of global data Autoimmun. Rev. 2021 102778
2 Curtis J.R. Johnson S.R. Anthony D.D. American college of rheumatology guidance for COVID-19 vaccination in patients with rheumatic and musculoskeletal diseases: version 3 Arthritis Rheumatol. 73 10 2021 e60 e75 34346564
3 Landewe R.B.M. Kroon F.P.B. Alunno A. EULAR recommendations for the management and vaccination of people with rheumatic and musculoskeletal diseases in the context of SARS-CoV-2: the November 2021 update Ann. Rheum. Dis. 12 2022 1628 1639 10.1136/annrheumdis-2021-222006
4 Santosa A. Xu C. Arkachaisri T. Recommendations for COVID-19 vaccination in people with rheumatic disease: developed by the Singapore Chapter of Rheumatologists Int J Rheum Dis 24 6 2021 746 757 33973379
5 Baden L.R. El Sahly H.M. Essink B. Efficacy and safety of the mRNA-1273 SARS-CoV-2 vaccine N. Engl. J. Med. 384 5 2021 403 416 33378609
6 Polack F.P. Thomas S.J. Kitchin N. Safety and efficacy of the BNT162b2 mRNA covid-19 vaccine N. Engl. J. Med. 383 27 2020 2603 2615 33301246
7 Connolly C.M. Ruddy J.A. Boyarsky B.J. Disease flare and reactogenicity in patients with rheumatic and musculoskeletal diseases following two-dose SARS-CoV-2 messenger RNA vaccination Arthritis Rheumatol. 74 1 2022 28 32 34346185
8 Machado P.M. Lawson-Tovey S. Strangfeld A. Safety of vaccination against SARS-CoV-2 in people with rheumatic and musculoskeletal diseases: results from the EULAR Coronavirus Vaccine (COVAX) physician-reported registry Ann. Rheum. Dis. 81 5 2022 695 709 34972811
9 Sattui S.E. Liew J.W. Kennedy K. Early experience of COVID-19 vaccination in adults with systemic rheumatic diseases: results from the COVID-19 Global Rheumatology Alliance Vaccine Survey RMD Open 7 3 2021
10 Ko T. Dendle C. Woolley I. SARS-COV-2 vaccine acceptance in patients with rheumatic diseases: a cross-sectional study Hum. Vaccines Immunother. 17 11 2021 4048 4056
11 Li X. Tong X. Yeung W.W.Y. Two-dose COVID-19 vaccination and possible arthritis flare among patients with rheumatoid arthritis in Hong Kong Ann. Rheum. Dis. 81 4 2022 564 568 34686479
12 Fan Y. Geng Y. Wang Y. Safety and disease flare of autoimmune inflammatory rheumatic diseases: a large real-world survey on inactivated COVID-19 vaccines Ann. Rheum. Dis. 81 3 2022 443 445 34824048
13 Barbhaiya M. Levine J.M. Bykerk V.P. Systemic rheumatic disease flares after SARS-CoV-2 vaccination among rheumatology outpatients in New York City Ann. Rheum. Dis. 80 10 2021 1352 1354 34158370
14 Izmirly P.M. Kim M.Y. Samanovic M. Evaluation of immune response and disease status in systemic lupus erythematosus patients following SARS-CoV-2 vaccination Arthritis Rheumatol. 74 2 2022 284 294 34347939
15 Rotondo C. Cantatore F.P. Fornaro M. Preliminary data on post market safety profiles of COVID 19 vaccines in rheumatic diseases: assessments on various vaccines in use, different rheumatic disease subtypes, and immunosuppressive therapies: a two-centers study Vaccines (Basel) 9 7 2021
16 Peet C.J. Papadopoulou C. Sombrito B.R.M. COVID-19 and autoinflammatory diseases: prevalence and outcomes of infection and early experience of vaccination in patients on biologics Rheumatol Adv Pract 5 2 2021 rkab043 34466775
17 Furer V. Eviatar T. Zisman D. Immunogenicity and safety of the BNT162b2 mRNA COVID-19 vaccine in adult patients with autoimmune inflammatory rheumatic diseases and in the general population: a multicentre study Ann. Rheum. Dis. 80 10 2021 1330 1338 34127481
18 Braun-Moscovici Y. Kaplan M. Braun M. Disease activity and humoral response in patients with inflammatory rheumatic diseases after two doses of the Pfizer mRNA vaccine against SARS-CoV-2 Ann. Rheum. Dis. 80 10 2021 1317 1321 34144967
19 MoH Singapore Vaccination Statistics 2022 [updated 26 June 2022. Available from: https://www.moh.gov.sg/covid-19/vaccination/statistics
20 Bechman K. Tweehuysen L. Garrood T. Flares in rheumatoid arthritis patients with low disease activity: predictability and association with worse clinical outcomes J. Rheumatol. 45 11 2018 1515 1521 30173149
21 Tzioufas A.G. Bakasis A.D. Goules A.V. A prospective multicenter study assessing humoral immunogenicity and safety of the mRNA SARS-CoV-2 vaccines in Greek patients with systemic autoimmune and autoinflammatory rheumatic diseases J. Autoimmun. 125 2021 102743
22 Medicine J.H.U. Coronavirus resource center 2022 [Available from: https://coronavirus.jhu.edu/region/singapore
23 Cho J. Lahiri M. Teoh L.K. Predicting flares in patients with stable systemic lupus erythematosus Semin. Arthritis Rheum. 49 1 2019 91 97 30660381
24 Furer V. Rondaan C. Heijstek M.W. Update of EULAR recommendations for vaccination in adult patients with autoimmune inflammatory rheumatic diseases Ann. Rheum. Dis. 79 1 2019 39 52 2020 31413005
25 Campos L.M. Silva C.A. Aikawa N.E. High disease activity: an independent factor for reduced immunogenicity of the pandemic influenza a vaccine in patients with juvenile systemic lupus erythematosus Arthritis Care Res. 65 7 2013 1121 1127
26 Collier D.A. Ferreira I. Kotagiri P. Age-related immune response heterogeneity to SARS-CoV-2 vaccine BNT162b2 Nature 596 7872 2021 417 422 34192737
| 36473406 | PMC9705203 | NO-CC CODE | 2022-12-05 23:15:17 | no | J Autoimmun. 2023 Jan 29; 134:102959 | utf-8 | J Autoimmun | 2,022 | 10.1016/j.jaut.2022.102959 | oa_other |
==== Front
Clinics (Sao Paulo)
Clinics (Sao Paulo)
Clinics
1807-5932
1980-5322
HCFMUSP. Published by Elsevier España, S.L.U.
S1807-5932(22)03351-8
10.1016/j.clinsp.2022.100150
100150
Original Articles
Anti-SARS-CoV-2 inactivated vaccine in patients with ANCA-associated vasculitis: immunogenicity, safety, antibody decay and the booster dose
Pereira Rosa M.R. a
Dagostin Marilia A. a⁎
Caparbo Valeria F. a
Sales Lucas P. a
Pasoto Sandra G. a
Silva Clovis A. b
Yuki Emily F.N. a
Saad Carla G.S. a
Medeiros-Ribeiro Ana C. a
Kupa Leonard V.K. a
Fusco Solange R.G. a
Martins Victor A.O. a
Martins Carolina C.M.F. c
Barbas Carmen Valente d
Shinjo Samuel K. a
Aikawa Nadia E. ab
Bonfa Eloisa a
a Rheumatology Division, Hospital das Clínicas, Faculdade de Medicina, Universidade de São Paulo (HCFMUSP), São Paulo, SP, Brazil
b Pediatric Rheumatology Unit, Instituto da Criança e do Adolescente, Hospital das Clínicas, Faculdade de Medicina, Universidade de São Paulo (HCFMUSP), São Paulo, Brazil
c Undergraduate Medical Student, Rheumatology Division, Hospital das Clínicas, Faculdade de Medicina, Universidade de São Paulo (HCFMUSP), São Paulo, SP, Brazil
d Pulmonary Division, Hospital das Clínicas, Faculdade de Medicina, Universidade de São Paulo (HCFMUSP), São Paulo, SP, Brazil
⁎ Corresponding author.
29 11 2022
29 11 2022
1001504 4 2022
10 10 2022
17 11 2022
© 2022 HCFMUSP. Published by Elsevier España, S.L.U.
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Objective
To evaluate inactivated CoronaVac prime vaccination, antibody decay, booster dose, and safety in ANCA-Associated Vasculitis (AAV) patients.
Methods
Fifty-three AAV patients and 106 Controls (CG) received CoronaVac on days: D0 (first dose), D28(second dose), and D210 (booster dose, 32 AAV: 32 CG). The primary outcome was immunogenicity after the second vaccine dose (day 69) assessed by Seroconversion Rates (SC) of anti-SARS-CoV-2 S1/S2 IgG and Neutralizing Antibodies (NAb). Secondary outcomes were safety, immunogenicity (D28/D240), 6-months antibody decay (D210) and the booster dose response (D240).
Results
At D69 SC (65.1% vs. 96.8%, p = 0.0001), GMT (21.3 UA/mL vs. 67.7 UA/mL, p < 0.001) and NAb- positivity (53.7% vs. 80.6%, p = 0.001) were moderate but lower in naïve-AAV patients than CG. Patients without SC used more often IS (93.3% vs. 53.3%, p = 0.015), mycophenolate mofetil (20% vs. 0%, p = 0.037) and prednisone (60.0% vs. 28.6%, p = 0.057) than seroconverted. NAb negativity in AAV patients was associated with prednisone treatment (57.9% vs. 18.2%, p = 0.015) and IS (84.2% vs. 55.0%, p = 0.046). Logistic regression analysis models showed that only prednisone was associated with lower seroconversion (OR = 0.2, 0,95% CI 0.05‒0.86, p = 0.030) and with lower NAb positivity (OR = 0.2, 0,95% CI 0.05‒0.88, p = 0.034). After six months (D69‒D210) a decrease in IgG positivity occurred in 32 AAV patients (15.7%, p = 0.074) and 32 CG (18.7%, p = 0.041). For the NAb positivity, the 6-month decrease was not significant (p = 0.114) whereas a major reduction occurred for CG (p < 0.001). A booster dose (D240) resulted in an increment in IgG-positivity (21.9%, p = 0.023) and NAb-positivity (34.4%, p = 0.006) in AAV patients. No moderate/severe adverse events attributable to the vaccine were observed.
Conclusion
This study provides novel data on the excellent safety and moderate immunogenicity of CoronaVac in AAV patients. A six-month mild antibody waning was observed with a good response to the booster dose, although levels remained lower than CG (CoronavRheum-NCT04754698).
Keywords
ANCA-associated vasculitis
Vaccine
SARS-CoV-2
Immunogenicity
==== Body
pmcIntroduction
Coronavirus Disease 2019 (COVID-19) causes Severe Acute Respiratory Syndrome and the agent Coronavirus 2 (SARS-CoV-2), emerged in 2019 and has spread rapidly since then. The death toll of the pandemic is estimated to be millions and brought major damage not only in health-related issues but also in social and economic aspects across the globe.1 , 2 By the time of this submission, more than 460 million people have been infected with SARS-CoV-2 and nearly 6 million died from COVID-19 (WHO ‒ https://covid19.who.int/).
Pharmacological antiviral therapy for COVID-19 patients is scarce and not widely available, and therefore supportive care measures such as ventilation oxygenation and fluid management remain the standard of care.3 Consequently, mass vaccination is the most effective strategy for controlling the pandemic so far. In the past 18 months, several vaccines have been developed and commercialized in record time, with proven efficacy in phase III trials,4, 5, 6 including CoronaVac,7 an inactivated virus vaccine against SARS-CoV-2, with emergency use approval by the World Health Organization (WHO) in several most populated countries, including Brazil.
Although there are a number of papers evaluating the safety and efficacy of the COVID-19 vaccines in overall Autoimmune Rheumatic Diseases (ARD)8, 9, 10, 11 none focused specifically on rare diseases such as AAV. These individuals are the ones that theoretically have the greatest benefit from vaccination since their condition is frequently aggravated by renal and lung function impairment with a consequent increase in the risk of severe SARS-CoV-2 infection and death.12, 13, 14 It is not known if high immunosuppression would impact immunogenicity and the dynamics of 6-months antibody decay or booster dose. In addition, regarding safety, there is a concern if the level of disease activity would influence vaccine immunogenicity or else if the vaccine may trigger or aggravate systemic inflammation.
The CoronavRheum trial, a large Brazilian phase 4 trial in 910 adults with ARD showed that this vaccine has an overall moderate short-term immunogenicity although lower than the control group.11 Similarly, an mRNA COVID-19 vaccine induced reduced immune response in a cohort of global ARD patients compared to the control group, including a very small sample of AAV patients.9
Therefore, the aim of this study is to analyze CoronaVac safety, immunogenicity, antibody decay, and booster dose response in AAV patients and the Control Group (CG). The authors also evaluated the impact of disease activity and immunosuppressive treatment on the vaccine response of these patients.
Materials and methods
This prospective controlled trial is within a large phase 4 study (CoronavRheum clinicaltrials.gov #NCT04754698) conducted at a single tertiary center in Sao Paulo (Brazil) that assessed immunogenicity and safety of the CoronaVac COVID-19 vaccine in a large sample of ARD patients.11 Data were collected and managed using REDCap electronic capture tools hosted at the studied Institution.15 , 16 The study was conducted according to the Declaration of Helsinki and local regulations and was approved by the local and national ethical committee (CAAE: 42566621.0.0000.0068). Written informed consent was obtained from all participants.
Patients and controls
Consecutive naïve patients (COVID-19 seronegative) diagnosed with Granulomatosis with Polyangiitis (GPA), Eosinophilic Granulomatosis with Polyangiitis (EGPA), or Microscopic Polyangiitis (MPA) by the American College of Rheumatology Classification Criteria17 , 18 and the Chapel Hill Conference Classification14 aged ≥ 18 years old and with regular follow-up in the Vasculitis Outpatient Clinic were invited to participate in the study.
Subsequently, a CG of administrative hospital workers and their relatives was invited to participate. The two groups were age and sex-balanced (± 5 years) in a 2:1 ratio (2 controls: 1 patient) using an in-house Excel program (Microsoft 2018) for random selection of participants in each group. Autoimmune rheumatic disease diagnosis, use of immunosuppressants, or Human Immunodeficiency Virus (HIV) infection were exclusion criteria for CG, though other well-controlled diseases were allowed.
Exclusion criteria for all participants were: acute febrile condition or symptoms suggestive of COVID-19 at baseline, previous anaphylactic response to vaccine components, demyelinating disease, severe heart failure (class III or IV), history of having received blood transfusion ≤6-months before study entry, inactivated virus vaccine ≤ 14-days before study entry, history of live virus vaccine ≤4-weeks before study entry, individuals who did not consent to participate in the study, hospitalized patients, prior immunization with any SARS-CoV-2 vaccine, and pre-vaccination positive COVID-19 Serology (anti-S1/S2 IgG) and/or NAb for immunogenicity analysis of naïve-AAV patients.
Vaccination and blood collection protocol
The study protocol consisted of five in-person visits that occurred on February 9th‒10th 2021 (D0 ‒ first vaccine dose and blood collection), on March 9th‒10th 2021 (D28 ‒ second vaccine dose and blood collection), on April 19th, 2021 (D69 – blood collection), on September 18th, 2021 (D210 – 3rd vaccine dose and blood collection) and on October 19th, 2021 (D240 – blood collection) at the Hospital Convention Center (São Paulo, Brazil). The vaccination protocol for all participants included three doses of ready-to-use syringes loaded with the CoronaVac vaccine (Sinovac Life Sciences, Beijing, China, batch #20200412), consisting of 3 μg in 0.5 mL of β-propiolactone inactivated SARS-CoV-2 (resultant from the CN02 strain of SARS-CoV-2 grown in African green monkey kidney cells ‒ Vero 25 cells) with aluminum hydroxide as an adjuvant and applied in the deltoid muscle.
Outcomes
The primary outcome was immunogenicity assessed by two co-primary endpoints: seroconversion of anti-SARS-CoV-2 S1/S2 IgG and the presence of NAb after the second vaccine dose (D69). Secondary outcomes were Geometric Mean Titers (GMT) and neutralizing activity at D69, immunogenicity parameters at D28, D210 (day of 3rd dose) and D240 (30 days after 3rd dose), and safety related to the vaccine doses. Additionally, factors associated with anti-SARS-Cov-2 IgG seroconversion and NAb positivity were evaluated.
To assess these outcomes, blood samples (20 mL) from all participants were obtained at all in-person visits.
Anti-S1/S2 SARS-CoV-2 antibodies
Human IgG antibodies against the S1 and S2 proteins of SARS-CoV-2 were measured by chemiluminescent immunoassay as described previously.14 Seroconversion Rates (SC) were measured by positive serology (≥15.0 UA/mL) after vaccination, considering that only patients with pre-vaccination negative serology were included. GMT and 95% Confidence Intervals (95% CI) of these antibodies were also determined at all-time points, attributing the value of 1.9 UA/mL (half of the lower limit of quantification 3.8 UA/mL) to above lower levels (< 3.8 UA/mL). The Factor Increase in GMT (FI-GMT) is the ratio of the GMT after immunization to the GMT before immunization which identifies the increase in titers.
Neutralizing antibodies (NAb)
The SARS-CoV-2 NAb analysis was performed according to manufacturer instructions using an sVNT Kit (GenScript, Piscataway, NJ, USA), as described previously.11 The samples were cataloged as “positive” (inhibition ≥30%) or “negative” (inhibition < 30%) according to the manufacturer.19 The frequency of seropositivity was calculated at all-time points. Medians (interquartile range) of the percentage of neutralizing activity were only measured for seropositive samples at all-time points.
Vaccine adverse events
Adverse Events (AE) were carefully followed throughout the study. Patients and CG were advised to report any adverse events of the vaccine and they received on D0 and at all visits a standardized diary for local and systemic manifestations. AE severity was classified according to WHO criteria.20
In addition, incident COVID-19 cases were assessed in all subjects with instruction to notify any symptom associated or not with COVID-19 (by telephone, smartphone instant messaging, or email) and the disease was confirmed by RT-PCR test. Independent vaccine experts monitored the study regarding adverse events for data safety.
Disease assessment
Demographic, clinical, and therapeutic data of the participants in the AAV group were recorded and compared regarding seroconversion and NAb positivity. The Birmingham Vasculitis Activity Score (BVAS) – version 321 was assessed in all patients at baseline (at the most recent outpatient visit before vaccination) and after the second dose of the vaccine (at the next outpatient visit), to analyze the possible impact of disease activity in the vaccine immunogenicity, as well as the potential risk of the vaccine to trigger disease activity. Disease-related damage, expressed by the Vasculitis Damage Index (VDI), was also included in the analysis as a potential impact on seroconversion and production of Nab.22 Since the beginning of the trial there was no evidence-based information on the effect of immunosuppression on vaccine immunogenicity, the protocol did not include tapering or discontinuation of any treatment, and doses of prednisone and other immunosuppressants were maintained as directed by the disease status.
Statistical analysis
Categorical variables were presented as numbers (percentage) and compared using the chi-square or Fisher's exact tests, as appropriate, and McNemar´s test for before and after comparisons in the same group. Continuous general data were presented as medians (minimum and maximum values) and compared using the Mann-Whitney test for intergroup comparisons and Wilcoxon signed rank test for before and after comparisons in the same group. Data regarding IgG titers were analyzed using Analysis of Variance (ANOVA) with repeated measures and two factors (two groups ‒ vasculitis versus CG ‒ at specified time points), followed by Bonferroni's multiple comparisons at neperian logarithm (ln)-transformed data. For patients with AAV, multivariate logistic regression analyses were performed using dependent variables SC or the presence of NAb at D69 (primary endpoints), and as independent variables with p < 0.05 in each univariate analysis. Statistical significance was defined as p < 0.05. All statistical analyses were performed using Statistical Package for the Social Sciences, version 20·0 (IBM-SPSS for Windows. 20.0. Chicago, IL, USA).
Results
Participants
Fifty-three AAV patients (GPA [n = 36], EGPA [n = 10] and MPA [n = 7]) and CG (n = 106) were initially included in the study and received two doses of CoronaVac vaccine (Fig. 1 ). AAV patients and CG were balanced for sex and age. Patients with AAV had a median disease duration of 7 years (range: 1 to 31). Comorbidities were more frequent in the AAV group (83%, p = 0.0007), with systemic arterial hypertension (50.9%) being the most prevalent comorbidity. A total of 19 AAV patients were using prednisone (35.8%), 71.7% of patients were under immunosuppressive drugs, and 15.1% individuals were under rituximab treatment (Table 1 ).Figure 1 Flow-chart of the present study.
Figure 1
Table 1 Baseline characteristics of ANCA-Associated Vasculitis (AAV) vasculitis patients and Controls (CG).
Table 1 AAV (n = 53) CG (n = 106) p-value
Demographics
Current age, years 52 (24‒75) 52 (24‒78) 0.770
Age at diagnosis, years 42 (3‒71) ‒ ‒
Disease duration, years 7 (1‒31) ‒ ‒
Female sex 31 (58.5) 62 (58.5) 1.000
Caucasian race 33 (62.3) 52 (49) 0.131
BMI, Kg/m2 28.1 (18.4‒38.5) 26.6 (17.3‒39.1) 0.142
Comorbidities 44 (83) 59 (55.7) 0.0007
Systemic arterial hypertension 27 (50.9) 33 (31.1) 0.023
Diabetes mellitus 8 (15.1) 18 (17) 0.824
Dyslipidemia 10 (18.9) 11 (10.4) 0.144
Obesity 17 (66) 29 (52.8) 0.580
Chronic cardiomyopathy 3 (5.7) 4 (4.7) 0.687
Chronic renal disease 6 (11.3) 0 0.001
Current smoking 2 (3.8) 9 (8.5) 0.339
Chronic obstructive pulmonary disease 2 (3.8) 1 (1,9) 0.258
Asthma 7 (13.2) 4 (3.8) 0.043
Interstitial lung disease 3 (5.7) 0 0.036
Pulmonary hypertension 1 (1.9) 0 0.333
Hematologic disease 0 0 ‒
Hepatic disease 1 (1.9) 0 0.333
Current cancer 1 (1.9) 0 0.333
Stroke 1 (1.9) 0 0.333
Current tuberculosis 0 0 ‒
HIV 0 0 ‒
Vasculitis Score ‒
BVAS 0 (0‒8) ‒ ‒
VDI 3 (0‒9) ‒ ‒
Current therapy ‒
Prednisone 19 (35.8) ‒ ‒
Immunosuppressive drugs 38 (71.7) ‒
Methotrexate 14 (26.4) ‒ ‒
Azathioprine 16 (30.2) ‒ ‒
Mycophenolate mofetil 5 (9.4) ‒ ‒
Cyclophosphamide 3 (5.7) ‒ ‒
Leflunomide 1 (1.9) ‒ ‒
Biologic therapy ‒
Rituximab 8 (15.1) ‒ ‒
Results are expressed as median (minimum and maximum values) and n (%).
For immunogenicity analysis 10 AAV patients were excluded due to: positive pre-vaccination COVID-19 serology (n = 7), hospitalization (n = 1) and loss of follow-up (n = 2). Thirteen individuals from the naïve-CG were also excluded from the immunogenicity analysis due to positive pre-vaccination COVID-19 serology. The final group comprised 43 AAV patients and 93 CG (Fig. 1). These 43 AAV patients and 93 CG were further invited to participate in the decay and booster dose extension protocol. Only 32 patients completed the protocol and the authors subsequently randomly selected among the CG group 32 sex and age-balanced (± 5-years) in a 1:1 ratio (1 control: 1 patient) (Fig. 1).
Immunogenicity
Anti-SARS-CoV-2 IgG antibodies production in 43 naïve-AAV and 93 naïve-CG groups at D69
The humoral response to CoronaVac is shown in Table 2 . Analysis of the SARS-CoV-2 S1/S2 IgG response revealed a moderate seroconversion rate in patients with AAV six weeks (D69) after the second vaccine dose, although lower compared to CG (65.1% vs. 96.8%, p = 0.0001). GMT and FI-GMT were also lower in patients with AAV compared to CG (p < 0.001 and p < 0.001, respectively).Table 2 Seroconversion rates and anti-SARS-CoV-2 S1/S2 IgG titers before and after the first and second doses of CoronaVac vaccination in ANCA-Associated Vasculitis (AAV) patients and Controls (CG).
Table 2 SC GMT (AUmL−1) FI-GMT
D28 D69 D0 D28 D69 D0 to D28 D0 to D69
AAV (n = 43) 5 (11.6) 28 (65.1) 2.2 (2.0‒2.4) 4.4 (3.2‒6.0) 21.3 (13.2‒34.5) 2.0 (1.53‒2.67) 9.8 (6.1‒15.8)
CG (n = 93) 32 (34.8) 90 (96.8) 2.4 (2.1‒2.6) 11.2 (8.4‒14.9) 67.7 (58.3‒78.6) 4.7 (3.7‒6.0) 28.7 (24.2‒34.1)
p (AAV vs. CG) 0.0065 0.0001 >0.999 <0.001 <0.001 <0.001 <0.001
SC is defined as post-vaccination titer ≥15 AU mL−1 by indirect ELISA, LIAISON SARS-CoV-2 S1/S2 IgG. Frequencies of SC are presented as number (%) and were compared using a two-sided chi-square test between AVV and CG at prespecified time points (D28 and D69). IgG antibody titers and FI-GMT are expressed as geometric means with 95% CI. Data regarding IgG titers were analyzed using ANOVA with repeated measures and two factors (two groups (vasculitis vs. CG) at three time points (D0, D28 and D69), followed by Bonferroni's multiple comparisons at ln-transformed data. The behavior of IgG titers was different for AAV and CG groups between D28 and D69: mean titers increased at each time point for AAV and CG (p < 0.001). FI-GMT values were compared using the Mann-Whitney U-test for intergroup comparisons in ln-transformed data at prespecified time points (D28 and D69). All analyses were two-sided.
NAb positivity in 41 naïve-AAV patients and 93 naïve-CG groups at D69
After the 2nd vaccine dose, more than half of the AAV patients had positive NAb, a frequency lower than the CG group (53.7% vs. 80.6%, p = 0.001). Of note, the median of NAb activity was similar after the second dose (69.3 [47.2‒90.0] vs. 61.2 [46.3-80.1], p = 0.240) in patients and CG (Table 3 ).Table 3 Frequency of neutralizing antibodies (NAb) and median percentage of neutralizing activity in positive cases, after the first and second doses of CoronaVac vaccination in ANCA-associated vasculitis (AAV) patients in comparison to controls (CG).
Table 3 D28 D69
Subjects with positive Nab, n (%) Neutralizing activity (%) Subjects with positive Nab, n (%) Neutralizing activity (%)
Median (interquartile range) Median (interquartile range)
AAV (n = 43) 5 (11.6) 62.6 (53.2‒65.2) 22 (53.7)a 69.3 (47.2‒90.0)
CG (n = 93) 36 (40) 49.9 (35.9‒80.4) 75 (80.6) 61.2 (46.3‒80.1)
p (AAV vs. CG) 0.001 0.952 0.001 0.240
Frequencies of subjects with positive NAb are expressed as number (%). Positivity for NAb was defined as neutralizing activity ≥ 30% (cPass sVNT Kit). Data were compared using a two-sided Chi-Square test between AAV and CG at prespecified time points (D28 and D69). Percentage of neutralizing activity among subjects with positive NAb is expressed as median (IQR). Data were compared using a two-sided Mann-Whitney U-test for comparison between AAV and CG, at prespecified time points (D28 and D69).
a In D69, AAV patients n = 41 due to unavailability of two NAb samples.
Factors associated with seroconversion and NAb positivity among naïve-AAV patients at D69
Analyzing the possible impact of disease activity on the immunogenicity of the vaccine, the higher frequency of seroconversion rates at D69 in naïve AAV patients with BVAS activity score = 0 compared to those with BVAS > 0 did not reach statistical significance (74.2% vs. 41.7%, p = 0.074). GMT was comparable in both groups at D69 (Table 4 ). With regard to the possible influence of vaccine on disease activity there was no change in this parameter with similar BVAS levels at baseline and after the 2nd vaccine dose (0.81 ± 1.64 vs. 1.07 ± 2.66, p = 0.71). GMT BVAS was also comparable at baseline and after the 2nd vaccine dose (0.81 ± 1.64 vs. 1.07±2.66, p = 0.71). There was no difference in the seroconversion rate among patients with AAV regarding baseline VDI (3.19 ± 0.35 with seroconversion vs. 3.17±0.37 with no seroconversion, p = 0.975).Table 4 Seroconversion rates at D69, anti-SARS-CoV-2 S1/S2 IgG titers comparing vasculitis activity score (BVAS = 0 vs. BVAS > 0) and frequency of NAb and median percentage of neutralizing activity after second dose (D69) of CoronaVac vaccination in ANCA-Associated Vasculitis (AAV) patients.
Table 4 SC GMT (AU mL−1)
D69
BVAS baseline = 0 (n = 31) 23 (74.2) 25.5 (14.6‒44.5)
BVAS baseline > 0 (n = 12) 5 (41.7) 14.2 (4.7‒42.9)
p (BVASbaseline = 0 vs BVASbaseline > 0) 0.074 0.330
D69
Subjects with positive NAb, n (%) Neutralizing activity (%) median (IQR)
BVAS baseline = 0 (n = 31) 15 (48.4) 68.2 (44.1–89.7)
BVAS baseline > 0 (n = 12) 7 (58.3) 74.3 (63.3–90.6)
p (BVASbaseline = 0 vs BVASbaseline > 0) 0.736 0.587
Frequency of subjects with seroconversion is expressed in n (%). Titers of IgG antibodies are expressed in geometric means with 95% CI. BVAS, Birmingham Vasculitis Activity Score; SC, Seroconversion; Nab, Neutralizing antibodies; GMT, Geometric Mean Titers.
Regarding treatment, the frequencies of immunosuppressive drugs (93.3% vs. 53.3%, p = 0.015) and mycophenolate mofetil (20% vs. 0%, p = 0.037) were significantly higher in AAV patients without SC compared to those with SC, and a trend of more frequent prednisone use (60% vs. 28.6%, p = 0.057). Negative NAb in AAV patients was associated with more frequent use of prednisone (57.9% vs. 18.2%, p = 0.015) and immunosuppressive drugs (84.2% vs. 55.0%, p = 0.046) compared to those with positive NAb (Table 5 ). Eight patients who had received rituximab within 6 months before the first dose of the vaccine were considered to be on rituximab treatment. The median cumulative dose of rituximab was 4.5g (minimum 2, maximum 8), and the median interval between the last rituximab cycle and the first vaccine dose was 2 months (minimum 0, maximum 5). Logistic regression analysis models showed that only the use of prednisone was associated with lower seroconversion (OR = 0.20, 95% CI 0.05‒0.86, p = 0.030) and lower NAb positivity (OR = 0.20, 95% CI 0.05‒0.88, p = 0.034).Table 5 Baseline characteristics of AAV patients with and without Seroconversion (SC) for anti-SARS-CoV-2 S1/S2 IgG antibodies and with and without positivity of Neutralizing Antibodies (NAb) after two doses of CoronaVac vaccination (day 69).
Table 5 Vasculitis patients with SC (n = 28) Vasculitis patients without SC (n = 15) p-value Vasculitis patients with NAb (n = 22) Vasculitis patients without NAb (n = 19) p-value
Demographics
Current age, years per median (mn ± max) 52.6±13.4 53.0±12.5 0.926 54.1±15.2 51.1±10.8 0.479
Current age > 60 years 6 (21.4) 5 (33.3) 0.473 8 (36.4) 3 (15.8) 0.173
Female sex 17 (60.7) 8 (53.3) 0.750 14 (63.6) 11 (57.9) 0.757
Caucasian race 17 (60.7) 11 (73.3) 0.512 15 (68.2) 12 (63.1) 0.754
Current therapy
Prednisone 8 (28.6) 9 (60) 0.057 4 (18.2) 11 (57.9) 0.015
Prednisone dose, mg 5 (1.7‒20) 10 (5‒40) 0.234 0 (0‒20) 1.7 (0‒40) 0.280
Prednisone ≥ 20 mg/day 2 (7.1) 4 (26.7) 0.161 3 (13.7) 2 (10.5) 1.000
Immunosuppressive drugs 15 (53.6) 14 (93.3) 0.015 11 (50) 16 (84.2) 0.046
Methotrexate 8 (28.6) 4 (26.6) 1.000 6 (27.3) 5 (26.3) 1.000
Azathioprine 7 (25) 5 (33.3) 0.723 5 (22.7) 7 (36.8) 0.493
Mycophenolate mofetil 0 (0) 3 (20) 0.037 0 (0) 3 (15.8) 0.091
Cyclophosphamide 0 (0) 2 (13.3) 0.116 0 (0) 1 (5.3) 0.463
Leflunomide 0 (0) 1 (6.6) 0.349 0 1 (5.3) 0.463
Biologic therapy
Rituximab 3 (10.7) 5 (33.3) 0.069 1 (4.5) 7 (36.8) 0.016
Results are expressed in median (minimum and maximum values) and n (%). SC, Seroconversion defined as a positive serology (IgG titer ≥15 AU/mL) for anti-SARS-CoV-2 S1/S2 IgG antibodies after vaccination (Indirect ELISA, LIAISON® SARS-CoV-2 S1/S2 IgG, DiaSorin, Italy). Positivity for Nabs defined as a neutralizing activity ≥ 30% (cPass sVNT Kit, GenScript, Piscataway).
Six-months (D210) immunogenicity decay in 32 AAV and 32 CG groups after the second vaccine dose
Antibody decay in six months was observed with a trend of 15.7% reduction in IgG seropositivity for 32 AAV patients (68.8% vs. 53.1%, p = 0.074) and 18.7% for 32 GC (100% vs. 81.3%, p = 0.041). GMT titers also had a significant reduction of 39.2% in AAV patients (26.5 [14.9–46.9] vs. 16.1 [8.7–29.9], p = 0.010) and an even more striking decrease of 54.8% for the CG (83.7 [69.3–101.3] vs. 37.8 [25.0–57.2], p < 0.001). For the NAb positivity the 6-month decrease in the rate for the 32 AAV patients (59.4% vs. 40.6%, p = 0.114) was not significant, whereas for the CG a 62.5% reduction was observed in CG (90.6% vs. 28.1%, p < 0.001).
Booster dose immunogenicity in 32 AAV patients and 32 CG from D210 to D240
The booster dose resulted in a 21.9% increase in anti-SARS-CoV-2 IgG antibodies positivity in AAV patients (53.1% vs. 75%, p = 0.023) and 18.7% in CG (81.3% vs. 100%, p = 0.041). AAV patients remained lower than the CG group at D240 (75% vs. 100%, p = 0.005). A 4.3-fold augmentation in GMT was observed for the AAV group (16.1 [8.7‒29.9] vs. D240 70.0 [34.8‒140.7] p < 0.0001) and 6.3-fold for CG (37.8 [25.0‒57.2] vs. D240 237.8 [195.8‒288.6] p < 0.0001). For NAb positivity the same pattern was observed but with a more relevant increment of 34.4% in the AAV patients (40.6% vs. 75%, p = 0.006) and 68.8% in CG (28.1% vs. 96.9%, p < 0.0001). NAb activity increased 1.4-fold for AAV patients (59.2 [48.9‒75.3] vs. D240 82.1 [57.2‒95.7] p = 0.006) and no change was detected for the CG (86.3 [54.1‒95.0] vs. D240 82.5 [62.7‒97.1] p = 0.065)].
Vaccine tolerance and safety 53 AAV patients and 106 CG at D28 and D69
No serious adverse reaction was observed in both group and the events were similar between the former group and CG, except for the hospitalization of one patient in the AAV group on the date of his second vaccine dose, due to a urinary tract infection, which was not considered to be a vaccine-related adverse event. This patient was excluded from the study for not completing the vaccine protocol at the scheduled interval but had a complete recovery from the infection and subsequently completed the vaccination. After the first dose of CoronaVac, there was a higher prevalence of malaise (p = 0.007), myalgia (p = 0.021), and sneezing (p = 0.017) in AAV patients when compared with CG, and after the second dose, the events were similar the former group and CG (p = 0.696) (Supplementary Table 1). For booster dose, only mild AE was observed in 12 (37.5%) AAV patients and 6 (18.8%) CG (p = 0.111). There was 1 incident case of COVID-19 in an AAV patient during the study and 2 cases in the control group, all with mild symptoms and no need for hospitalization.
Discussion
ANCA-associated vasculitis patients are among the high-risk groups of SARS-CoV-2 serious infection and death. The present results show that two doses of the inactivated CoronaVac had moderate immunogenicity in naïve AAV patients, lower than the control group. AAV patients had a mild decrease in humoral response in six months and a good response with a booster dose. Furthermore, the authors showed that this vaccine was safe in this group of patients.
The immunogenicity in AAV patients was moderate and the inclusion of only naïve patients may have influenced this finding. In fact, the authors have previously demonstrated that naïve ARD and COVID-19 pre-exposed ARD patients have distinct dynamics of vaccine response, with a significantly lower antibody production in the former group.23 In spite of that, naïve AAV had a lower response compared to naïve Systemic Lupus (SLE) patients (70.2%) immunized with the same vaccine and also reduced when compared to the naïve CG. For the healthy control group, age and sex, known relevant factors to impair vaccine response, is not the likely explanation since groups were balanced for these parameters. With regard to SLE, the older age and the distinct sex distribution of AAV may have contributed to the reduced vaccine-induced antibody response in these patients.24 In addition, other and the authors have demonstrated that the major factor to influence vaccine response is therapy and in fact, the frequency of methotrexate and rituximab was higher in AAV patients than in lupus.9 , 11 Unexpectedly, neutralizing activity of anti-SARS-CoV-2 antibodies was comparable to the control group, a reassuring finding since this parameter was reported to be a more precise maker of disease protection than anti-SARS-CoV-2 IgG positivity.25
Regarding factors that influence vaccine response, the authors observed herein, that the use of prednisone is a major contributor to the decreased immunogenicity in AAV patients. A similar finding was reported for mRNA and inactivated vaccines in general ARD patients,9 , 11 , 26 , 27 but none evaluated specifically AAV patients. Regarding the impact of drugs in the present study, mycophenolate mofetil appeared to reduce seroconversion, while rituximab impacted mainly NAb production (Table 5). Accordingly, rituximab and methotrexate, two medications used very often to treat AAV, were reported to decrease vaccine response.28, 29, 30 However, this impact was not confirmed in the multivariate analysis, probably due to the small representation of the patients under these therapies. Reinforcing this possibility, in the large CoronavRheum trial, of which the present study is part, the multivariate analysis revealed that methotrexate, mycophenolate, and rituximab had a deleterious effect on vaccine response.11
Immunogenicity waning is a major concern for these immunocompromised patients and a 65% reduction in IgG levels and 70% for neutralizing antibody concentrations in 6-months was reported for participants with immunosuppression immunized with mRNA vaccine.31 A lower reduction in IgG levels (38%) and neutralizing antibody activity (54%) was observed by the group for a large overall ARD population with 818 patients immunized with inactivated vaccine.32 The pattern of antibody decay for AAV was distinct with a non-significant reduction in NAb activity (15.1%) in 6-months.
The booster dose was effective in increasing seroconversion and NAb positivity in both groups 6-months after the first vaccine dose, reinforcing the recommendation of this strategy for this group of patients, as suggested by the Center of Disease Control.33 Anti-SARS-CoV-2 IgG positivity was, however, lower than observed in overall ARD patients after the third dose34 probably related to higher frequencies of patients under rituximab in the present study.
Importantly, the authors demonstrated that the vaccine was safe in this group of patients, with no serious adverse effects and a general safety profile compared with the control group. Duran and colleagues35 showed that infection by the SARS-CoV-2 has triggered a few cases of AAV in previously healthy individuals. The present data suggest that the vaccine does not worsen disease control in patients with pre-existing vasculitis. Another important observation was that disease activity, measured by the BVAS, did not seem to impact vaccine immunogenicity.
The main limitation of the present study is the small sample size related to the rarity of the disease and the lack of cellular immune response assessment.
In conclusion, this study provides novel data on moderate immunogenicity and an excellent safety profile of CoronaVac in AAV patients. A six-months mild antibody waning was observed with a good response to the booster dose, although levels remained lower than CG.
Conflicts of interest
The authors declare no conflicts of interest.
Appendix Supplementary materials
Image, application 1
Funding
This study was sponsored by grants from Fundação de Amparo à Pesquisa do Estado de São Paulo (#2015/03756–4 to C.A.S. S.G.P., N.E.A., E.B; #2019/17272-0 to L.V.K.K; #2020/09367-8, #2018/09937-9 to V.A.O.M and #2021/08455-3 to CCMFM), Conselho Nacional de Desenvolvimento Científico e Tecnológico (#305556/2017-7 to R.M.R.P; #304984/2020-5 to C.A.S.; #303379/2018-9 to S.K.S and #305242/2019-9 to E.B), B3-Bolsa de Valores do Brasil and Instituto Todos pela Saúde (ITPS 01/2021, C1313 to E.B., C.A.S., N.E.A. and S.G.P.). Instituto Butantan supplied the study product and had no other role in the trial.
Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.clinsp.2022.100150.
==== Refs
References
1 Hosseini ES Kashani NR Nikzad H Azadbakht J Bafrani HH Kashan HH. The novel coronavirus Disease-2019 (COVID-19): Mechanism of action, detection and recent therapeutic strategies Virology 551 2020 1 9 33010669
2 Autrán-Gómez AM Favorito LA. The Social, economic and sanitary impact of COVID-19 Pandemic Int Braz J Urol 46 suppl. 1 2020 3 5
3 Gavriatopoulou M Ntanasis-Stathopoulos I Korompoki E Fotiou D Migkou M Tzanninis I-G Emerging treatment strategies for COVID-19 infection Clin Exp Med 21 2 2021 167 179 33128197
4 Krammer F. SARS-CoV-2 vaccines in development Nature 586 7830 2020 516 527 32967006
5 Ita K. Coronavirus Disease (COVID-19): Current Status and Prospects for Drug and Vaccine Development Arch Med Res 52 1 2021 15 24 32950264
6 Tumban E. Lead SARS-CoV-2 Candidate Vaccines: Expectations from Phase III Trials and Recommendations Post-Vaccine Approval Viruses 13 1 2020 54 33396343
7 Tanriover MD Doğanay HL Akova M Güner HR Azap A Akhan S CoronaVac Study Group. Efficacy and safety of an inactivated whole-virion SARS-CoV-2 vaccine (CoronaVac): interim results of a double-blind, randomised, placebo-controlled, phase 3 trial in Turkey Lancet 398 10296 2021 213 222 34246358
8 Simon D Tascilar K Fagni F Krönke G Kleyer A Meder C SARS-CoV-2 vaccination responses in untreated, conventionally treated and anticytokine-treated patients with immune-mediated inflammatory diseases Ann Rheum Dis 80 10 2021 1312 1316 33958324
9 Furer V Eviatar T Zisman D Peleg H Paran D Levartovsky D Immunogenicity and safety of the BNT162b2 mRNA COVID-19 vaccine in adult patients with autoimmune inflammatory rheumatic diseases and in the general population: a multicentre study Ann Rheum Dis 80 10 2021 1330 1338 34127481
10 Pablos JL Galindo M Carmona L Lledó A Retuerto M Blanco R Clinical outcomes of hospitalised patients with COVID-19 and chronic inflammatory and autoimmune rheumatic diseases: a multicentric matched cohort study Ann Rheum Dis. 79 12 2020 1544 1549 32796045
11 Medeiros-Ribeiro AC Aikawa NE Saad CGS Yuki EFN Pedrosa T Fusco SRG Immunogenicity and safety of the CoronaVac inactivated vaccine in patients with autoimmune rheumatic diseases: a phase 4 trial Nat Med 27 10 2021 1744 1751 34331051
12 Gao YD Ding M Dong X Zhang J-J Azkur AK Azkur D Risk factors for severe and critically ill COVID-19 patients: A review Allergy 76 2 2021 428 455 33185910
13 Bruchfeld A. The COVID-19 pandemic: consequences for nephrology Nat Rev Nephrol 17 2 2021 81 82 33257872
14 Jennette JC Falk RJ Bacon PA Cid MC Ferrario F Flores-Suarez LF 2012 revised International Chapel Hill Consensus Conference Nomenclature of Vasculitides Arthritis Rheum. 65 1 2013 1 11 23045170
15 Harris PA Taylor R Thielke R Payne J Gonzalez N Conde JG. Research electronic data capture (REDCap) ‒ A metadata-driven methodology and workflow process for providing translational research informatics support J Biomed Inform 42 2 2009 377 381 18929686
16 Harris PA Taylor R Minor BL Elliott V Fernandez M O'Neal L The REDCap consortium: Building an international community of software platform partners J Biomed Inform 95 2019 103208
17 Leavitt RY Fauci AS Bloch DA Michel BA Hunder GG Arend WP The American College of Rheumatology 1990 criteria for the classification of Wegener's granulomatosis Arthritis Rheum 33 8 1990 1101 1107 2202308
18 Masi AT Hunder GG Lie JT Michel BA Bloch DA Arend WP The American College of Rheumatology 1990 Criteria for the Classification of Churg-Strauss Syndrome (Allergic Granulomatosis and Angiitis) Arthritis Rheum 33 8 1990 1094 1100 2202307
19 Taylor SC Hurst B Charlton CL Bailey A Kanji JN McCarthy MK A new SARS-CoV-2 dual-purpose serology test: highly accurate infection tracing and neutralizing antibody response detection J Clin Microbiol 59 4 2021 e02438-20
20 World Health Organization. WHO Draft Guidelines for Adverse Event Reporting and Learning Systems. 2021. https://apps.who.int/iris/bitstream/handle/10665/69797/WHO-EIP-SPO-QPS-05.3-eng.pdf?sequence=1&isAllowed=y.
21 Mukhtyar C Lee R Brown D Carruthers D Dasgupta B Dubey S Modification and validation of the Birmingham Vasculitis Activity Score (version 3) Ann Rheum Dis 68 2009 1827 1832 19054820
22 Exley AR Bacon PA Luqmani RA Kitas GD Gordon C Savage CO Development and initial validation of the vasculitis damage index for the standardized clinical assessment of damage in the systemic vasculitides Arthritis Rheum 40 1997 371 380 9041949
23 Aikawa NE Kupa LVK Pasoto SG Medeiros-Ribeiro AC Yuki EFN Saad CGS Immunogenicity and safety of two doses of the CoronaVac SARS-CoV-2 vaccine in SARS-CoV-2 seropositive and seronegative patients with autoimmune rheumatic diseases in Brazil: a subgroup analysis of a phase 4 prospective study Lancet Rheumatol 4 2 2022 e113 e124 34901885
24 Yuki EFN Borba EF Pasoto SG Seguro LP Lopes M Saad CGS Impact of distinct therapies on antibody response to SARS-CoV-2 Vaccine in systemic lupus erythematosus Arthritis Care Res (Hoboken) 74 4 2022 562 571 34806342
25 Khoury DS Cromer D Reynaldi A Schlub TE Wheatley AK Juno JA Neutralizing antibody levels are highly predictive of immune protection from symptomatic SARS-CoV-2 infection Nat Med 27 7 2021 1205 1211 34002089
26 Bugatti S De Stefano L Balduzzi S Greco MI Luvaro T Cassaniti I Methotrexate and glucocorticoids, but not anticytokine therapy, impair the immunogenicity of a single dose of the BNT162b2 mRNA COVID-19 vaccine in patients with chronic inflammatory arthritis Ann Rheum Dis 80 12 2021 1635 1638 34172502
27 Deepak P, Kim W, Paley MA, Yang M, Carvidi AB, El-Qunni AA, et al. Glucocorticoids and B Cell Depleting Agents Substantially Impair Immunogenicity of mRNA Vaccines to SARS-CoV-2. medRxiv [Preprint]. 2021:2021.04.05.21254656.
28 Friedman MA Curtis JR Winthrop KL. Impact of disease-modifying antirheumatic drugs on vaccine immunogenicity in patients with inflammatory rheumatic and musculoskeletal diseases Ann Rheum Dis 80 10 2021 1255 1265 34493491
29 Spiera R Jinich S Rituximab Jannat-Khah D. but not other antirheumatic therapies, is associated with impaired serological response to SARS- CoV-2 vaccination in patients with rheumatic diseases Ann Rheum Dis 80 10 2021 1357 1359 33975857
30 Mahil SK Bechman K Raharja A Domingo-Vila C Baudry D Brown MA The effect of methotrexate and targeted immunosuppression on humoral and cellular immune responses to the COVID-19 vaccine BNT162b2: a cohort study Lancet Rheumatol 3 9 2021 e627 e637 34258590
31 Levin EG Lustig Y Cohen C Fluss R Indenbaum V Amit S Waning Immune Humoral Response to BNT162b2 Covid-19 Vaccine over 6 Months N Engl J Med 385 24 2021 e84 34614326
32 Bonfa E, Silva C, Medeiros-Ribeiro A, Kupa L, Yuki E, Pasoto SS, et al. Anti-SARS-CoV-2 immunogenicity decay and incident cases six months after Sinovac-CoronaVac inactivated vaccine in autoimmune rheumatic diseases patients: phase 4 prospective trial. 23 November 2021. PREPRINT (Version 1) available at Research Square https://doi.org/10.21203/rs.3.rs-1054476/v1.
33 Mbaeyi S, Oliver SE, Collins JP, et al. Morbidity and Mortality Weekly Report The Advisory Committee on Immunization Practices’ Interim Recommendations for Additional Primary and Booster Doses of COVID-19 Vaccines-United States, 2021. 2021. https://www.cdc.gov/vaccines/hcp/acip-recs/vacc-specific/covid-19.html.
34 Aikawa NE Kupa LVK Medeiros-Ribeiro AC Saad CGS Yuki EFN Pasoto SG Increment of immunogenicity after third dose of a homologous inactivated SARS-CoV-2 vaccine in a large population of patients with autoimmune rheumatic diseases Ann Rheum Dis 81 7 2022 1036 1043 35277389
35 Izci Duran T Turkmen E Dilek M Sayarlioglu H Arik N. ANCA-associated vasculitis after COVID-19 Rheumatol Int 41 8 2021 1523 1529 34100115
| 0 | PMC9705212 | NO-CC CODE | 2022-12-16 23:21:38 | no | Clinics (Sao Paulo). 2023 Nov 29 January-December; 78:100150 | utf-8 | Clinics (Sao Paulo) | 2,022 | 10.1016/j.clinsp.2022.100150 | oa_other |
==== Front
Journal of Purchasing and Supply Management
1478-4092
1478-4092
The Authors. Published by Elsevier Ltd.
S1478-4092(22)00028-0
10.1016/j.pursup.2022.100773
100773
Article
One crisis, different paths to supply resilience: The case of ventilator procurement for the COVID-19 pandemic
Dube Nonhlanhla a∗
Li Qiujun a
Selviaridis Kostas a
Jahre Marianne b
a Lancaster University, Dept. of Management Science, Lancaster, United Kingdom
b BI Norwegian Business School, Dept. of Accounting and Operations Management, Oslo, Norway
∗ Corresponding author.
12 5 2022
12 2022
12 5 2022
28 5 100773100773
9 8 2021
16 4 2022
6 5 2022
© 2022 The Authors
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
This research explores supply resilience through an equifinality lens to establish how buying organizations impacted differently by the same extreme event can strategize and all successfully secure supply. We conduct case study research and use secondary data to investigate how three European governments sourced for ventilators during the first wave of COVID-19. The pandemic had an unprecedented impact on the ventilator market. It disrupted already limited supply and triggered a demand surge. We find multiple paths to supply resilience contingent on redundant capacity and local sourcing options at the pandemic's onset. Low redundancy combined with limited local sourcing options is associated with more diverse strategies and flexibility. The most notable strategy is spurring supplier innovation by fostering collaboration among actors in disparate industries. High redundancy combined with multiple local sourcing options is associated with more focused strategies and agility. One (counter-intuitive) strategy is the rationalization of the supply base.
Keywords
Supply resilience
Extreme events
Equifinality
COVID-19
Public procurement
==== Body
pmc1 Introduction
Extreme events, such as the COVID-19 global pandemic, are difficult to predict and present unique risks to supply chains (Knight et al., 2022; Van Hoek and Loseby, 2021; Vanpoucke and Ellis, 2019). Such events threaten livelihoods and the continuity of supply due to their unprecedented impact in scale, duration, and scope (Craighead et al., 2020; Van Hoek, 2020). Buying organizations need to be resilient to overcome their disruptive effects (Tukamuhabwa et al., 2015; Walker, 2020). While “supply chain resilience” refers to supply chain-wide decisions and activities (e.g., Scholten et al., 2019), in this research we use the term “supply resilience” to highlight the focus on buying organizations and how they adapt to secure supply following disruptions caused by extreme events. Despite its widely recognized importance for dealing with major disruptions, the purchasing and supply chain management (SCM) literature is still grappling with what resilience entails (Hohenstein et al., 2015; Wieland and Durach, 2021; Wiedmer et al., 2021). We posit that, in part, this is because empirical research underplays the fact that investing in all types of resilience capabilities – encompassing at least 13 attributes and 84 managerial practices – is too costly (Ali et al., 2017). Buying organizations, therefore, make different trade-offs when allocating limited resources based on their priorities (Robinson and Sahin, 2006: in Mentzer et al., 2008). This leads to inherently varying (resilience) capabilities and vulnerabilities (Blackhurst et al., 2011; Pettit et al., 2013; Tukamuhabwa et al., 2017). By extension, buying organizations’ initial conditions, both in terms of challenges and opportunities, will differ at the onset of any extreme event. Therefore, considering the implications of different resource allocation decisions is crucial for understanding different paths to supply resilience.
This study explores the mechanisms underpinning different pathways to supply resilience given varying initial conditions at the onset of extreme events. We investigate two research questions: (1) How do the initial conditions of buying organizations following an extreme event influence the ways in which buyers employ sourcing strategy to respond to disruptions (i.e., response strategies)? (2) To achieve supply resilience, which response strategies are better aligned what initial conditions? To explore these relationships, we employ an equifinality lens. Equifinality relates to the situation whereby different organizations achieve similar performance outcomes through different strategies (Fernández and Kekäle, 2005; Jaspers, 2007; Katz and Kahn, 1978). To this end, we study how three governments with different initial conditions during the first wave of the COVID-19 pandemic in 2020 successfully secured supply for ventilators despite unprecedented demand surges globally and disrupted supply. We employ a qualitative, case-based research design to investigate the problem through an underexplored lens (i.e., resilience through an equifinality lens), address contextual complexity (Eisenhardt and Graebner, 2007), and enhance the quality of explanations for equifinal outcomes (Mills et al., 2013). Accordingly, we generate in-depth insights into how and why (Voss et al., 2002) buying organizations with different initial conditions successfully bridged an extraordinary demand-supply gap under exceptional circumstances.
Pre-COVID, the annual global demand for ventilators was stable at around 77,000. During the first-wave infections New York city, alone, needed an additional 33,000 ventilators (Netland, 2020). As ventilators are high-tech products, it was anticipated that already disrupted supply chains would not quickly match surging demand. In efforts to avoid shortages, governments applied different response strategies including agile procurement, fostering cross-sectoral collaboration and merging supply chains (Feizabadi et al., 2021; Fearne et al., 2021). The public procurement context is theoretically interesting because, unlike in private sector contracting, procurement professionals are typically constrained by rules, regulations, and norms that can hinder resilience, e.g., impede close collaboration with suppliers and discourage flexibility (Selviaridis and Spring 2022). At the same time, public procurement can be used strategically to help implement public policy goals such as innovation, and to coordinate response to emergency situations (Harland et al., 2021a; Selviaridis, 2021). In crisis periods such as the COVID-19 pandemic, governments have the power to override normal procurement processes, steer public-private collaboration and use more agile approaches to procurement of critical healthcare products and services (Harland et al., 2021b; Fearne et al., 2021).
Our study contributes to research at the intersection between strategic sourcing and supply resilience (e.g., Pereira et al., 2014) by exploring how procurement can enable adaptation to unprecedented changes. We also extend prior research addressing contingencies (e.g., Bode et al., 2011; Namdar et al., 2018; Roscoe et al., 2020; Wiedmer et al., 2021) and reconcile some conflicting empirical findings in the literature by showing that the most suitable response strategies depend on buying organizations’ initial conditions. To the best of our knowledge, this study is also the first to empirically investigate equifinality in public procurement. We add to research stressing the critical role of public procurement in responding to emergencies (e.g., Harland et al., 2021a) by unearthing different pathways leading to the same successful outcomes, despite varied approaches to allocation resources in this setting. Despite our focus on public procurement, our results are largely transferable to private-sector procurement settings given that governmental organizations and private-sector firms have some common characteristics as buying organizations e.g., level of procurement centralization. Consistent with the call of this Special Issue (Kähkönen et al., 2020), we further provide insights regarding how and why certain sourcing strategies foster innovation and enable rapid supply capacity development to bridge unprecedented demand-supply gaps.
The rest of this paper is organized as follows. The theoretical background is presented in Section 2 followed by the methodology in Section 3. Section 4 presents the key results of the study. The discussion and conclusions follow in Section 5.
2 Theoretical background
2.1 Supply chain resilience – A brief overview
The increasing occurrence and consequences of disruptions make resilience an important capability in supply chains (Brandon-Jones et al., 2014; Gunasekaran et al., 2015; Ponomarov and Holcomb, 2009). Despite differences in the understanding of what resilience entails (Wieland and Durach, 2021), most definitions stress the adaptive capability to quickly respond to unexpected disruptions, maintain some functionality, and recover to the original state or better (e.g., Ali et al., 2017; Mena et al., 2020; Sheffi and Rice, 2005). Even though recent works adopt new terminology for adaptation and transformation to a better state – e.g., Nikookar et al. (2021) introduce antifragility while Craighead et al. (2020) argue for transiliency – we view both terms as being captured in long-standing definitions of resilience.
Understanding resilience in the context of extreme events is essential for two reasons. Firstly, the required state before and after the disruption is often different (Ponomarov and Holcomb, 2009; Wieland and Durach, 2021). Thus, it is important to establish how organizations can adapt and/or transform themselves when extreme events bring inevitable change (Walker, 2020). Secondly, because investing in developing all resilience capabilities is expensive (Ali et al., 2017; Pettit et al., 2019) there are trade-offs to be made in the strategic allocation of resources (Jüttner and Maklan, 2011; Mentzer et al., 2008; Pereira et al., 2014). Subsequently, there can never be guarantees that risk mitigation measures will be sufficient to resist the unforeseen and/or unprecedented effects of extreme events. Indeed, some major vulnerabilities in most organizations’ supply chains became known during COVID-19 (Van Hoek and Loseby, 2021). Thus, a crucial question concerns how organizations can make different trade-offs and equally be resilient to the disruptive impact of extreme events.
2.2 Key resilience capabilities
The key resilience capabilities identified in prior literature are flexibility, redundancy, agility, collaboration, and visibility (Ali et al., 2017; Hohenstein et al., 2015). Since definitions of these capabilities differ (Rice and Caniato, 2003; Sheffi and Rice, 2005; Jüttner and Maklan, 2011), Table 1 presents those we adopt in this research. Resilience capabilities are interrelated and/or can be mutually reinforcing (e.g., Ali et al., 2017; Christopher and Peck, 2004; Pettit et al., 2013). Therefore, Table 1 also shows some of the identified linkages between them. Interestingly, we could not identify literature stating that flexibility contributes to other capabilities. Some definitions merge capabilities, e.g., incorporating agility in definitions of flexibility (Ali et al., 2017; Jüttner and Maklan, 2011). These observed interrelationships reinforce the idea of multiple pathways to resilience, as investing in one capability can enhance or lead to another.Table 1 Resilience capabilities and their definitions.
Table 1Resilience capability Definition and brief description Relationships between capabilities
Flexibility The ability of the supply chain to adapt by redeploying various resources in response to disruptions. Flexibility ensures adaptability.
Sources: Ali et al. (2017); Jüttner and Maklan (2011); Rice and Caniato (2003); Jüttner and Maklan (2011); Wallace and Choi (2011).
Redundancy The ability to reserve some resources for use if disruptions occur. For example, additional capacity or stock is often maintained before needed, even though it may not be used, to make up for the capacity loss that could be caused by disruptions. Redundancy ensures continuity of function when supply is disrupted.
Sources: Rice and Caniato (2003); Sheffi and Rice (2005). Some redundancy strategies contribute to flexibility and agility, e.g., using backup suppliers and building buffers for essential components, respectively.
Sources: Lee (2004); Xie et al. (2019).
Agility The ability of the supply chain to respond quickly to sudden and unexpected changes in demand and/or supply.
Sources: Christopher and Peck (2004); Lee (2004). Improves flexibility by accelerating processes for responding to disruptions.
Sources: Ali et al. (2017); Christopher and Peck (2004).
Collaboration The ability of supply chain members to work together effectively for their mutual benefit in the face of risk and uncertainty.
Sources: Jüttner and Maklan (2011); Pettit et al. (2010) Some collaborative practices improve flexibility and agility, e.g., information-sharing and mutual knowledge creation.
Sources: Christopher and Peck (2004); Scholten and Schilder (2015); Wieland and Wallenburg (2013).
Visibility The ability to access relevant information between or across supply chain tiers and is largely seen as a capability that enables the identification of root causes of supply chain issues.
Sources: Brandon-Jones et al. (2014); Azevedo et al., 2013. Visibility, like collaboration, improves situation awareness, warning strategies and recovery times. Although visibility may be less relevant for anticipating the manifestation and impact of extreme events, it provides a basis for collaboration and improves agility
Sources: Brandon-Jones et al. (2014); Christopher and Peck (2004); Christopher and Rutherford (2004); Vanpoucke and Ellis (2019).
2.3 Sourcing strategy and supply resilience
The strategic sourcing literature recognizes the unique impact of extreme events on supply (e.g., Arani et al., 2016; Mandal, 2020) and, hence, the need for buying organizations to be resilient. Although strategic sourcing decisions are enduring in nature, to achieve supply resilience they should also include supply risk mitigation strategies that increase response capabilities (Craighead et al., 2020; Vanpoucke and Ellis, 2019; Wieland and Durach, 2021).
Key decisions related to risk mitigation in strategic sourcing include supplier location, number of suppliers, buyer-supplier relationship types, and approaches to supplier development (Arani et al., 2016; Namdar et al., 2018; Pereira et al., 2014; Scholten and Schilder, 2015). Related sourcing practices are echoed in the resilience literature; for example, building a flexible supplier base, reserving excess capacity, information sharing, investing in suppliers’ capability to recover quickly from a disruption, and joint planning (Kochan and Nowicki, 2018; Namdar et al., 2018). Furthermore, practices such as cost-sharing or incorporating deductible elements (Erkoc and Wu, 2005), and revenue-sharing (Zeng and Xia, 2015) are essential. They improve collaboration and provide the means for suppliers to respond (Wiedmer et al., 2021; Zeng and Xia, 2015).
Resource scarcity implies the need to ensure that investments made match the risk level (Pettit et al., 2019) and that any combination of strategies will lead to trade-offs. For extreme events, the former only becomes apparent after the fact. For example, under-investments in pandemics only became clear when COVID-19 struck. The extant literature maps out some of the trade-offs made by adopting different sourcing strategies. In this study, we consider decisions on local vs. global sourcing (i.e., supplier location) and single vs. multiple sourcing (i.e., number of suppliers) as all other sourcing decisions follow from these.
2.3.1 Single versus multiple sourcing
Research on whether single or multiple sourcing is better for supply resilience is inconclusive, suggesting that there is no ideal strategy. Single sourcing, for instance, enables the establishment of collaborative relationships that promote mutual profit (Van Weele, 2010). However, it exposes the buying organization to greater risks of disruption if that sole supplier fails (Svensson, 2004). Incorporating redundancy, e.g., backup agreements whereby the supplier reserves a certain portion of products or capacity for the buying organization can mitigate this risk (Namdar et al., 2018). Multiple sourcing is the dominant strategy in uncertain contexts (Namdar et al., 2018) because it can facilitate responsiveness to disruptions (Mehrjerdi and Shafiee, 2020). In contrast, Wiedmer et al. (2021) find that multiple sourcing worsens the impact of a disruptive event at its onset but contributes to faster recovery of supply volumes ex post. Furthermore, it is only viable under certain conditions which may be difficult to assess ex ante, making it a complex strategy to implement. For example, suppliers must be selected based on their diversity of strategies for coping with disruptions (Kahiluoto et al., 2020). In conclusion, there are trade-offs associated with either strategy in general, and specifically in the case of extreme events.
2.3.2 Local versus global sourcing
Local sourcing ensures better responsiveness to disruptions by, for example, increasing agility and flexibility (Jüttner and Maklan, 2011; Van Hoek, 2020). Nevertheless, labor costs, local resource shortages (e.g., input materials and labor), and other restrictive conditions (e.g., regulations) have led to the rise of global sourcing which broadens supply options (Gunasekaran et al., 2015). Global supply chains, however, are susceptible to disruptions that are difficult to recover from, especially when triggered by extreme events (Gunasekaran et al., 2015). Thus, neither local nor global sourcing inherently improves a buying organization's potential supply situation in future extreme events. A compromise is to settle for geographical dispersion of suppliers. Even then, success further depends on adequate investment in supply chain visibility and flexibility to deal with heightened supply chain complexity, costs resulting from dealing with multiple geographically dispersed suppliers, and to enable product or process modification (Azevedo et al., 2013; Brandon-Jones et al., 2014; Sawik, 2021). In conclusion, there are trade-offs associated with either strategy in general and particularly in the case of extreme events.
2.4 Achieving supply resilience: An equifinality perspective
The essence of equifinality is that organizations can reach a common end state through different strategies (Gresov and Drazin, 1997; Katz and Kahn, 1978). Equifinality is based on the concept of fit (Bozarth and McDermott, 1998), aiming to help answer the question of “which strategies are best”. Our study argues that different initial conditions, i.e., sourcing strategies ahead of an extreme event (ex ante), are also essential and that buying organizations must adapt their response strategies to fit those initial conditions to secure supply. Hence, a key concern vis-à-vis supply resilience is which strategies are most aligned with what initial conditions.
Despite its recognized usefulness, research adopting the equifinality concept remains very limited in procurement and SCM (e.g., Cagliano et al., 2004; Fernández and Kekäle, 2005; Kosmol et al., 2018; Marcolin and Ross, 2005; Sousa and Voss, 2008). A few papers use it to explore internal or external strategic fit in manufacturing (Bozarth and McDermott, 1998; Cagliano et al., 2004; Fernández and Kekäle, 2005). Marcolin and Ross (2005) apply it to information systems sourcing while Kosmol et al. (2018) demonstrate how different supply quality management strategies can lead to similar quality achievements. To the best of our knowledge, there is no study linking sourcing strategy and equifinal outcomes in extreme events, nor did we identify any study in the context of public procurement. Our study seeks to provides theoretical and empirical insights to this end.
3 Methods
3.1 Research approach and setting
We adopt a case-based research approach as it is suitable for studying complex real-life phenomena in their natural setting and enables an in-depth exploration of “how and why” questions (Eisenhardt, 1989; Voss et al., 2002; Yin, 2014). A multiple case study design also lends itself well to investigations on equifinality because of its inherent assumption that individual cases have unique local details that have implications for realized outcomes (Jaspers, 2007; Mills et al., 2013).
Studying extreme or “unusual” events has been argued to lead to some of the most significant contributions to theory (Bamberger and Pratt, 2010; Craighead et al., 2020), making COVID-19 suitable for our purposes (Van Hoek, 2020; Sodhi et al., 2021). We chose to focus on the ventilator supply crisis because of the extraordinary demand-supply gap triggered by the pandemic (Netland, 2020) and the complexity of manufacturing ventilators with their hundreds of intricate parts coupled with the fact that suppliers are concentrated in a handful of countries (Elsahn and Siedlok, 2021; S080; S120). We focused on the first wave of the pandemic to ensure that our analysis covered a period of never-before-experienced impact.
3.2 Case design and selection
Our unit of analysis is the buying organization – a central government or a governmental agency with a procurement remit. Our case study design explicitly considered the theory-method link (Dubois and Araujo, 2007). Specifically, we embedded equifinality in the research design, and selected countries with different initial conditions but who all succeeded in avoiding ventilator shortages. Since we sought to make general statements about different paths to supply resilience, we selected cases of “polar types” varying significantly along important theoretical dimensions (Eisenhardt and Graebner, 2007; Miles and Huberman, 1994) of initial conditions: (1) Pre-existing redundancy measured by available ventilator capacity in healthcare systems at the onset of the pandemic; and (2) available sourcing strategies, i.e., local versus global and single versus multiple sourcing. Applying these case sampling criteria, we selected the United Kingdom (UK), Switzerland, and Germany. Furthermore, due to limited data during the first wave of COVID-19, we selected these three countries because we could access much publicly available data on their response pathways. Table 2 summarizes the cases and their initial conditions.Table 2 Case countries and initial conditions at the onset of the pandemic's first wave.
Table 2Case Country Hospital availability pre-COVID (per 100,000)
(Proxy for redundancy) Global Market Share of Local Suppliers (Proxy for local Sourcing Options)
Emergency ventilatorsa Mobile Ventilatorsb
UK 7.49
Low redundancy «2%
Very limited domestic sourcing options «2%
Very limited domestic sourcing options
Switzerland 9.88
Low redundancy 22%c
Multiple domestic sourcing options 18%c
Limited domestic sourcing options
Germany 30.5
High redundancy 19%
Multiple domestic sourcing options 45%
Multiple domestic sourcing options
a Sweden highest at 22%, Germany second highest, China third highest at 10%.
b Germany highest, Switzerland second highest, US third highest at 5%.
c Contribution of a US/Switzerland firm. So, effective capacity could be less.
(Sources: S003; S004; S011; S016; S067; IPG Research in: S080; S085-6; S100; S109-11; S124)
Germany and the UK were the most polar cases, the UK having the least favorable initial conditions and Germany the most favorable ones. UK-based suppliers were few and mostly producing basic ventilators that could not be used for critical care (S012; S090; S099). In Germany there was high redundancy in hospitals (S011) and abundant local supply options including two large manufacturers of emergency ventilators for intensive care. Switzerland had relatively low redundancy in healthcare facilities but hosts the largest ventilator supplier in the world, Hamilton, among others (S079).
3.3 Data collection
We collected data from multiple reliable secondary data sources in two main stages to enhance completeness: (1) April–August 2020; and (2) April 2021–January 2022. For each of the countries, we searched for articles addressing ventilator supply and government response strategies using key words “Ventilator”, “Supply”, “Shortage”, “COVID-19 pandemic” in English and German. The main data sources are shown in Table 3 . In total, 124 online documents were used (numbered S001 to S124). The source links are available as a supplement and PDF files are available on request. Some of the data from these sources were instrumental in contextualizing key findings, thereby improving sensemaking and mitigating the risk of over-attributing outcomes (Eisenhardt, 1989; Miles and Huberman, 1994) to the response strategies we identified. For instance, we tracked data on demand management strategies employed in each country, e.g., postponing or cancelling some medical procedures to free up ventilator capacity and imposing lockdowns and cross-border restrictions to slow down infection rates (e.g., S003; S007; S014; S024-8; S031-2; S052-3; S056; S064; S089; S099; S109-10). This helped us to explain the differing ventilator needs in the three countries and assisted with validating findings from limited data. For example, the Swiss government did not publicly publish ventilator numbers but the evidence of an aggressive demand management approach leading to lower infection rates assured us that what we could glean from the limited sources on ventilator quantities was reliable.Table 3 Sources of the data for the case study.
Table 3Case Country Type of Data Publication/Online Sources
United Kingdom (UK) News CNN, BBC News, Financial Times, The Guardian, The Washington Post, Bloomberg
Government Websites The Government Websites of the UK (gov.uk), Office for National Statistics
Statements The Statements of the Ventilator Challenge UK Consortium
Company Websites Smiths Medical Official Website, Penlon Official Website
Germany News Ärztezeitung, Tagesschau, BBC News, Die Zeit, Reuters, Der Spiegel, ZDF
Government Websites The Website of the German Federal Government (Bundesregierung),
The Website of The German Federal Parliament (Deutscher Bundestag)
Reports from Institutions DIVI-Intensivregister
Company Websites Drägerwerk AG Official Website, Löwenstein Medical Official Website
Switzerland News Tages-Anzeiger, Aargauer Zeitung, Handelszeitung, SWI swissinfo.ch
Government Websites The Website of the Swiss Federal Government (Der Bundesrat admin.ch)
Company Websites Hamilton Medical Official Website
Reports from Institutions Swiss Society of Intensive Care Medicine
To ensure credibility and internal validity (Eisenhardt and Graebner, 2007; Voss et al., 2002), we triangulated data from different sources. In the few instances where we found conflicting accounts for the same piece of information, we gave more weight to official government publications. If different government publications provided different estimates and we could not ascertain the accuracy of any of the sources, we used a range. For example, we worked with a range of 700–850 ventilators for Switzerland after multiple searches and triangulation of sources did not yield a definitive result.
3.4 Data coding and analysis
We conducted a qualitative content analysis to analyze the data (Miles and Huberman, 1994). We deductively coded the data for information on initial conditions and the response strategies by each of the three governments. We also developed inductive codes for emerging themes that helped to refine insights of strategies and decisions made. One such example is the deployment of regulatory instruments by all the governments in efforts to improve their supply situation, in addition to sourcing-related strategies. Sample deductive and inductive codes for the strategies and decisions are provided in Supplement 2. Measures to ensure trustworthiness and generalizability (Miles and Huberman, 1994; Voss et al., 2002) include careful selection of secondary sources, coding by two of the co-authors, and iterating between the data and the literature during the analysis process.
We conducted both within- and cross-case analyses (Yin, 2014). Within-case analyses helped us to develop an in-depth understanding of the unique characteristics, initial conditions and response strategies in each country. We accounted for temporal aspects (Craighead et al., 2020) by constructing a timeline of response strategies against ventilator supply/availability for each country. We sought evidence of shortages defined as an instance whereby a ventilator could not be allocated to a patient in need. The timelines were also used to assess agility by comparing how long it took the different countries to initiate specific responses. Cross-case analysis subsequently helped us to identify theoretically important patterns across the three countries in terms of response strategies and paths to supply resilience.
4 Results
4.1 United Kingdom
The UK Government employed multiple response strategies, often simultaneously, to fill the demand-supply gap and improve supply (S079; S090). At the onset of the pandemic, official estimates for ventilator needs for seriously ill COVID-19 patients were at least 30,000, and there was far less than one third available (S070; S099). Despite its dire initial prognosis, the country avoided shortages (S082; S084). Fig. 1 shows the response and supply timeline with the initial conditions captured at the beginning. We present the response strategies in greater detail next.Fig. 1 Timeline of the UK's response strategies and supply in 2020
Fig. 1
4.1.1 Reallocation of available ventilators
The first thing that the UK Government did was to work with private hospitals and the army to reallocate existing ventilators to the publicly funded NHS (National Health Service). The NHS reached agreements with private hospitals and the Ministry of Defence for the reallocation of thousands of ventilators (old and new stocks) to NHS hospitals (S007).
4.1.2 Local sourcing
The UK Government sought to secure supply locally in several ways with the bulk of the responses being initiated between mid-March and early April 2020. We present them in turn.
Developing new ventilator models - Due to global ventilator shortages, the main path of procurement from existing suppliers was going to be too slow to allow the Government to meet rapidly growing demand. For that reason, the Government set in motion initiatives to stimulate local product development and production in order to avoid disruption problems experienced by global supply chains (S041; S093). To facilitate a speedy response, the Medicines and Healthcare Products Regulatory Agency (MHRA), accelerated the approval process for new products (S090).
By mid-March, the UK Government turned to local companies to develop ventilators that could be mass-produced quickly and locally (S054; S093; S117; S123). Within one week, a Continuous Positive Airway Pressure (CPAP) device for less critically ill patients had been developed (S054; S087; S092; S096; S117; S123). This was the result of a collaboration between the University College London (UCL), University College London Hospitals (UCLH) and Mercedes Formula One (S087; S096). Formula One Teams were credited for their use of high-speed techniques to quickly generate solutions to a time-sensitive matter (S063). By the end of March, the device had been approved for use on hospitalized patients by the MHRA. There were several other innovative bids to address potential shortages through developing simple designs. For example, a team from Oxford University was developing vital related substitute devices for ventilators (S051).
The UK Government also incentivized domestic businesses to design new emergency ventilators. Actors from multiple industries participated. For example, a team of academics, engineers and doctors created a prototype of a ventilator to treat coronavirus patients (S097). The three main participants in this initiative were British companies Dyson and Meggitt as well as UK-based GTECH. The project of Dyson, a technology company known for its vacuum cleaners and hair dryers, started in partnership with the Technology Partnership, a Cambridge-based medical equipment company (S087; S092); the defense company Babcock later joined this project (S040). This partnership led to the design of the “CoVent” ventilator for treating COVID-19 patients. It could be produced at speed and at volume, and Dyson bore the full cost of development - more than £20 million (S005; S107). Meggitt, a firm specializing in producing components for the aerospace, defense and energy industries, led a consortium of aerospace and automotive companies (S095). GTECH, which specializes in cordless vacuum cleaners and garden power tools, worked on its own (S005).
Local manufacturing of existing and new models - The first set of new ventilators to be manufactured in the UK were the CPAP devices. Forty devices had been delivered to hospitals by the end of March (S054; S117; S123). Because of its simple design, the production rate of the CPAP device was as high as 1000 per day and 10,000 were subsequently procured by the government for the NHS (S051; S090; S099; S123).
To ramp up the local production of existing, modified, and new ventilator designs, the Government promoted collaboration between the few ventilator manufacturers and manufacturers from other industries (S010; S041; S092; S114). Some manufacturers were forced to close (some of) their production lines or factories (S038) due to a drastic decline in demand caused by loss of income and uncertainty for consumers (S085). Thus, there were manufacturers with idle capacity that repurposed their facilities for the production of ventilators (S005; S040-2; S092). Notably, the UK Ventilator Challenge was set up and commenced on March 14, 2020 (S093). A Ventilator Challenge UK consortium was formed consisting of technology and engineering companies from the aerospace, automotive and medical sectors (S094, S096, S113).
The consortium accelerated the production of two ventilator models: Smiths' paraPAC and Penlon's Prima ESO2 (S041; S092; S114). Smiths' paraPAC was a pre-existing lightweight mobile ventilator which can be used in ambulances or on arrival at the hospital, but not for long-term intensive care (S106). Penlon's Prima ESO2 was modified to conform to the rapid manufacturing specification and can be used on critically ill patients (S057). The Penlon Prima ESO2 devices are typically deployed in operating theatres and can be used for more acute patients (S057; S093). The typical combined manufacturing capacity of Penlon and Smiths was only between 50 and 60 ventilators a week. The Ventilator Challenge UK consortium scaled up the production of the models to more than 100 devices per day and peaked at more than 400 per day (S008; S012; S059; S099; S114).“The Ventilator Challenge helped scale up the production of three models (paraPAC, Vivo65 and Nippy4+) and helped guide one newly adapted model, the Penlon ESO 2, all the way through regulatory approval.” (S059)
The Ventilator Challenge UK program was eventually opened to companies outside the UK. The ventilator models Vivo65 and Nippy4+ from the Swedish company Breas Medical were added to the list of Ventilator Challenge devices (S055). The UK government assisted Breas in “negotiating with suppliers to source critical components and expediting shipments of key parts from around the world” leading to the delivery of a first batch of 150 ventilators in early May (S055). Subsequently, these four models, i.e., Penlon ESO 2, paraPAC, Vivo65 and Nippy4+, received continued support from the UK Government while support for other devices ended. The main reasons were that the former had been approved by the MHRA and projections showed that the suppliers would meet remaining demand (S055; S059; S082). Dyson had received an initial order of 10,000 CoVent ventilators which was subsequently cancelled. The company later announced that they were hoping to make CoVent available to the global market and were not looking to recoup any costs from the government (S009; S044; S107).
Ultimately, the UK Government created local sourcing channels by substitution through funding innovation and supply chain compression. In addition, some regulations were relaxed and approval processes fast-tracked. All participants of the ventilator challenge were absolved of any legal liabilities and compensated for the direct costs incurred.“Cabinet Office committed to covering participants' reasonable direct costs and indemnified them against legal actions from inadvertently breaching intellectual property rights, competition and procurement law, and some aspects of product failure. It estimates it will spend £113 million (excluding VAT) on design costs, components and factory capacity for ventilators it did not buy because the design was not viable or not needed to meet the government’s targets.” (S082, p. 11)
4.1.3 Global sourcing
To spread risk and complement the efforts of domestic companies to find solutions to the ventilator supply, the UK Government still placed orders for ventilators from the EU region and other nations (S041; S051). At the early stage of the outbreak, British embassies around the world were asked to help the UK Government to tackle the shortage problem of medical equipment in the NHS, including ventilators (S070).
To shorten lead time and boost supply through this international sourcing route, the UK Government temporarily lifted import duty requirements on vital medical items including ventilators (S039). This decision resulted in 8000 ventilators being brought into the UK duty-free (S039). In order to expedite the transportation process, some ventilators were transported to the UK by air instead of the normal sea freight mode (S041).
4.1.4 UK main outcomes
Shortages were avoided even if the bulk of the ventilators were supplied past the initial peak of the infections, as shown in Fig. 1 (S082; S084). By then, the demand estimate had been revised down to 18,000 (S090; S099). The UK Government ultimately spent £569 million (over US$780 million) on 20,900 new ventilators (S033; S035; S082). This figure excludes any investments or incentives towards boosting local ventilator development and production.
4.2 Germany
Germany had the most favorable initial conditions at the start of the pandemic. Starting off with high redundancy compared with other countries (Table 2), and having aggressively sought to control the spread of COVID (S029; S032; S089; S109; S110; the government no longer needed additional ventilators after the 4000 (from emergency orders for 20,000) units were delivered (S011).“Compared with other European states, Germany is by far best equipped to deal with the outbreak. Not only does it have a good number of intensive care beds — around 28,000 — it also possesses 25,000 ventilators, with 10,000 more on the way.” (S011)
The Federal Government of Germany initiated one major response to fill the potential demand-supply gap. In particular, they secured early supply/manufacturing capacity from local suppliers, at the first month of the pandemic outbreak (S001; S077). Fig. 2 shows the response and supply timeline as well as the initial conditions. We present the details of this response strategy next.Fig. 2 Timeline of Germany's response strategies and supply in 2020
Fig. 2
4.2.1 Local sourcing
In Germany, prior to COVID-19, ventilator suppliers directly transacted with hospitals and clinics. When COVID-19 was declared a pandemic, the Federal Government swiftly centralized procurement and secured supply for meeting pandemic-driven demand, which was projected to exceed existing capacity (S001; S030).“It’s unusual for a government to order medical gear directly (…). Normally customers in Germany are hospitals and clinics.” (S078)
The Federal Government sought to learn from the COVID-19 treatment experience of China and Italy where infections peaked well ahead of other countries. Based on lessons learned, it ordered ventilators suitable for three performance levels: high-end intensive care, life-support intensive care, and simple ventilators (S046). Before April 2020, the Ministry of Health had concluded contracts for more than 20,000 ventilators with several suppliers. Most of the ventilators were procured from two domestic ventilator manufacturers: Drägerwerk was to supply 10,000 (S010; S049; S050; S073; S078-9; S118)1 while Löwenstein Medical was to supply 6500. Due to supply issues and the unprecedented demand, the production process and lead times were expected to take several months (S004; S047). Löwenstein Medical was to manufacture and deliver the ventilators to the Federal Government for distribution to health facilities over a three-month period (S075). The order of 10,000 ventilators was the largest Drägerwerk had ever received and equivalent to their annual production volume (S078). To ensure delivery in full within 12 months (S062; S078; S111), Drägerwerk expanded its production capacity in Lübeck (S049).
4.2.2 Germany main outcomes
Shortages were avoided during the initial peak of the infections because Germany already had high redundancy within healthcare facilities. In addition, the country faced very low infection numbers and hospitalizations compared to most other countries (S011; S073). The bulk of the ventilators were procured at the start of the pandemic, but delivery was spread over a one-year period. In response to lower demand than initially expected, the German government reduced order quantities with several suppliers and retained some of the ventilators as emergency stock for future emergencies (S001).
4.3 Switzerland
Switzerland initiated two key responses to fill the demand-supply gap caused by the COVID-19 pandemic: reallocating ventilators to potential pressure points and used local sourcing to secure supply. Ultimately, and partially because of its highly aggressive infection control strategy (S017-19), the country successfully avoided ventilator shortages: there was overcapacity in the healthcare industry, and most hospital beds prepared for COVID-19 patients remained vacant (S069). Fig. 3 shows the Swiss Federal Government's initial conditions, its response strategy, and the supply timeline.Fig. 3 Timeline of Switzerland's response strategies and supply in 2020
Fig. 3
4.3.1 Reallocation of available resources
At the start of the pandemic, it turned out that the Swiss Intensive Care Medicine Association, SGI, had reserves of mobile ventilation devices as well as portable ventilation devices which could be immediately added to the hospital capacity; there was also sufficient capacity in the rescue services and emergency stations (S100, S102). The Koordinierter Sanitätsdienst (KSD), another Swiss medical services organization, also had additional ventilation equipment (S102). The existing ventilator capacity of these organizations and others were allocated to hospitals and clinics as frontline care providers S102).
4.3.2 Local sourcing
In mid-March 2020, the Federal Government entered an exclusive supply arrangement with a Swiss medical technology company and the largest ventilator supplier in the world, Hamilton Medical. The Government purchased 900 ventilators from Hamilton Medical in response to rapidly rising infection rates (S065; S066; S103). Ventilators of the type HAMILTON-T1 Military, which are transport-intensive-care ventilation devices suitable for use both inside and outside intensive care units, were ordered for Swiss hospitals. The Federal Government further instructed Hamilton Medical to supply locally manufactured ventilators exclusively to the Government. Therefore, during the duration of the contract, Hamilton Medical could not process other individual local or international orders. Swiss hospitals could only purchase these devices through the Federal Resources Administration. Effectively – although the government stated that this was not the case (S105) – exports were banned. Nonetheless, Hamilton was reported to have delivered 400 ventilators to Italy around mid-March (S079) and were engaging with stakeholders impacting their ability to source for components or supply different customers. For example, they had indicated that they would contest an export ban. The company's CEO, Andreas Wieland, was quoted as saying:“The Federal Council could ban us from exporting on the basis of the extraordinary measures introduced. But we would try to oppose that. If they let us do our work and support us, we will do everything we can to make enough material available for Switzerland.” (S103)
Hamilton also stocked up on components in anticipation of a sudden increase in demand when they heard about a mysterious respiratory virus from Chinese associates (S078) and prioritized customers most in need of ventilators (S103, S105).
4.3.3 Switzerland main outcomes
Switzerland avoided ventilator shortages during the first wave of the pandemic and even had excess capacity between its healthcare facilities, rescue services, and emergency stations (S069; S100). Given the relatively low infection rates recorded in the country largely because of stringent lockdown measures (S024-8; S100), the redundant capacity in multiple locations already covered a substantial amount of the demand.
4.4 Cross-case analysis
Table 4 summarizes each government's initial conditions and how they adapted their sourcing strategies in response to the ventilator supply crisis. We find that governments, as buying organizations, were responsible for setting the supply objectives and facilitating their achievement, while the suppliers put in the work necessary to close the demand-supply gap.Table 4 Initial conditions and response strategies per buying organization.
Table 4 Case Country
UK Switzerland Germany
Initial Conditions • Very low redundancy
• Limited sourcing options
• Low redundancy
• Multiple sourcing options
• High redundancy
• Multiple sourcing options
Government response strategies Ensure continuity• Reallocate redundant capacity (from private sector and army)
• Build stockpiles for future outbreaks
Secure additional supply• Global sourcing from existing ventilator suppliers
• Increase local supply options (spurring innovation)o Pool resources from disparate industries as a source of innovation
o Financial incentives⁃ Absorb/reduce suppliers’ additional material and product development costs of innovative bids to boost supply
⁃ Waive import tariffs
o Regulatory changes⁃ Relax supply-related regulations for developers and manufacturers
⁃ Expedite product approval process
⁃ Sourcing by central government
Ensure continuity• Reallocate redundant capacity (from private sector and emergency and rescue services)
Secure additional supply• Exclusive arrangements with local ventilator suppliers (bulk of order quantity placed with one supplier)
• Regulatory changeso Partially restrict exports (but not recognized as such (e.g., in S104 the Swiss government argued that, unlike Germany, they were not restricting exports))
o Restrict sale only to central government
Ensure continuity
No evidence found
Secure additional supply• Exclusive arrangements with local ventilator suppliers (bulk of order quantity placed with one supplier)
• Regulatory changeso Restrict exports
o Restrict sale only to central government
Supplier response strategies Product Innovation (to enable manufacturing at scale)• Reverse engineering existing models
• Developing new models
• Developing components for new and existing models
Ramp up supply• Scale up capacity of existing ventilator manufacturers
• Repurpose manufacturing facilities production
• Split development and/or manufacturing roles across sectors
Product Innovation
No evidence found
Ramp up supply• Scale up manufacturing
• Engage in dialogue with stakeholders impacting supply
• Prioritize resource allocation to ease pressure on supply
Product Innovation
No evidence found
Ramp up supply• Scale up manufacturing
4.4.1 Response strategies – buying organizations
Our first RQ refers to initial conditions of buying organizations and how these conditions influenced their response strategies to the same extreme event. Ultimately, all strategies were employed to bridge the demand-supply gap. There are similarities and differences among the three cases in relation to sourcing strategy as a route to supply resilience, and the enduring effects of the employed strategies.
Local versus global sourcing - Local sourcing played the most critical role in securing supply in all three cases. The UK had major challenges with sourcing globally. Curiously, the bulk of its supply came from new local market entrants. While Germany and Switzerland secured supply through a rationalized supply base, the existing suppliers significantly ramped up supply. The German supplier Drägerwerk, for example, reported that it was ramping up supply to meet Germany's demand which accounted for one year of their capacity and demand from other countries.
Single versus multiple sourcing – There was divergence in the number of suppliers the three governments used to secure supply. The UK Government dealt with three broad categories of ventilator suppliers: international, existing domestic, and new domestic, most of whom participated in the UK Ventilator Challenge. The number of suppliers was eventually rationalized to four once demand had stabilized. Switzerland opted for a single-sourcing approach. Germany placed orders with different suppliers but relied mainly on two suppliers for more than 80% of the ventilators ordered.
Diversity of strategies to secure supply - The three governments implemented different response strategies or used the same strategies differently to achieve secure supply. The UK Government employed the most strategies (see Table 4). In addition to pursuing global sourcing, the Government provided the financial incentives and regulatory space and pace needed to spur innovation. This led to increased local availability. Innovation needs in this setting entailed product modification, new product development, and repurposing of manufacturing. In turn, suppliers engaged in collaborative ventures (mostly horizontal and spanning multiple sectors) and developed product and manufacturing flexibility capabilities at record speed. This is evidenced by, for example, the fact that lead times of the new market entrants were better than those of existing suppliers (S082). Germany did more of the same: enhance redundancy. Disregarding the demand control measures of Switzerland, the same conclusion can be drawn for this case: the focus of the Swiss Government was on building further redundancy.
Regulatory instruments - Regulations were crucial for enabling all three governments to achieve their objectives. The UK employed regulatory instruments the most to increase supplier agility and flexibility. Switzerland and Germany applied fewer such instruments to preclude competition for scarce supplier resources (through import restrictions) and control local sourcing channels (through centralized procurement) and enhance coordination.
4.4.2 Response strategies - suppliers
All suppliers of the three governments, new and old, had to supply unprecedented volumes. We identified different ways in which they achieved this.
Increasing manufacturing capacity - For most established ventilator suppliers, the use of conventional strategies of increasing production capacity were the most dominant. The exception comprised UK suppliers that also modified their existing designs so that they could be manufactured at scale. Even then, they could not produce at a higher level than the new entrants. This points to a systemic scalability challenge for established suppliers.
Adapting ventilators for manufacturing at scale – Upon realizing that existing ventilator models could not be produced fast enough to close the demand-supply gap, UK-based suppliers embarked on the task of changing the ventilator products instead. With UK Government support and intense collaboration with others including universities and medical professionals, the first such ventilators (CPAP) had been reverse-engineered, approved by the MHRA, manufactured, and delivered to healthcare facilities within a matter of weeks.
Proactively managing risks – From a demand and supply perspective, our data shows that at least one of the suppliers managed risks both before and during the pandemic (we imagine that they were not unique in this respect). They foresaw the demand surge and stocked up on components which enabled them to immediately ramp up supply when the crisis hit. They also engaged with stakeholders impacting their ability to source for components or supply different customers and/or prioritized customers most in need of ventilators.
4.4.3 Different paths to resilience
Our second RQ concerned the ways in which different response strategies led to supply resilience. Although all three cases successfully closed the demand-supply gap, there were differences in the resilience capabilities linked to the adopted response strategies. Table 5 shows the main resilience capabilities linked to each government and its suppliers.Table 5 Comparison of dominant supply resilience capabilities across the three cases.
Table 5“+” relatively strong, “-” relatively weak.
“++” relatively much stronger, “--” relatively much weaker.
Combining the cross-case findings and time lines of each country, several patterns emerge. All countries appeared to achieve visibility through centralized control and coordination. Suppliers had different visibility concerns, e.g., UK suppliers focused more on upstream collaboration and Switzerland's suppliers focused more on establishing actual ventilator demand downstream. German and Swiss governments were more agile, issuing contracts very early. The UK Government was comparatively slower to initiate its response strategies (lower agility). However, they were comparatively more agile than the UK Government, having taken the initial decision to secure supply far sooner. The UK, though, implemented a wider range of strategies (higher flexibility) and worked closely with diverse stakeholders (more intense collaboration) to secure supply. However, the most significant collaborations were observed on the supply-side (e.g., consortiums with participants from multiple sectors including universities and manufacturing organizations). Since the UK Government was slower to make the critical sourcing decisions (based on their timeline and compared to the other two countries), UK-based suppliers had less time to close the demand-supply gap compared to their German- and Switzerland-based counterparts. Thus, fittingly, the UK's suppliers were highly agile, adapting/developing new ventilator models and producing them at record speed. The Ventilator Challenge UK closed within four months of being established, having successfully bridged the demand-supply gap. In comparison, Germany's biggest supplier had been given a year to produce the required quantities. It appears that the early government response eased the pressure on suppliers to dramatically ramp up supply, making agility less critical.
In sum, all three countries were successful but for different reasons. Germany had highly redundant ventilator capacity at the start of the pandemic and Switzerland ran a successful infection control campaign leading to lower demand. To maintain their advantageous positions and minimize the use of resources, they had to move fast. In contrast, because of its disadvantageous position, the UK needed to adopt more diverse strategies and invest more.
5 Discussion
Our results show different paths to supply resilience of three governments contingent on their initial conditions following supply-side disruptions and a demand surge for ventilators during the first wave of the COVID-19 pandemic. In line with previous research, local sourcing was crucial for all three governments (e.g., Jüttner and Maklan, 2011; Van Hoek, 2020). New suppliers contributed the most to rapidly ramping up supply. Although existing suppliers were eager to close the demand-supply gap, there were limits to what they could achieve in terms of ramping up supply. This may reflect the difficulty of increasing production capacity in high-tech industries (Elsahn and Siedlok, 2021); meaningful capacity improvement costs can range from hundreds of millions to billions of dollars and can take years to achieve (Trivedi, 2021). However, it raises questions about how new entrants could outperform existing suppliers both in developing new models and manufacturing at scale. We discuss these and other differences in more depth next and formulate related propositions for validation in future research.
5.1 Propositions
In the context of extreme events, generally speaking, strategies that entail investments before or after an extreme event has occurred carry inherent risks and trade-offs. A pertinent question is how to move forward when the event does occur. We now discuss the different paths to supply resilience and present the accompanying propositions.
5.1.1 Low versus high redundancy
The risk appetite of buying organizations can help explain differences in resource allocation for extreme events (Namdar et al., 2018). Higher risk aversion is associated with prioritizing reliability over costs and more diverse strategies to mitigate risks. Conversely, lower risk aversion is related to a greater focus on cost and a tendency to improve relationships and collaborate with a few preferred suppliers (Namdar et al., 2018). Research also suggests that risk propensity may be determined by environmental factors, e.g., if there are high risks but low response capacity, buying organizations tend to be risk averse (Mena et al., 2020).
Our results do not lend support to these findings in relation to strategies in place pre-COVID as well as in response to the pandemic. Specifically, the countries we studied had arm's-length relationships with ventilator suppliers pre-COVID (e.g., S012, S065, S078). They neither had collaborative relationships with suppliers nor multiple strategies in place. This may be because, pre-COVID, the ventilator market was deemed stable. However, since the ventilator supply chain typically works on a make-to-order basis, with lead times of at least a few months (S082), risk averse countries would likely have built more redundancy into their systems. Alternatively, countries with lower redundancy may have shared the same concerns but allocated their resources differently.
From this starting point, it is interesting that the buying government with the lowest redundancy (the UK) took the longest to act and secure more supply. This further supports the risk aversion argument. That said, given the ventilators are a high-tech product and that sourcing from existing suppliers was difficult, lower redundancy might imply that the options available for securing supply under conditions where demand far outstrips supply can be highly costly and risky (e.g., sourcing poor quality from non-vetted suppliers). As a result, more time would be needed to carefully weigh options. Thus, we argue that lower redundancy is positively associated with less agile decision-making at first (primarily because of increased risks and costs) while higher redundancy (primarily driven by risk aversion) is associated with more agile decision-making in relation to securing additional supply. We thus propose:Proposition 1 At the onset of an extreme event, agile execution is better aligned with high redundancy, while flexibility is better aligned with low redundancy.
A major implication of this scenario is that, since buying organizations starting off with low redundancy lose time agonizing on the best approaches or seeking alternatives, their suppliers must be much more agile to make up for lost time. Our results show such agility by the UK’s local suppliers and, not surprisingly, this comes at a significant cost to the UK Government.
An issue that reduces the options of buying organizations is the intense competition for resources. Therefore, buying organizations with low redundancy must be more creative. In case of global products, other governments can take measures that worsen the situation (Craighead et al., 2020). Indeed, some countries temporarily imposed export bans on ventilator suppliers based in their countries (Hodgson, 2020). This meant that some countries had to build the supply chain from scratch, and do so fast. Because of the complexity of ventilator products, all necessary changes entailed innovation: new product development, product modification, repurposing and scaling up of manufacturing (Elsahn and Siedlok, 2021). Buying organizations with high redundancy, however, could take less drastic measures because of the high redundancy (excess finished products on hand) that enables continuity (e.g., Jüttner and Maklan, 2011; Wallace and Choi, 2011). In the face of diminishing global sourcing options, though, they also had to boost supply capacity at the very least. Entering into early contract arrangements with suppliers would have signaled to suppliers how much they needed to expand capacity and enabled them to ascertain if the subsequent volume targets were realistic. Therefore, we further argue that buying organizations with lower redundancy can only significantly improve their supply position by spurring innovation by local suppliers, while those with higher redundancy may, ceteris paribus, need to facilitate scaling up of supply through early contracting of local suppliers. Accordingly, we propose:
Proposition 2 At the onset of an extreme event, the type of supplier incentives is related to the level of existing redundancy: spurring local supplier innovation is better aligned with low redundancy, while early contracting of suppliers is better aligned with high redundancy.
5.1.2 Limited versus multiple sourcing options
The strategic sourcing literature largely focuses on the decision about how many suppliers to use for a single product as a risk mitigation strategy. Single sourcing is argued to improve responsiveness because of the intimacy of the buyer-supplier relationship which increases supplier willingness to respond fast to buyer's changing needs (Van Weele, 2010). Multiple sourcing can improve responsiveness too, but in a different way: by providing multiple supply options (Mehrjerdi and Shafiee, 2020). In high-tech industries, it is preferable to have supplier switching capabilities and a flexible sourcing strategy, e.g., in volume, mix, and delivery (Azevedo et al., 2013). Furthermore, major capacity constraints make it sensible to split orders and/or reserve capacity with multiple suppliers (Erkoc and Wu, 2005). The literature, however, appears to be silent on what happens if some buying organizations find themselves with fewer sourcing options than others as the result of varied implications of the same extreme event.
Our findings are counter-intuitive in that the case with the least sourcing options (UK Government) combined innovation with a broad base of horizontally collaborating suppliers, new and old, originating in disparate industries. Given that multiple sourcing is inherently complex, the combination with the need for innovation is an intriguing finding. Equally interesting is that the other two countries with multiple sourcing options chose to further rationalize their supply base and sourced the bulk of their ventilators from one or two suppliers. Under high uncertainty, the tendency is to increase the supply base (Namdar et al., 2018). A possible explanation for this outcome is that the limited suppliers available to the UK neither had the technical expertise nor the capacity to meet the UK's demand. Consequently, the UK Government had the herculean task of building both the expertise and production capacity rapidly. For Germany and Switzerland, however, all the suppliers available to them had sufficient technical expertise and could ramp up supply to meet their needs if they could focus solely on their government customers, respectively. However, there was competition for the suppliers' available capacity and the two governments were faced with the choice to either impose export bans or motivate a few suppliers to focus on satisfying their demand. Ultimately, and partly because of the backlash from earlier decisions to impose export bans, rationalizing supply by picking the best performing suppliers was the chosen route to securing supply. It would be more efficient for the contracted suppliers to process one big order from the same customer than to split and manage supply among multiple customers with small orders. In turn, this would increase reliability under uncertainty. Therefore, we argue that for high-tech products like ventilators, having limited sourcing options at the onset of the extreme event require greater flexibility to innovate and to extend the supply base. On the other hand, if there are multiple sourcing options, the concern is how to ensure supplier reliability and commitment. The latter appears to be achieved by reducing the number of suppliers. Thus, we propose:Proposition 3 At the onset of an extreme event, the intensity of collaboration is related to the number of sourcing options available: intense collaboration (horizontal and vertical) is better aligned with having limited sourcing options, while limited collaboration is better aligned with having multiple sourcing options.
5.1.3 The role of regulations
A major capability that governments have as buying organizations is to wield regulatory instruments to encourage or discourage specific supplier behaviors. In the UK, local supplier flexibility and agility were needed. Flexible regulatory processes appear to facilitate agility as, for example, suppliers do not have to worry about breaking competition laws or being liable for unforeseen product failure further down the road. However, stringent approval processes remained in place for safety-related aspects of new or modified ventilator designs. Flexibility in this regard would otherwise set a dangerous precedent and could cost lives (Elsahn and Siedlok, 2021). Agility in implementing regulatory processes like product approval, however, could enhance supplier flexibility. Suppliers would be willing to expend more effort and resources if they know that their efforts have a good chance of paying off. It also gives the assurances needed for suppliers to bear risks on behalf of the buying organization.
The opposite effects are observed in countries with multiple local sourcing options where the main concern is to guarantee supply from already capable, but highly sought after, suppliers. Thus, regulations were tightened more (lower regulatory flexibility) and this negatively impacted suppliers' agility as they could not take any decisive action while obliged by law not to sell to other parties (S104). The speed with which these regulations were imposed also reduced supplier flexibility. For instance, pursuing measures to dramatically ramp up supply would be futile if those restrictions were not lifted for a long time. Suppliers in this sector are generally less keen to build stocks (S082) and would, therefore, likely not increase flexibility (significantly) under such conditions. This is partly supported by the finding that delays in imposing export bans in Switzerland enhanced the key supplier's flexibility (Hamilton) who declared that they would improve supply to serve customers in Switzerland and other countries. This suggests that not imposing bans would be better for flexibility and, by extension, availability. All governments also quickly moved to take direct charge of procurement and became the go-between for supplier and healthcare providers. This enabled buying governments to have oversight over demand and supply while also allowing them to respond rapidly to emerging issues. In sum, we propose:Proposition 4 At the onset of the extreme events, regardless of their sourcing options, buying organizations can improve:a. supplier flexibility and agility through higher regulatory flexibility.
b. their own visibility and agility through centralized control of procurement
5.2 Research implications and contributions
Our study introduces equifinality in supply resilience research. We make four key contributions to the literature. Firstly, we extend research at the intersection between strategic sourcing and resilience (e.g., Mandal, 2020; Namdar et al., 2018; Pereira et al., 2014) by employing an equifinality perspective to demonstrate that there are different pathways to supply resilience and to develop theoretical insights accordingly (see propositions). More generally, we respond to calls for the use of equifinality in procurement and supply research (e.g. Cagliano et al., 2004; Fernández and Kekäle, 2005; Kosmol et al., 2018) and demonstrate the analytical usefulness of this approach in relation to supply risk management and supply resilience.
Secondly, we add to prior research addressing contingency factors (e.g., Bode et al., 2011; Namdar et al., 2018; Roscoe et al., 2020) in two ways. Whereas the equifinality literature considers fit between enduring strategies and enduring environmental characteristics, our research considers fit between initial conditions following the onset of an adverse event (i.e., environmental disturbance) and subsequent response strategies. Thus, we extend the equifinality concept to include fit during periods of disturbance and subsequent change. Furthermore, we unveil specific conditions under which certain strategies are effective. In doing so, we reconcile conflicting empirical findings in the literature, e.g., with respect to the effectiveness of multiple sourcing strategies as opposed to using one or two suppliers (Wiedmer et al., 2021). Our focus on the buying organizations' diverse response pathways shows that the most suitable strategies depend on the buying organization's initial conditions. Hence, single and multiple sourcing, as well as local and global sourcing, can help avoid shortages but through different pathways.
Thirdly, to the best of our knowledge this is the first study applying an equifinality perspective in the context of public procurement. We add to research stressing the imperative role of public procurement in responding to crisis situations (Harland et al., 2021a; Fearne et al., 2021) by showing that public buying organizations such as government departments can make different resource allocation decisions in preparedness for extreme events and still be able to achieve desired outcomes – in this case, avoid ventilator shortages – by employing different strategies at an event's onset.
Fourthly, our study generates empirical insights with respect to how and why some sourcing strategies foster innovation and help create new market entrants in a very short space of time, thereby responding to the call of this Special Issue (Kähkönen et al., 2020) to advance knowledge concerning the implications of the COVID-19 pandemic for capacity building in the supply market, and for supplier-enabled innovation. Our results show the merits of pursuing adaptive and transformative routes to supply resilience (Feizabadi et al., 2021; Nikookar et al., 2021), as these paths hold promise for rapidly closing the demand-supply gap in unforeseen future global crises.
5.3 Implications for practice
Our results and propositions point to a need for practitioners to rethink supply resilience to extreme events. Preparing for the unknowable might be a fool's errand and may deplete resources that will be needed when the unthinkable manifests. Whatever the future entails, there are three clear messages for practitioners.
Firstly, Proposition 1, Proposition 2, Proposition 3 collectively suggest that buying organizations facing unfavorable initial conditions (e.g. in terms of low redundancy and limited sourcing options) need to facilitate unlikely, yet intense, collaborations; take decisive action to increase agility; continuously evaluate decisions and options; and be open to explore new solutions. In other words, survival and success will increasingly depend on rapid innovation from unlikely places. Our study shows that public buying organizations have a pivotal role to play in fostering cross-sectoral collaboration, building the innovative capacity of available suppliers, and helping to on-board new suppliers. Proposition 4 highlights the relevance of procurement centralization for increasing the buying organization's visibility. In addition, public organizations should be prepared to accept regulatory flexibilities (e.g. regarding competition) as these can increase suppliers' ability to respond to emergency situations in an agile fashion.
Secondly, those buying organizations that are fortunate enough to have favorable initial conditions should work with others to address the problems caused by the extreme event, or at least do no harm. For instance, the export bans not only worsened ventilator shortages but also may have robbed existing suppliers of the opportunity to adapt and transform themselves and become more resilient to similar future events. Buying organizations must also consider the long-term implications of addressing immediate concerns through short-term measures. For critical supplies, if a buying organization undermines other organizations’ ability to secure supply, this can also stifle innovation which could enable equitable access to scarce resources. Stated differently, especially for global crises, supply resilience of one organization should not be achieved at the detriment of others.
Thirdly, given the challenges facing practitioners with respect to scarce resources, the essence of allocating these resources strategically is to determine those aspects of procurement and SCM that are worth expending resources on now, while having a good understanding of the limits and risks they pose for responding to future extreme events. For governments, the key question relates to the critical public goods that should be prioritized as part of emergency preparedness and the associated immediate and opportunity costs. For example, in addition to health, food security and infrastructure are other key areas of concern given the expected impact of climate change.
5.4 Limitations and future research
Our research approach was suitable for the purpose at hand, i.e., establishing how buying organizations with different initial conditions can achieve supply resilience to the same extreme event. Future research could test our propositions, using primary data in other contexts. For example, because our study focuses on public procurement, we unearth insights into the implications of trade-offs made ex ante in allocating limited resources to secure the supply of life-saving public goods ex post following the onset of extreme events. However, public procurement is distinct from commercial procurement in that organizations can specialize in their product offering and target specific segments of the population. Furthermore, the power dynamics are different. Governments can compel supplier behavior through regulatory instruments, while private companies mostly rely on the power dynamics of the relationship. Given this key difference, it is worthwhile investigating how equifinality of outcomes is achieved in extreme events that impact the private sector.
Our study raised multiple other important questions. Firstly, under which conditions are different strategies cost effective? For example, there are conflicting findings in the literature regarding cost efficiency of different sourcing strategies, such as single and multiple sourcing (Van Weele, 2010). Another is the cost trade-off between preparedness through redundancy and responsiveness through flexibility and agility. The former involves amortizing the cost of preparedness over a long time period while the latter entails expending vast resources in a short space of time. For extreme events, picking the best strategy is difficult because it is not possible to predict if/when the resources are needed.
Secondly, our case countries are all well-resourced and have large procurement budgets. Future research can include resource-poor settings where the capacity to increase supply is diminished (Craighead et al., 2020). These would be interesting cases for developing a deeper understanding of supply resilience and, potentially, its limits. A comparative study on both settings could further reveal the impact of the behavior of well-resourced buying organizations (e.g., hoarding, accepting unjustified price hikes, and imposing export bans) on system outcomes (e.g., equitable distribution).
Finally, there is also a need for longitudinal and multidisciplinary research to better understand the wider and long-term implications of varied response strategies for supply resilience to global crises. In addition to its relevance for the aforementioned research directions, longitudinal research can help to uncover the causes and effects of different approaches over time. This is crucial given the complexity of global supply chains and the enduring effects of extreme events on their functioning.
Funding
This work was supported by the 10.13039/501100005416 Research Council of Norway – GLOBVAC, grant number 312715.
Author statement
All authors listed have contributed substantially to the project.
Nonhlanhla Dube: Supervision, Conceptualization, Formal Analysis, Methodology, Visualization, Validation, Investigation, Writing (Original draft, Review and Editing).
Qiujun Li: Methodology, Investigation, Formal Analysis, Data curation, Visualization, Writing (Original draft).
Kostas Selviaridis: Methodology, Writing (Original draft, Review and Editing), Validation, Supervision.
Marianne Jahre: Writing (Original draft, Reviewing and Editing), Funding acquisition.
Declaration of competing interest
The research is partially funded by the Research Council of Norway – GLOBVAC (grant number 312715). Nonetheless, to the best of our knowledge, no conflict of interest, financial or otherwise, exists.
Dr. Nonhlanhla Dube is an Assistant Professor of Operations Management at Lancaster University in the UK. Her areas of research interest are operations and supply chain management in the humanitarian and public sectors. Her research explores resilience in humanitarian and medicine supply chains as well as the implications of insecurity for logistics and operations strategy in the humanitarian sector. At present, she is also part of a large research group investigating cost-effective measures for dealing with shortages in medicine supply chains. Her maiden work is published in the Journal of Operations Management and Risk, Hazards & Crisis in Public Policy.
Qiujun Li completed a double degree in Logistics and Supply Chain Management (ISM Frankfurt) and Marketing Analytics (Lancaster University) in 2021. Prior to this, she did a Bachelor in Biomedical Engineering. She is interested in combining theories and analytical techniques from her studies to understand and solve problems in medicine supply chains.
Dr. Kostas Selviaridis is an Associate Professor of Operations Management at the Department of Management Science, Lancaster University Management School. His research concerns the governance of inter-organizational relationships in the supply chain, with a focus on contracting. His current work places particular emphasis on public contracting and its role in promoting innovation, resilience and sustainability goals in supply chains. Through such research Kostas has cultivated an interest, more broadly, in the interplay between supply chain management and public policy. Kostas’ research has attracted external funding from multiple sources including the British Academy for Humanities and Social Sciences and the Research Council of Norway. His work has appeared in international outlets such as the Journal of Supply Chain Management and the International Journal of Operations and Production Management. Kostas serves as an Associate Editor in the Journal of Purchasing and Supply Management.
Prof. Marianne Jahre is Professor of Logistics at Lund University and BI Norwegian Business School. She has co-edited and co-authored several books and published articles among others in JOM, IJOPM, and JHLSCM. Jahre has been working with disaster relief logistics research and teaching since 2007 in cooperation with IFRC, UNHCR, UNFPA, UNICEF, Norwegian Red Cross, and the Norwegian Refugee Council. She undertook projects on health supply chains in Uganda for UNICEF. She now heads research projects on drug shortage, including COVID-19 in cooperation with the Norwegian Institute of Public Health (NIPH) and 5 other universities.
Appendix A Supplementary data
The following are the Supplementary data to this article:Multimedia component 1
Multimedia component 1
Multimedia component 2
Multimedia component 2
1 This is the single biggest order mentioned in most sources; S045 mentions 16,000 to an unnamed supplier but we could not find other sources supporting this.
Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.pursup.2022.100773.
==== Refs
References
Ali A. Mahfouz A. Arisha A. Analysing supply chain resilience: integrating the constructs in a concept mapping framework via a systematic literature review Supply Chain Manag. 22 1 2017 16 39
Arani W. Mukulu E. Waiganjo E. Wambua J. Strategic sourcing an antecedent of supply chain resilience in manufacturing firms in Kenya Int. J. Acad. Res. Bus. Soc. Sci. 6 10 2016 1 18
Azevedo S.G. Govindan K. Carvalho H. Cruz-Machado V. Ecosilient Index to assess the greenness and resilience of the upstream automotive supply chain J. Clean. Prod. 56 2013 131 146
Bamberger P. Pratt M. Moving forward by looking back: reclaiming unconventional research contexts and samples in organizational scholarships Acad. Manag. J. 53 2010 665 671
Blackhurst J. Dunn K.S. Craighead C.W. An empirically derived framework of global supply resiliency J. Bus. Logist. 32 4 2011 374 391
Bode C. Petersen K.J. Ellram L.M. Understanding responses to supply chain disruptions: insights from information processing and resource dependence perspectives Acad. Manag. J. 54 4 2011 833 856
Bozarth C. McDermott C. Configurations in manufacturing strategy: a review and directions for future research J. Oper. Manag. 16 4 1998 427 439
Brandon-Jones E. Squire B. Autry C.W. Petersen K.J. A contingent resource-based perspective of supply chain resilience and robustness J. Supply Chain Manag. 50 3 2014 55 73
Cagliano R. Caniato F. Spina G. Lean, agile and traditional supply: how do they impact manufacturing performance? J. Purch. Supply Manag. 10 4/5 2004 151 164
Christopher M. Peck H. Building the resilient supply chain Int. J. Logist. Manag. 15 2 2004 1 13
Christopher M. Rutherford C. Creating supply chain resilience through agile six sigma Critical Eye, June/Aug 2004 24 28
Craighead C.W. Ketchen D.J. Darby J.L. Pandemics and supply chain management research: toward a theoretical toolbox Decis. Sci. J. 51 4 2020 838 866
Dubois A. Araujo L. Case research in purchasing and supply management: opportunities and challenges J. Purch. Supply Manag. 13 3 2007 170 181
Eisenhardt K.M. Building theories from case study research Acad. Manag. Rev. 14 1989 532 550
Eisenhardt K.M. Graebner M.E. Theory building from cases: opportunities and challenges Acad. Manag. J. 50 2007 25 32
Elsahn Z. Siedlok F. Can we build it? Yes, we can!” complexities of resource re-deployment to fight pandemic Ind. Market. Manag. 93 2021 191 207
Erkoc M. Wu D. Managing high-tech capacity expansion via reservation contracts Prod. Oper. Manag. 14 2 2005 232 251
Fearne A. Wagner B. McDougall N. Loseby D. The power of purpose – lessons in agility from the Ventilator Challenge Supply Chain Manag.: Int. J. 26 6 2021 753 766
Feizabadi J. Gligor D. Choi T. Examining the resiliency of intertwined supply networks: a jury-rigging perspective Int. J. Prod. Res. 2021 10.1080/00207543.2021.1977865
Fernández I. Kekäle T. The influence of modularity and clock speed on reverse logistics strategy: implications for the purchasing function J. Purch. Supply Manag. 11 2005 193 205
Gresov C. Drazin R. Equifinality: functional equivalence in organization design Acad. Manag. Rev. 22 2 1997 403 428
Gunasekaran A. Subramanian N. Rahman S. Supply chain resilience: role of complexities and strategies Int. J. Prod. Res. 53 22 2015 6809 6819
Harland C. Essig M. Lynch J. Patrucco A. Policy-led public procurement: does strategic procurement deliver? J. Public Procure. 21 3 2021 221 228
Harland C. Knight L. Patrucco A. Lynch J. Telgen J. Peters E. Tatrai T. Ferk P. Practitioners' learning about healthcare supply chain management in the COVID-19 pandemic: a public procurement perspective Int. J. Oper. Prod. Manag. 41 13 2021 178 189
Hodgson C. Export Bans Blocked Signed Contracts to Buy PPE, MPs Told 2020 Financial Times https://www.ft.com/content/8c0a29fc-a523-4901-a190-fe5a2dcc8faa Accessed
Hohenstein N. Feinsel E. Hartman E. Giunipero L. Research on the phenomenon of supply chain resilience Int. J. Phys. Distrib. Logist. Manag. 45 2015 90 117
Jaspers F. Case study research: some other applications besides theory building J. Purch. Supply Manag. 13 3 2007 210 212
Jüttner U. Maklan S. Supply chain resilience in the global financial crisis: an empirical study Supply Chain Manag.: Int. J. 16 4 2011 246 259
Kahiluoto H. Mäkinen H. Kaseva J. Supplying resilience through assessing diversity of responses to disruption Int. J. Oper. Prod. Manag. 40 3 2020 271 292
Kähkönen A.-K. Patrucco A. Appolloni A. van Raaij E. Purchasing and supply management learning from the pandemic: transforming for better crisis management. Special issue call for papers J. Purch. Supply Manag. 2020
Katz D. Kahn R.L. The Social Psychology of Organizations second ed. 1978 Wiley New York
Knight L. Tate W. Carnovale S. Di Mauro C. Bals L. Caniato F. Gualandris J. Johnsen T. Matopoulos A. Meehan J. Miemczyk J. Patrucco A. Schoenherr T. Selviaridis K. Touboulic A. Wagner S. Future business and the role of purchasing and supply management: opportunities for “business-not-as-usual” PSM research J. Purch. Supply Manag. 28 1 2022 1 12
Kochan C.G. Nowicki D.R. Supply chain resilience: a systematic literature review and typological framework Int. J. Phys. Distrib. Logist. Manag. 48 8 2018 842 865
Kosmol T. Reimann F. Kaufmann L. Co-alignment of supplier quality management practices and cognitive maps – a neo-configurational perspective J. Purch. Supply Manag. 24 1 2018 1 20
Lee H.L. The triple-A supply chain Harv. Bus. Rev. 82 10 2004 102 113
Mandal S. Impact of supplier innovativeness, top management support and strategic sourcing on supply chain resilience Int. J. Prod. Perform. Manag. 2020 Print
Marcolin B.L. Ross A. Complexities in IS sourcing: equifinality and relationship management ACM SIGMIS - Data Base: the DATABASE for Advances in Information Systems 36 4 2005 29 46
Mehrjerdi Y.Z. Shafiee M. Multiple-sourcing in sustainable closed-loop supply chain network design: tire industry case study Int. J. Supply Oper. Manag. 7 3 2020 202 221
Mena C. Melnyk S.A. Baghersad M. Zobel C.W. Sourcing decisions under conditions of risk and resilience: a behavioral study Decis. Sci. J. 51 4 2020 985 1014
Mentzer J.T. Stank T.P. Esper T.L. Supply chain management and its relationship to logistics, marketing, production, and operations management J. Bus. Logist. 29 1 2008 31 46
Miles M.B. Huberman A.M. Qualitative Data Analysis second ed. 1994 Sage Thousand Oaks, California
Mills A. Durepos G. Wiebe E. Equifinality Encyclopedia of Case Study Research 2013
Namdar J. Li X. Sawhney R. Pradhan N. Supply chain resilience for single and multiple sourcing in the presence of disruption risks Int. J. Prod. Res. 56 6 2018 2339 2360
Netland T. Solving the Ventilator Shortage 2020 https://ethz.ch/en/news-and-events/eth-news/news/2020/04/blog-netland-ventilator.html Accessed
Nikookar E. Varsei M. Wieland A. Gaining from disorder: making the case for antifragility in purchasing and supply chain management J. Purch. Supply Manag. 27 3 2021 in press)
Pereira C. Christopher M. Lago Da Silva A. Achieving supply chain resilience: the role of procurement Supply Chain Manag. 19 5/6 2014 626 642
Pettit T.J. Fiksel J. Croxton K.L. Ensuring supply chain resilience: development of a conceptual framework J. Bus. Logist. 31 1 2010 1 21
Pettit T.J. Croxton K.L. Fiksel J. Ensuring supply chain resilience: development and implementation of an assessment tool J. Bus. Logist. 34 1 2013 46 76
Pettit T.J. Croxton K.L. Fiksel J. The evolution of resilience in supply chain management: a retrospective on ensuring supply chain resilience J. Bus. Logist. 40 1 2019 56 65
Ponomarov S.Y. Holcomb M.C. Understanding the concept of supply chain resilience Int. J. Logist. Manag. 20 1 2009 124 143
Rice J.B. Caniato F. Building a secure and resilient supply network Supply Chain Manag. Rev. 7 5 2003 22
Roscoe S. Skipworth H. Aktas E. Habib F. Managing supply chain uncertainty arising from geopolitical disruptions: evidence from the pharmaceutical industry and Brexit Int. J. Oper. Prod. Manag. 40 9 2020 1499 1529
Sawik T. On the risk-averse selection of resilient multi-tier supply portfolio Omega 101 2021 102267
Scholten K. Schilder S. The role of collaboration in supply chain resilience Supply Chain Manag.: Int. J. 20 4 2015 471 484
Scholten K. Stevenson M. van Donk D.P. Dealing with the unpredictable: supply chain resilience Int. J. Oper. Prod. Manag. 40 1 2019 1 10
Selviaridis K. Effects of public procurement of R&D on the innovation process: evidence from the UK small business research initiative J. Public Procure. 21 3 2021 229 259
Selviaridis K. Spring M. Fostering SME supplier-enabled innovation in the supply chain: the role of innovation policy J. Supply Chain Manag. 58 1 2022 92 123
Sheffi Y. Rice B.J. A supply chain view of the resilient enterprise MIT Sloan Manag. Rev. 47 1 2005 41 48
Sodhi M. Tang C. Willenson E. Research opportunities in preparing supply chains of essential goods for future pandemics Int. J. Prod. Res. 2021 in press)
Sousa R. Voss C.A. Contingency research in operations management practices J. Oper. Manag. 26 6 2008 697 713
Svensson G. Supplier segmentation in the automotive industry: a dyadic approach of a managerial model Int. J. Phys. Distrib. Logist. Manag. 34 1 2004 12 38
Trivedi A. Toyota Broke its Just-In-Time Rule Just in Time for the Chip Shortage 2021 https://www.bloomberg.com/opinion/articles/2021-02-15/toyota-broke-its-just-in-time-rule-just-in-time-for-the-chip-shortage Accessed
Tukamuhabwa B. Stevenson M. Busby J. Supply chain resilience in a developing country context: a case study on the interconnectedness of threats, strategies and outcomes Supply Chain Manag.: Int. J. 22 6 2017 486 505
Tukamuhabwa Stevenson M. Busby J. Zorzini M. Supply chain resilience: definition, review and theoretical foundations for further study Int. J. Prod. Res. 53 18 2015 5592 5623
Van Hoek R. Research opportunities for a more resilient post-COVID-19 supply chain – closing the gap between research findings and industry practice Int. J. Oper. Prod. Manag. 40 4 2020 341 355
Van Hoek R. Loseby D. Beyond COVID-19 supply chain heroism, no dust settling yet – lessons learned at Rolls Royce about advancing risk management thinking Int. J. Oper. Prod. Manag. 41 10 2021 1579 1592
Van Weele A. Purchasing & Supply Chain Management: Analysis, Strategy, Planning, and Practice fifth ed. 2010 Cengage Learning Andover
Vanpoucke E. Ellis S.C. Building supply-side resilience – a behavioural view Int. J. Oper. Prod. Manag. 40 1 2019 11 33
Voss C. Tsikriktsis N. Frohlich M. Case research in operations management Int. J. Oper. Prod. Manag. 22 2 2002 195 219
Wallace S.W. Choi T.M. Flexibility, information structure, options, and market power in robust supply chains Int. J. Prod. Econ. 134 2 2011 284 288
Walker B.H. Resilience: what it is and is not Ecol. Soc. 25 2 2020 11
Wiedmer R. Rogers Z. Polyviou M. Mena C. Chae S. The dark and bright sides of complexity: a dual perspective on supply network resilience J. Bus. Logist. 42 3 2021 336 359
Wieland A. Durach C.F. Two perspectives on supply chain resilience J. Bus. Logist. 2021 in press)
Wieland A. Wallenburg M.C. The influence of relational competencies on supply chain resilience: a relational view Int. J. Phys. Distrib. Logist. Manag. 43 4 2013 300 320
Xie J. Wei L. Zhang W. Xia Y. Li J. A responsive pricing grid operator sourcing from competing generators under uncertain supply and demand Sustainability 11 15 2019 4061
Yin R.K. Case Study Research: Design and Methods fifth ed. 2014 SAGE Publications Thousand Oaks, CA
Zeng A.Z. Xia Y. Building a mutually beneficial partnership to ensure backup supply Omega 52 2015 77 91
| 0 | PMC9705261 | NO-CC CODE | 2022-12-01 23:19:34 | no | 2022 Dec 12; 28(5):100773 | utf-8 | null | null | null | oa_other |
==== Front
Sci Total Environ
Sci Total Environ
The Science of the Total Environment
0048-9697
1879-1026
Elsevier B.V.
S0048-9697(22)07676-8
10.1016/j.scitotenv.2022.160573
160573
Article
Sewershed surveillance as a tool for smart management of a pandemic in threshold countries. Case study: Tracking SARS-CoV-2 during COVID-19 pandemic in a major urban metropolis in northwestern Argentina
Cruz Mercedes Cecilia a1
Sanguino-Jorquera Diego a
González Mónica Aparicio a
Irazusta Verónica Patricia ab
Poma Hugo Ramiro a
Cristóbal Héctor Antonio ab
Rajal Verónica Beatriz acd⁎
a Instituto de Investigaciones para la Industria Química (INIQUI), Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET), Universidad Nacional de Salta (UNSa), Av. Bolivia 5150, 4400 Salta, Argentina
b Facultad de Ciencias Naturales, UNSa, Salta, Argentina
c Facultad de Ingeniería, UNSa, Salta, Argentina
d Singapore Centre for Environmental Life Sciences Engineering (SCELSE), Nanyang Technological University, Singapore, Singapore
⁎ Corresponding author at: INIQUI, CONICET, UNSa, Av. Bolivia 5150, 4400 Salta, Argentina.
1 Present address: Department of Biological Sciences, Marquette University, 1428 W Clybourn Street, Milwaukee, WI 53233, USA.
29 11 2022
29 11 2022
16057331 8 2022
24 11 2022
25 11 2022
© 2022 Elsevier B.V. All rights reserved.
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Wastewater-based epidemiology is an economical and effective tool for monitoring the COVID-19 pandemic. In this study we proposed sampling campaigns that addressed spatial-temporal trends within a metropolitan area. This is a local study of detection and quantification of SARS-CoV-2 in wastewater during the onset, rise, and decline of COVID-19 cases in Salta city (Argentina) over the course of a twenty-one-week period (13 Aug to 30 Dec) in 2020. Wastewater samples were gathered from 13 sewer manholes specific to each sewershed catchment, prior to convergence or mixing with other sewer lines, resulting in samples specific to individual catchments with defined areas. The 13 sewershed catchments selected comprise 118,832 connections to the network throughout the city, representing 84.7 % (534,747 individuals) of the total population. The number of COVID19-related exposure and symptoms cases in each area were registered using an application developed for smartphones by the provincial government. Geographical coordinates provided by the devices were recorded, and consequently, it was possible to geolocalise all app-cases and track them down to which of the 13 sampling catchments belonged. RNA fragments of SARS-CoV-2 were detected in every site since the beginning of the monitoring, anticipating viral circulation in the population. Over the course of the 21-week study, the concentrations of SARS-CoV-2 ranged between 1.77 × 104 and 4.35 × 107 genome copies/L. There was a correspondence with the highest viral load in wastewater and the peak number of cases reported by the app for each catchment. The associations were evaluated with correlation analysis. The viral loads of SARS-CoV-2 in wastewater were a feasible means to describe the trends of COVID-19 infections. Surveillance at sewershed scale, provided reliable and strategic information that could be used by local health stakeholders to manage the COVID-19 pandemic.
Graphical abstract
Unlabelled Image
Keywords
SARS-CoV-2
Wastewater
Monitoring
COVID-19
Wastewater-based epidemiology
SALTA COVID app
Editor: Warish Ahmed
==== Body
pmcData availability
Data will be made available on request.
| 36460114 | PMC9705263 | NO-CC CODE | 2022-12-10 23:15:27 | no | Sci Total Environ. 2023 Mar 1; 862:160573 | utf-8 | Sci Total Environ | 2,022 | 10.1016/j.scitotenv.2022.160573 | oa_other |
==== Front
Infect Prev Pract
Infect Prev Pract
Infection Prevention in Practice
2590-0889
Published by Elsevier Ltd on behalf of The Healthcare Infection Society.
S2590-0889(22)00062-2
10.1016/j.infpip.2022.100261
100261
Article
Infection of healthcare workers despite a high vaccination rate during the fifth wave of COVID-19 due to Omicron variant in Hong Kong
Wong Shuk-Ching a
Chan Veronica Wing-Man a
Yuen Lithia Lai-Ha a
AuYeung Christine Ho-Yan a
Leung Jessica Oi-Yan a
Li Chi-Kuen a
Kwok Monica Oi-Tung a
So Simon Yung-Chun b
Chen Jonathan Hon-Kwan b
Chiu Kelvin Hei-Yeung b
Tam Anthony Raymond cb
Hung Ivan Fan-Ngai c
Kai-Wang To Kelvin d
Lo Janice Yee-Chi e
Yuen Kwok-Yung d
Cheng Vincent Chi-Chung ab∗
a Infection Control Team, Queen Mary Hospital, Hong Kong West Cluster, Hong Kong Special Administrative Region, China
b Department of Microbiology, Queen Mary Hospital, Hong Kong Special Administrative Region, China
c Department of Medicine, Li Ka Shing Faculty of Medicine, The University of Hong Kong, Hong Kong Special Administrative Region, China
d Department of Microbiology, Li Ka Shing Faculty of Medicine, The University of Hong Kong, Hong Kong Special Administrative Region, China
e Centre for Health Protection, Department of Health, Hong Kong Special Administrative Region, China
∗ Corresponding author. Address: Infection Control Team, Queen Mary Hospital, Hong Kong West Cluster, Hong Kong Special Administrative Region, China. Tel.: +852 22552351, Fax: +852 23523698.
29 11 2022
29 11 2022
10026127 9 2022
22 11 2022
© 2022 The Authors
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Background
No nosocomial infection was recorded in our healthcare workers (HCWs) during the early phase of the coronavirus disease 2019 (COVID-19) pandemic. With the emergence of the Omicron variant of increased transmissibility, infection in HCWs occurred as expected. We aimed to study the epidemiology of infection in HCWs and to describe the infection control measures during the outbreak of the Omicron variant.
Methods
With daily rapid antigen testing and molecular confirmation test for COVID-19, infected HCWs were interviewed by infection control nurses (ICNs) to investigate the potential source of infection. The epidemiology of COVID-19 in Hong Kong served as reference.
Results
During the fifth wave of COVID-19 (31 December 2021 to 31 May 2022), 1,200,068 cases were reported (incidence 95 times higher than in preceding waves in Hong Kong; 162,103 vs 1,707 per million population respectively, P<0.001). The proportion of infected HCWs was significantly higher than that of the general population (24.9%, 1,607/6,452 vs 16.2%, 12, 000, 068/7,403,100 respectively; P<0.01). The proportion of infected non-clinical staff was significantly higher than that of clinical staff (31.8%, 536/1,687 vs 22.5%, 1,071/4,765 respectively; P<0.001). Of 82.8% (1,330/1,607) infected HCWs interviewed by ICNs, 99.5% (1,324/1,330) had been fully vaccinated; 49.5% (659/1,330) had no identifiable source; 40.7% (541/1,330) were probably infected from household members; 9.8% (130/1,330) had possible exposure to confirmed patients or HCWs, but no lapse in infection control measures or inappropriate use of personal protective equipment was recalled.
Conclusion
Omicron variant is highly transmissible such that breakthrough infection occurred despite high level of vaccination.
==== Body
pmcIntroduction
Coronavirus disease 2019 (COVID-19), caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), was described as the first Disease X on the World Health Organization list of priority diseases requiring urgent research and development attention [1,2]. As of 18 September 2022, over 609 million confirmed cases and over 6.5 million deaths have been reported globally [3]. Infection and death among healthcare workers (HCWs) were not unusual, especially during the early phase of the pandemic [4]. In a systemic review of COVID-19 infection among HCWs, 152,888 infections and 1,413 deaths were reported globally as of 8 May 2020. Deaths among HCWs constituted 0.52% of total population of COVID-19 deaths [5].
In Hong Kong, infection control and public health response was immediately activated following the official announcement of the community-acquired pneumonia outbreak in Wuhan, Hubei Province by the National Health Commission of the People's Republic of China on 31 December 2019 [6,7]. Protection of HCWs from nosocomial acquisition of SARS-CoV-2 was considered paramount because 8 HCWs died during the outbreak of severe acute respiratory syndrome (SARS) in 2003. With stringent infection control measures, there was no nosocomial infection with SARS-CoV-2 among HCWs up till the pre-Omicron phase [[8], [9], [10]].
The emergence of Omicron sublineages (BA.1, BA.2, to BA.4 and BA.5) of increasing transmissibility [11] has resulted in explosive outbreaks in the community [[12], [13], [14], [15]]. Models simulating the household setting underscored the contribution of airborne transmission of the Omicron variant, especially during asymptomatic or pre-symptomatic infection, as compared with the ancestral strain [16]. HCWs could also be infected with SARS-CoV-2 in the community, especially during the Omicron wave. Here, we reported the epidemiology of COVID-19 infection among our HCWs and described our infection control measures during the outbreak of the SARS-CoV-2 Omicron variant.
Material and methods
Epidemiology of the fifth wave of COVID-19 Hong Kong
From the outset, the daily number of laboratory-confirmed COVID-19 cases was reported in the public domain of the Centre for Health Protection (CHP), Department of Health, the Government of Hong Kong Special Administrative Region, China [17]. The evolution of the COVID-19 epidemic from the first to the fourth wave in Hong Kong has been summarized previously [14]. The epidemic curve and public health responses were described.
Control of COVID-19 and burden of COVID-19 patients in the hospital
Infection control measures have evolved to minimize the risk of nosocomial outbreaks during the COVID-19 pandemic [[8], [9], [10]]. Briefly, the measures included staff training, directly observed donning and doffing of personal protective equipment (PPE), enforcement of hand hygiene and environmental disinfection, as well as proactive screening and isolation of COVID-19 patients in airborne infection isolation rooms (AIIRs). In the fifth wave, some of the general wards were temporarily converted into negative pressure wards (NPWs) for COVID-19 patients in view of the high occupancy of the AIIRs in Queen Mary Hospital (QMH), a 1,700-bed university-affiliated hospital. COVID-19 patients with clinical improvement were transferred to four extended-care hospitals in our healthcare network. For hospitalized patients, universal admission screening by real-time reverse transcription polymerase chain reaction (RT-PCR) using deep throat saliva (DTS) or nasopharyngeal swab (NPS) specimens was performed. While awaiting RT-PCR results upon admission, patients were managed in designated cubicles, namely surveillance cubicles, in general wards with 6 air changes per hour. Portable air purifiers were placed inside the surveillance cubicles to improve air dilution. Patients with positive RT-PCR results for SARS-CoV-2 were immediately transferred to AIIRs or NPWs which were designated for COVID-19 patients. Patients with negative SARS-CoV-2 results were transferred to other cubicles in general wards. In general wards, repeated testing was undertaken if the patient had clinical features suggestive of COVID-19. The extent of contact tracing for potential secondary cases would depend on risk assessment by the infection control team. For the purpose of the current study, patients with positive SARS-CoV-2 RNA detection only 3 or more days after hospitalization were defined as nosocomial cases, so as to include as many nosocomial cases as possible for analysis. For nosocomial COVID-19 cases, sporadic cases referred to isolated cases, whereas clusters referred to ≥ 2 cases. The number of COVID-19 patient admissions and bed occupancy in our network during the fifth wave were recorded.
COVID-19 testing among HCWs
HCWs were encouraged to refrain from work if they had fever or respiratory symptoms. Self-collected DTS specimens were sent to the hospital microbiology laboratory for SARS-CoV-2 RNA detection by RT-PCR if HCWs had symptoms or any epidemiological exposure to COVID-19. Since 21 February 2022, all HCWs were required to perform a rapid antigen test (GLINE-2019-nCoV Ag, BGI, China) by self-collected nasal swabs before work. If the rapid antigen test result was positive, HCWs were required to perform another rapid antigen test to ensure a consistent positive result. HCWs with immediately repeated positive rapid antigen test results were required to refrain from work and to proceed with a confirmatory test by RT-PCR using DTS. Infected HCWs would be interviewed by an ICN to assess the potential source of infection. Possible exposure to COVID-19 was defined as HCWs having contacted a COVID-19 case in either hospital or community setting, regardless of appropriateness of PPE, in the past 5 days. Nosocomial COVID-19 infection among HCWs was defined as staff who had inappropriate PPE when caring for a COVID-19 patient in the past 5 days. Appropriate PPE included the use of a surgical respirator, cap, face shield, gown, and gloves.
COVID-19 vaccination among HCWs
Two formulations of COVID-19 vaccines were available for HCWs since March 2021: CoronaVac, inactivated whole cell vaccine, Sinovac Biotech (Hong Kong) Limited, and BNT162b2 mRNA vaccine, BioNTech, Fosun Pharma in collaboration with the German drug manufacturer. HCWs with direct patient care were required to complete the second dose of COVID-19 vaccination by 16 February 2022. All uninfected HCWs were required to complete the second dose by 1 April 2022. The deadline of receiving the third dose was 16 May 2022, or within 150 calendar days from the second dose, whichever later. HCWs who recovered from COVID-19 were required to complete 2 doses of COVID-19 vaccination unless they had already completed the second or third dose of vaccination before infection. Exemption of vaccination was provided to staff with medical contraindication.
Viral load assessment of respiratory specimens
For clinical specimens, total nucleic acid extraction was performed using 250μL of the specimen by the eMAG extraction system (bioMérieux, Marcy-l'Etoile, France) following the manufacturer's instructions. Quantification of SARS-CoV-2 RNA was performed by RT-PCR as previously described [18].
This study was approved by the Institutional Review Board of The University of Hong Kong/Hospital Authority Hong Kong West Hospital Cluster.
Statistical analysis
The Chi-Square test was used to compare independent categorical variables between groups. All reported P values were two-sided. A P value of <0.05 was considered statistically significant. Computation was performed using the SPSS Version 15.0 for Windows.
Results
Epidemiology of the fifth wave of COVID-19 in Hong Kong
Soon after the importation of the SARS-CoV-2 Omicron variant [19], the fifth wave of COVID-19 in Hong Kong commenced on 31 December 2021, which was the second anniversary of the official announcement of community-acquired pneumonia of unknown etiology in Wuhan, Hubei Province [6]. During the fifth wave (defined as 31 December 2021 to 31 May 2022 in the current study), a total of 1,200,068 COVID-19 cases were reported to CHP. The incidence of COVID-19 during the fifth wave was 95 times higher than the total from the first to fourth wave (n=12,636) in Hong Kong (162,103 vs 1,707 per million population respectively, P<0.001). The daily number of new cases exceeded 10,000 on 25 February 2022 and reached a peak of >50,000 on 3 March 2022, amounting to 1,352 and >7,676 per million population respectively. As of 31 May 2022, there were 9,165 deaths within the fifth wave (1,238 per million population), of which 6,507 (71.0%) and 1,534 (16.7%) were aged ≥ 80 and 70–79 years respectively. Among the fatal cases, 6,923 out of 9,165 (72.0%) had not received any COVID-19 vaccine, while the vaccination coverage of the first, second, and third dose in the Hong Kong population was 6,700,406 (90.5%), 6,332,041 (85.5%), and 3,876,310 (52.4%) respectively [20]. The epidemic curve and public health responses by managing cases in community isolation facilities in the Asia World-Expo (∼1,000 beds), public rental buildings (∼3,000 residential flats), fangcang shelters (∼20,000 beds), and Kai-Tak cruise terminal (∼1,000 beds) are shown in Figure 1 .Figure 1 Epidemic curve of COVID-19 during the fifth wave in Hong Kong (31 December 2021 to 31 May 2022). Note. CTF in AWE, community treatment facility of around 1,000 beds in Asia World-Expo which was re-opened on January 2, 2022; FC, fangcang shelter for the purpose of community isolation facility. FC1 denotes fangcang shelter located in Tsing Yi which was opened on March 1, 2022; FC2 denotes fangcang shelter located in San Tin which was opened on March 9, 2022; FC3 denotes fangcang shelter located in Hong Kong-Zhuhai-Macao Bridge which was opened on March 12, 2022; FC4 denotes fangcang shelter located in Fanling which was opened on March 13, 2022; FC5 denotes fangcang shelter located in Hung Shui Kiu which was opened on March 17, 2022; FC6 denotes fangcang shelter located in Tam Mi which was opened on March 24, 2022. All six fangcang shelters were purposely built as community isolation facilities to provide a total of 20,000 beds; KTCT denotes Kai Tak cruise terminal which was converted into community isolation facility and opened on March 21, 2022 to provide around 1,000 beds; PRB denotes public rental buildings, Queens Hill Estate, located in Fanling, and Heng King House of Lai King Estate, located in Kwai Chung, which were converted into community isolation facility and opened on February 24, 2022 to provide around 3,000 residential flats.
Figure 1
Control of COVID-19 and burden of COVID-19 in the hospital
With increasing COVID-19 patient admissions during the fifth wave, up to 50% of beds in the general wards were temporarily converted into NPWs caring for COVID-19 patients. Mobile Modular High Efficiency Particulate Arrestance Filter Units (MMHUs) and exhaust fans were installed in each cubicle to increase air-changes per hour from 6 to 10. Negative pressure was established with airflow from the corridor to the cubicles. Of 2,320 COVID-19 patients admitted to QMH, 1,284 were male and 1,036 female, with a median age of 73 years (range: 8 days to 107 years). The daily numbers of COVID-19 patient admissions and bed occupancy during the fifth wave are shown in Figure 2 .Figure 2 Number of COVID-19 cases in Queen Mary Hospital and Hong Kong West Cluster during the fifth wave of COVID-19. Note. HKWC, Hong Kong West Cluster; QMH, Queen Mary Hospital. Hong Kong West Cluster is a healthcare network comprising Queen Mary Hospital, a 1,700-bed university-affiliated tertiary referral center, and another 5 extended-care hospitals with a total of 1,700 beds. During the fifth wave of COVID-19, four out of 5 extended-care hospitals received clinically stable COVID-19 from Queen Mary Hospital.
Figure 2
During the study period, 37 nosocomial COVID-19 patients were diagnosed at a median of 12 days (range, 3–158 days) after hospitalization in 16 general wards without conversion to NPWs. There were 10 sporadic cases in 10 different general wards. The remaining 27 patients comprised 6 clusters in 6 different general wards, including 12 cases in 1 cluster. ICNs coordinated infection control measures when a nosocomial COVID-19 patient was diagnosed. Nosocomial COVID-19 patients were immediately transferred to AIIRs to reduce the risk of further transmission in the general wards. Terminal disinfection of the general ward was performed. Hand hygiene among HCWs and patients were enforced based on our previous experience on outbreak prevention and control [[21], [22], [23], [24], [25]]. All exposed patients and HCWs in the general ward were tested for SARS-CoV-2 by RT-PCR daily for 7 days for early recognition of secondary cases.
COVID-19 infection among HCWs
During the fifth wave, a total of 1,607 (24.9%) of 6,452 HCWs in QMH were tested positive for COVID-19 till 31 May 2022 (Figure 3 ). The proportion of infected HCWs was significantly higher than that of the general population in Hong Kong (24.9%, 1,607/6,452 vs 16.2%, 1,200,068/7,403,100 respectively; P<0.001). The proportion of infected non-clinical staff was significantly higher than that of clinical staff (31.8%, 536/1,687 vs 22.5%, 1,071/4,765 respectively; P<0.001). Among 1,607 infected HCWs, 1,330 (82.8%) were interviewed by an ICN by phone. None had history of past infection of COVID-19. Symptomatic infection was reported in 1,050 (78.9%) HCWs. Symptoms included sore throat (53.8%, 565), cough (27.1%, 285), fever (19.9%, 209), runny nose (12.8%, 134) and headache (7.0%, 74). Among the 1,330 interviewees, 659 (49.5%) did not report any exposure to cases in the household or hospital. Another 541 (40.7%) reported contact with infected household members. The remaining 130 (9.8%) recalled possible exposure in the past 5 days to patients or staff within or outside the hospital, who were subsequently diagnosed to have COVID-19. Among this group of 130 HCWs, 113 were clinical staff, with 22 (19.5%) having performed aerosol generating procedures (AGPs) for COVID-19 patients with full PPE (Figure 4 ). Six (0.45%) out of 1,330 infected HCWs had not received any COVID-19 vaccine. Only 1% of the staff was exempted from COVID-19 vaccination in our healthcare network as of 31 May 2022.Figure 3 Number of healthcare workers in Queen Mary Hospital infected with COVID-19. Note. The peak of HCW infection occurred earlier than the peak of COVID-19 cases in the community and the hospital. It was the result of effort of ICNs to remind HCWs to avoid social gathering and obey social distancing when there was increasing HCW infection.
Figure 3
Figure 4 Analysis of COVID-19 infected healthcare workers in Queen Mary Hospital (31 December 2021 to 31 May 2022). Non-clinical staff includes administrative, clerical, and non-care related supporting staff. AGP, aerosol generating procedures; PPE, personal protective equipment.
Figure 4
Discussion
During the fifth wave of COVID-19 predominantly caused by Omicron subvariant BA.2 in Hong Kong, an unprecedented number of COVID-19 infections was reported among our HCWs even though infection control practices in the healthcare setting remained stringently enforced throughout the COVID-19 pandemic. Our finding of HCWs infection was in contrast with the observation in the early phase of the COVID-19 pandemic in Hong Kong that nosocomial infection among HCWs was not recorded in hospitals [9,10], temporary test centers [26], and community isolation and treatment facilities [27]. Nosocomial infection of nine HCWs was only reported in a COVID-19 outbreak in late 2020 [18]. In this study, about one-fourth of HCWs were infected in our hospital and the proportion was comparable to a private hospital of 600-bed in Hong Kong [28]. It is important to explore if the infection control measures remained appropriate to protect our HCWs.
Before the onset of the fifth wave of COVID-19, we adopted a hospital-based approach that all suspected or confirmed COVID-19 patients were isolated in healthcare facilities, including AIIRs in the hospitals or community treatment facilities. This approach of institutionalization for containment could have minimized community transmission of COVID-19. In the healthcare facilities, HCWs were provided with full PPE including a surgical respirator, cap, face shield, gown, and gloves during patient care. Directly observed donning and doffing was enforced to maximize staff protection and minimize the risk of self-contamination upon removal of PPE. With the evolution of SARS-CoV-2, RNA of the virus has been increasingly detected in air samples collected in AIIRs with 12 air-changes per hour [29,30]. However, the quantitation of viral RNA was as low as 0.005 genome copies per litre of air [30]. The provision of full PPE should be sufficient to protect HCWs from inhalation of SARS-CoV-2.
With the emergence of SARS-CoV-2 Omicron subvariant BA.2 in the fifth wave, there were explosive outbreaks in the community, especially during mask-off activities in restaurants [31] as well as vertical transmission in high-rise residential buildings [13,14]. Although the operation of community isolation facilities in the Asia World-Expo, public rental buildings, fangcang shelters, and Kai-Tak cruise terminal had provided some 25,000 additional beds to manage patients with mild or asymptomatic COVID-19 infection, this was still insufficient to cope with the approach of institutionalization for containment during the peak of the fifth wave, with more than 10,000 confirmed cases per day. HCWs thus could also have acquired infection in the community. In fact, about 50% of HCWs had no identifiable source of infection in the household or the hospital setting based on our epidemiological analysis, suggesting widespread transmission of COVID-19 in the community.
The proportion of infected HCWs was significantly higher than that of our general population. It may be related to the policy of daily COVID-19 testing among the HCWs that the number of tests per HCW was significantly higher than that of the general population during the study period (unpublished data) according to the statistics on testing for COVID-19 in Hong Kong [32]. In fact, daily COVID-19 testing detected an additional 19% of asymptomatic cases among our HCWs in this study. In a systemic analysis of COVID-19 infection, frontline HCWs were more likely to report a positive COVID-19 test when compared with community individuals in the United Kingdom and the United States [33]. In our study, it is interesting to observe that the proportion of infected non-clinical staff was significantly higher than that of clinical staff. The alertness of clinical staff may be higher as a result of on-going infection control training. Hand hygiene among our clinical staff remained highly compliant during the COVID-19 pandemic [34].
Nosocomial acquisition of SARS-CoV-2 remains difficult to ascertain because our HCWs, especially clinical staff, were not under closed-loop management in the hospital as in mainland China [35]. Clinical staff who had performed AGP for COVID-19 patients reported having donned full PPE during patient care. None of our clinical staff fulfilled the case definition of nosocomial COVID-19 infection. However, about 10% of HCWs, including clinical and non-clinical staff, recalled possible exposure to patients in the general wards or infected HCWs. The possibility of COVID-19 transmission from patient to HCWs or HCWs to HCWs could not be excluded. In fact, 80% of air samples collected in NPWs caring for COVID-19 patients were positive for SARS-CoV-2 RNA [36]. Based on the results, air dispersal of SARS-CoV-2 in general wards with unrecognized COVID-19 patients would be possible. This may also explain the presence of 37 hospitalized patients with nosocomial acquisition of SARS-CoV-2 in general wards during our study period.
Universal masking has been adopted in our community and healthcare settings even before the announcement of the COVID-19 pandemic [37,38]. Universal masking was also recommended by the Centers for Disease Control and Prevention of the United States [39], and maximizing the fitness of the mask was further emphasized [40]. Experimental studies demonstrated that surgical masks were effective at preventing virus spread under conditions of low virus load, whereas more advanced masks were required in potentially virus-rich indoor environments including medical centers and hospitals [41]. Recent studies showed that the universal use of surgical respirators as an additional infection preventive measure contributed to the rapid control of Omicron transmission in the hospital [42]. Use of the surgical respirator by an infected HCW may be one of the reasons for the lack of Omicron transmission in a cohort of immunosuppressed patients [43]. These preliminary findings may provide indirect epidemiological evidence of airborne transmission of the Omicron variant in the clinical areas. The use of surgical masks or surgical respirators to minimize the risk of Omicron transmission in the healthcare setting deserves further investigation.
There are several limitations in this study. Firstly, we did not perform case-control analysis to investigate the risk factors for COVID-19 infection among HCWs, because 50% of them could not identify the source of infection during overwhelming transmission of COVID-19 in the community. Secondly, we were not able to perform detailed epidemiological analysis for HCWs who recalled possible exposure to infected patients or staff within or outside the hospital. The route of transmission thus could not be definitively ascertained. Thirdly, we did not perform whole genome sequencing analysis to establish the transmission relationships, as the predominant virus circulating in both the community and hospitals was already known to be subvariant BA.2 during the period [13,36]. Fourthly, we defined nosocomial COVID-19 cases as patients with positive SARS-CoV-2 RNA detection 3 or more days after hospitalization. As the mean incubation period of COVID-19 was 3.42 days for the Omicron variant [44], misclassification of a small proportion of nosocomial cases might have occurred. Finally, we did not analyze the relationship between the regimen of COVID-19 vaccination and the risk of infection. The impact of COVID-19 vaccination on HCW infections during circulation of the SARS-CoV-2 Omicron variant has been reported previously. Using the two-dose BNT162b2 regimen as reference, two-dose CoronaVac recipients had a significantly higher risk of infection, whereas three-dose BNT162b2 and two-dose CoronaVac plus BNT162b2 booster regimens were associated with a lower risk of infection [28].
Conclusions
SARS-CoV-2 Omicron variant was highly transmissible such that breakthrough infection might occur despite a high level of vaccination coverage in the population. Although a significant proportion of HCWs were infected, none had severe infection, underpinned by a high vaccination rate. Nosocomial transmission of COVID-19 from patients to HCWs could not be ascertained by epidemiological analysis in view of the component of airborne transmission.
Funding statement
This study was partially supported by the Health and Medical Research Fund (10.13039/501100005847 HMRF ) Commissioned Research on Control of Infectious Disease (Phase IV), CID-HKU1-16, Health Bureau, Hong Kong SAR Government.
Conflicts of interest statement
All authors report no conflicts of interest relevant to this article.
Credit Author Statement
Shuk-Ching Wong: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Writing - original draft, Writing - review and editing. Veronica Wing-Man Chan: Investigation. Lithia Lai-Ha Yuen: Investigation. Christine Ho-Yan AuYeung: Investigation. Jessica Oi-Yan Leung: Investigation. Chi-Kuen Li: Investigation. Monica Oi-Tung Kwok: Investigation. Simon Yung-Chun So: Data curation, Formal analysis. Jonathan Hon-Kwan Chen: Resources. Kelvin Hei-Yeung Chiu: Resources. Anthony Raymond Tam: Resources. Ivan Fan-Ngai Hung: Resources. Kelvin Kai-Wang To: Resources. Janice Yee-Chi Lo: Writing - review and editing. Kwok-Yung Yuen: Funding acquisition, Supervision. Vincent Chi-Chung Cheng: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Supervision, Writing - original draft, Writing - review and editing. All authors have read and agreed to the published version of the manuscript.
Acknowledgments
We are grateful to the contribution of our frontline staff and laboratory staff in enforcing the infection control measures and performing the laboratory work in the Queen Mary Hospital.
==== Refs
References
1 Jiang S. Shi Z.L. The First Disease X is Caused by a Highly Transmissible Acute Respiratory Syndrome Coronavirus Virol Sin 35 2020 263 265 10.1007/s12250-020-00206-5 32060789
2 Bloom D.E. Cadarette D. Infectious Disease Threats in the Twenty-First Century: Strengthening the Global Response Front Immunol 10 2019 549 10.3389/fimmu.2019.00549 30984169
3 Covid-19 Weekly Epidemiological Update. Edition 109 published 14 September 2022. World Health Organization. https://www.who.int/publications/m/item/weekly-epidemiological-update-on-covid-19-21-september-2022 (Accessed 24 September 2022).
4 COVID-19 has infected some 570,000 health workers and killed 2,500 in the Americas. Pan American health organization. World Health Organization. Available at: https://www.paho.org/en/news/2-9-2020-covid-19-has-infected-some-570000-health-workers-and-killed-2500-americas-paho (Accessed 19 September 2022).
5 Bandyopadhyay S. Baticulon R.E. Kadhum M. Alser M. Ojuka D.K. Badereddin Y. Infection and mortality of healthcare workers worldwide from COVID-19: a systematic review BMJ Glob Health 5 2020 10.1136/bmjgh-2020–3097 e003097
6 The Centre for Health Protection closely monitors cluster of pneumonia cases on Mainland. Press release of the Department of Health, Hong Kong Special Administrative Region. https://www.info.gov.hk/gia/general/201912/31/P2019123100667.htm. (Accessed 19 September 2022).
7 Cheng V.C.C. Wong S.C. To K.K.W. Ho P.L. Yuen K.Y. Preparedness and proactive infection control measures against the emerging novel coronavirus in China J Hosp Infect 104 2020 254 255 10.1016/j.jhin.2020.01.010 31962139
8 Cheng V.C.C. Wong S.C. Chen J.H.K. Yip C.C.Y. Chuang V.W.M. Tsang O.T.Y. Escalating infection control response to the rapidly evolving epidemiology of the coronavirus disease 2019 (COVID-19) due to SARS-CoV-2 in Hong Kong Infect Control Hosp Epidemiol 41 2020 493 498 10.1017/ice.2020.58 32131908
9 Cheng V.C.C. Wong S.C. Chuang V.W.M. So S.Y.C. Chen J.H.K. Sridhar S. Absence of nosocomial transmission of coronavirus disease 2019 (COVID-19) due to SARS-CoV-2 in the prepandemic phase in Hong Kong Am J Infect Control 48 2020 890 896 10.1016/j.ajic.2020.05.018 32461068
10 Cheng V.C. Wong S.C. Tong D.W. Chuang V.W. Chen J.H. Lee L.L. Multipronged infection control strategy to achieve zero nosocomial coronavirus disease 2019 (COVID-19) cases among Hong Kong healthcare workers in the first 300 days of the pandemic Infect Control Hosp Epidemiol 43 2022 334 343 10.1017/ice.2021.119 33736729
11 Shrestha L.B. Foster C. Rawlinson W. Tedla N. Bull R.A. Evolution of the SARS-CoV-2 omicron variants BA.1 to BA.5: Implications for immune escape and transmission Rev Med Virol 32 2022 10.1002/rmv.2381 e2381
12 Kim E.Y. Choe Y.J. Park H. Jeong H. Chung J.H. Yu J. Community Transmission of SARS-CoV-2 Omicron Variant, South Korea Emerg Infect Dis 2022 28 2021 898 900 10.3201/eid2804.220006
13 Cheng V.C.C. Ip J.D. Chu A.W.H. Tam A.R. Chan W.M. Abdullah S.M.U. Rapid Spread of Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) Omicron Subvariant BA.2 in a Single-Source Community Outbreak Clin Infect Dis 75 2022 e44 e49 10.1093/cid/ciac203 35271728
14 Cheng V.C. Wong S.C. Au A.K. Zhang C. Chen J.H. So S.Y. Explosive outbreak of SARS-CoV-2 Omicron variant is associated with vertical transmission in high-rise residential buildings in Hong Kong Build Environ 109323 2022 221 10.1016/j.buildenv.2022.109323
15 Chen Z. Deng X. Fang L. Sun K. Wu Y. Che T. Epidemiological characteristics and transmission dynamics of the outbreak caused by the SARS-CoV-2 Omicron variant in Shanghai, China: A descriptive study Lancet Reg Health West Pac 100592 2022 29 10.1016/j.lanwpc.2022.100592
16 Ji S. Xiao S. Wang H. Lei H. Increasing contributions of airborne route in SARS-CoV-2 omicron variant transmission compared with the ancestral strain Build Environ 109328 2022 221 10.1016/j.buildenv.2022.109328
17 Covid-19 Thematic Website. Centre for Health Protection. Department of Health. The Government of the Hong Kong Special Administrative Region. https://www.coronavirus.gov.hk/eng/index.html (Accessed 24 September 2022).
18 Cheng V.C. Fung K.S. Siu G.K. Wong S.C. Cheng L.S. Wong M.S. Nosocomial Outbreak of Coronavirus Disease 2019 by Possible Airborne Transmission Leading to a Superspreading Event Clin Infect Dis 73 2021 e1356 e1364 10.1093/cid/ciab313 33851214
19 Wong S.C. Au A.K. Chen H. Yuen L.L. Li X. Lung D.C. Transmission of Omicron (B.1.1.529) - SARS-CoV-2 Variant of Concern in a designated quarantine hotel for travelers: a challenge of elimination strategy of COVID-19 Lancet Reg Health West Pac 100360 2022 18 10.1016/j.lanwpc.2021.100360
20 Statistics on 5th Wave of COVID-19. Centre for Health Protection. Department of Health. The Government of the Hong Kong Special Administrative Region. https://www.coronavirus.gov.hk/eng/5th-wave-statistics.html# (Accessed 24 September 2022).
21 Cheng V.C. Tai J.W. Li W.S. Chau P.H. So S.Y. Wong L.M. Implementation of directly observed patient hand hygiene for hospitalized patients by hand hygiene ambassadors in Hong Kong Am J Infect Control 44 2016 621 624 10.1016/j.ajic.2015.11.024 26777285
22 Cheng V.C. Tai J.W. Chau P.H. Lai C.K. Chuang V.W. So S.Y. Successful control of emerging vancomycin-resistant enterococci by territory-wide implementation of directly observed hand hygiene in patients in Hong Kong Am J Infect Control 44 2016 1168 1171 10.1016/j.ajic.2016.03.050 27291818
23 Cheng V.C.C. Wong S.C. Wong I.W.Y. Chau P.H. So S.Y.C. Wong S.C.Y. The challenge of patient empowerment in hand hygiene promotion in health care facilities in Hong Kong Am J Infect Control 45 2017 562 565 10.1016/j.ajic.2016.12.007 28131422
24 Cheng V.C.C. Wong S.C. Wong S.C.Y. Yuen K.Y. Directly observed hand hygiene - from healthcare workers to patients J Hosp Infect 101 2019 380 382 10.1016/j.jhin.2018.11.016 30496764
25 Wong S.C. Chau P.H. So S.Y. Lam G.K. Chan V.W. Yuen L.L. Control of Healthcare-Associated Carbapenem-Resistant Acinetobacter baumannii by Enhancement of Infection Control Measures Antibiotics (Basel) 11 2022 1076 10.3390/antibiotics11081076 36009945
26 Wong S.C. Leung M. Lee L.L. Chung K.L. Cheng V.C. Infection control challenge in setting up a temporary test centre at Hong Kong International Airport for rapid diagnosis of COVID-19 due to SARS-CoV-2 J Hosp Infect 105 2020 571 573 10.1016/j.jhin.2020.05.006 32437823
27 Wong S.C. Leung M. Tong D.W. Lee L.L. Leung W.L. Chan F.W. Infection control challenges in setting up community isolation and treatment facilities for patients with coronavirus disease 2019 (COVID-19): Implementation of directly observed environmental disinfection Infect Control Hosp Epidemiol 42 2021 1037 1045 10.1017/ice.2020.1355 33280617
28 Zee S.T. Kwok L.F. Kee K.M. Fung L.H. Luk W.P. Chan T.L. Impact of COVID-19 Vaccination on Healthcare Worker Infection Rate and Outcome during SARS-CoV-2 Omicron Variant Outbreak in Hong Kong Vaccines (Basel) 1322 2022 10 10.3390/vaccines10081322
29 Cheng V.C. Wong S.C. Chan V.W. So S.Y. Chen J.H. Yip C.C. Air and environmental sampling for SARS-CoV-2 around hospitalized patients with coronavirus disease 2019 (COVID-19) Infect Control Hosp Epidemiol 41 2020 1258 1265 10.1017/ice.2020.282 32507114
30 Wong S.C. Yuen L.L. Chan V.W. Chen J.H. To K.K. Yuen K.Y. Airborne transmission of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2): What is the implication of hospital infection control? Infect Control Hosp Epidemiol 43 2022 1522 1523 10.1017/ice.2021.318 34250877
31 Cheng V.C. Lung D.C. Wong S.C. Au A.K. Wang Q. Chen H. Outbreak investigation of airborne transmission of Omicron (B.1.1.529) - SARS-CoV-2 variant of concern in a restaurant: Implication for enhancement of indoor air dilution J Hazard Mater 128504 2022 430 10.1016/j.jhazmat.2022.128504
32 Statistics on testing for COVID-19 in Hong Kong. https://www.chp.gov.hk/files/pdf/statistics_on_covid_19_testing.pdf (Accessed 10 November 2022).
33 Nguyen L.H. Drew D.A. Graham M.S. Joshi A.D. Guo C.G. Ma W. Coronavirus Pandemic Epidemiology Consortium. Risk of COVID-19 among front-line health-care workers and the general community: a prospective cohort study Lancet Public Health 5 2020 e475 e483 10.1016/S2468-2667(20)30164-X 32745512
34 Wong S.C. AuYeung C.H. Lam G.K. Leung E.Y. Chan V.W. Yuen K.Y. Is it possible to achieve 100 percent hand hygiene compliance during the coronavirus disease 2019 (COVID-19) pandemic? J Hosp Infect 105 2020 779 781 10.1016/j.jhin.2020.05.016 32422309
35 Bai D.S. Geng P. Wang Z.D. Wang X.L. Xu G.R. Ye Q. Practice and experience of regional medical center entrance linkage and closed-loop management under the wartime situation of the COVID-19 in China Ann Transl Med 112 2022 10 10.21037/atm-22-61
36 Wong S.C. Chan V.W. Yuen L.L. AuYeung C.H. Leung J.O. Li C.K. Air dispersal of SARS-CoV-2 in general wards: implication on hospital infection control during the fifth wave of COVID-19 due to Omicron variant in Hong Kong Nfect Control Hosp Epidemiol 2022 Oct 24 1 4 10.1017/ice.2022.258 Epub ahead of print
37 Cheng V.C. Wong S.C. Chuang V.W. So S.Y. Chen J.H. Sridhar S. The role of community-wide wearing of face mask for control of coronavirus disease 2019 (COVID-19) epidemic due to SARS-CoV-2 J Infect 81 2020 107 114 10.1016/j.jinf.2020.04.024 32335167
38 Wong S.C. Lam G.K. AuYeung C.H. Chan V.W. Wong N.L. So S.Y. Absence of nosocomial influenza and respiratory syncytial virus infection in the coronavirus disease 2019 (COVID-19) era: Implication of universal masking in hospitals Infect Control Hosp Epidemiol 42 2021 218 221 10.1017/ice.2020.425 32799965
39 Honein M.A. Christie A. Rose D.A. Brooks J.T. Meaney-Delman D. Cohn A. CDC COVID-19 Response Team Summary of Guidance for Public Health Strategies to Address High Levels of Community Transmission of SARS-CoV-2 and Related Deaths MMWR Morb Mortal Wkly Rep 2020 69 December 2020 1860 1867 10.15585/mmwr.mm6949e2
40 Brooks J.T. Beezhold D.H. Noti J.D. Coyle J.P. Derk R.C. Blachere F.M. Maximizing Fit for Cloth and Medical Procedure Masks to Improve Performance and Reduce SARS-CoV-2 Transmission and Exposure MMWR Morb Mortal Wkly Rep 2021 70 2021 254 257 10.15585/mmwr.mm7007e1
41 Cheng Y. Ma N. Witt C. Rapp S. Wild P.S. Andreae M.O. Face masks effectively limit the probability of SARS-CoV-2 transmission Science 372 2021 1439 1443 10.1126/science.abg6296 34016743
42 Baker M.A. Rhee C. Tucker R. Badwaik A. Coughlin C. Holtzman M.A. Rapid Control of Hospital-Based Severe Acute Respiratory Syndrome Coronavirus 2 Omicron Clusters Through Daily Testing and Universal Use of N95 Respirators Clin Infect Dis 75 2022 e296 e299 10.1093/cid/ciac113 35137035
43 Karan A. Ferguson J. Salinas J.L. Lack of severe acute respiratory coronavirus virus 2 (SARS-CoV-2) transmission from a healthcare worker to a cohort of immunosuppressed patients during the SARS-CoV-2 omicron variant surge, California Infect Control Hosp Epidemiol 2022 Jul 7 2022 1 2 10.1017/ice.2022.175 Epub ahead of print
44 Wu Y. Kang L. Guo Z. Liu J. Liu M. Liang W. Incubation Period of COVID-19 Caused by Unique SARS-CoV-2 Strains: A Systematic Review and Meta-analysis JAMA Netw Open 5 2022 e2228008. 10.1001/jamanetworkopen.2022.28008
| 36465098 | PMC9705264 | NO-CC CODE | 2022-12-09 23:15:08 | no | Infect Prev Pract. 2023 Mar 29; 5(1):100261 | utf-8 | Infect Prev Pract | 2,022 | 10.1016/j.infpip.2022.100261 | oa_other |
==== Front
J Hosp Infect
J Hosp Infect
The Journal of Hospital Infection
0195-6701
1532-2939
The Healthcare Infection Society. Published by Elsevier Ltd.
S0195-6701(22)00366-8
10.1016/j.jhin.2022.11.012
Article
Risk of transmission of COVID-19 from healthcare workers returning to work after a 5-day isolation, and kinetics of shedding of viable SARS-CoV-2 variant B.1.1.529 (Omicron)
Jung Jiwon 12a
Kang Sung Woon 1a
Lee Sojeong 2a
Park Heedo 3a
Kim Ji Yeun 1a
Kim Sun-Kyung 2
Park So Yeon 2
Lim Young-Ju 2
Kim Eun Ok 2
Lim So Yun 1
Chang Euijin 1
Bae Seongman 1
Kim Min Jae 1
Chong Yong Pil 1
Lee Sang-Oh 1
Choi Sang-Ho 1
Kim Yang Soo 1
Park Man-Seong 3∗∗
Kim Sung-Han 12∗
1 Department of Infectious Diseases, Asan Medical Center, University of Ulsan College of Medicine, Seoul, South Korea
2 Office for Infection Control, Asan Medical Center, Seoul, South Korea
3 Department of Microbiology, Institute for Viral Diseases, Vaccine Innovation Center, College of Medicine, Korea University, Seoul, South Korea
∗ Corresponding author. Department of Infectious Diseases, Asan Medical Center, University of Ulsan College of Medicine, 88 Olympic-ro, 43-gil, Songpa-gu, Seoul, South Korea. Tel.: +82 2 3010-3305
∗∗ Corresponding author. Department of Microbiology and Institute for Viral Diseases, Korea University College of Medicine, 73 Goryeodae-ro, Seongbuk-gu, Seoul, 02841, South Korea. Tel.: +82 2 2286-1312
a These authors contributed equally to the work.
29 11 2022
29 11 2022
8 9 2022
16 11 2022
19 11 2022
© 2022 The Healthcare Infection Society. Published by Elsevier Ltd. All rights reserved.
2022
The Healthcare Infection Society
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Objectives
There have been limited data on the risk of onward transmission from individuals with Omicron variant infections who return to work after a 5-day isolation. We evaluated the risk of transmission from healthcare workers (HCWs) with Omicron variant who returned to work after a 5-day isolation and the viable virus shedding kinetics.
Methods
This investigation was performed in a tertiary care hospital, Seoul, South Korea. In a secondary transmission study, we retrospectively reviewed the data of HCWs confirmed as COVID-19 from March 14 to April 3, 2022 in units with 5 or more COVID-19-infected HCWs per week. In the viral shedding kinetics study, HCWs with Omicron variant infection who agreed with daily saliva sampling were enrolled between February and March, 2022.
Results
Of the 248 HCWs who were diagnosed with COVID-19 within 5 days of the return of an infected HCW, 18 (7%) had contact with the returned HCW within 1 to 5 days after their return. Of these, 9 (4%) had an epidemiologic link other than with the returning HCW, and 9 (4%) had contact with the returning HCW, without any other epidemiologic link. In the study of the kinetics of virus shedding (n=32), the median time from symptom onset to negative conversion of viable virus was 4 days (95% CI, 3 to 5 days).
Conclusions
Our data suggest that the residual risk of virus transmission after 5 days of isolation following diagnosis or symptom onset is low.
Keyword
COVID-19
SARS-CoV-2
transmission
isolation
==== Body
pmcIntroduction
After the surge in Omicron cases, the Centers for Disease Control and Prevention (CDC) shortened the isolation period for people with COVID-19 from 10 days after symptom onset or positive viral test to 5 days, and recommended wearing a well-fitting mask for 5 days after the 5-day isolation period [1]. Likewise, on January 26, 2022 the Korea Disease Control and Prevention Agency (KDCA) changed the isolation period for the general population with mild disease from 10 days to 7 days after positive viral test [2]. They also announced that the isolation period for HCWs confirmed with COVID-19 and up-to-date with vaccination could be reduced to 3-5 days, and these HCWs could work wearing well-fitted masks after they returned [3].
A previous study has reported that no contacts developed SARS-CoV-2 infection after exposure to a COVID-19 patient 6 days or more after the onset of that patient’s symptoms [4], but there was a worry concerning the risk of transmission from HCWs who return earlier to healthcare facilities. In this study, we used two cohorts to determine, respectively, the risk of onward transmission from HCWs with Omicron variant infections who return to work after a 5-day isolation, and the kinetics of shedding of viable virus from HCWs infected with the Omicron variant.
Methods
This cohort study included the epidemiologic investigation study on the risk of transmission form HCWs who returned to work after a 5-day isolation (cohort 1) and a viral kinetics shedding study (cohort 2).
Cohort 1: risk of transmission from HCWs who return to work after a 5-day isolation
This study was performed in Asan Medical Center, a 2,700-bed tertiary care hospital, in Seoul, South Korea. During the study period, HCWs were recommended to wear universal masking (KF94 [FFP2-equivalent mask]) and eye protection (goggle or faceshield) regardless of whether they were caring for a COVID-19 patients. Patients with COVID-19 were isolated in the single room with or without negative pressure. In a study carried out in units with 5 or more COVID-19-infected HCWs per week from March 14 to April 3, 2022 when the Omicron subvariant BA.2 was dominant [5], we evaluated the number of HCWs subsequently confirmed as COVID-19 occurring 1 to 5 days after the return of HCWs who had been diagnosed as COVID-19 and had completed a 5-day isolation. During the study period, all HCWs with COVID-19 were isolated in their homes from the day of a positive viral test (day 1) to day 5, and they filled out a mobile epidemiologic investigation form which was sent to the infection control team in our hospital. We investigated symptoms at diagnosis, history of exposure to returned HCWs, households, and patients with COVID-19. Exposure to a COVID-19 case (close contact) was defined as being within 2 meters of that case for more than 3 minutes regardless of mask wearing by the index and contact or as being exceeds 2 meters, if HCWs and COVID-19 case taking off the mask to eat or drink in a narrow space. Confirmed SARS-CoV-2 infection was defined as a positive RT-PCR result from a nasopharyngeal (NP) swab or from a rapid antigen test performed by a trained health care professional. HCWs with mild illness or asymptomatic infections and a positive viral test, and who had received 3rd dose vaccinations with mRNA vaccine, were allowed to return after a 5-day isolation. These HCWs received guidance on behavior for 2 days after returning; they were advised to always wear well-fitting KF94 (FFP2-equivalent) masks, dine and relax in a separate space, minimize non-essential contact with colleagues, and restrict face-to-face meetings.
The institutional review board of Asan Medical Center evaluated and approved the medical, scientific, and ethical aspects of the study protocol concerning cohort 1 (2021-0024).
Cohort 2: kinetics of shedding of the Omicron variant
We enrolled HCWs with SARS-CoV-2 infections confirmed as due to the Omicron variant who agreed with daily saliva sampling at Asan Medical Center, in February and March 2022. All enrolled HCWs were instructed to submit daily saliva samples. The institutional review board of Asan Medical Center evaluated and approved the medical, scientific, and ethical aspects of the study protocol concerning cohort 2 (2020-0297).
Saliva Sample Collection and Laboratory Procedures
During the isolation period, all HCWs were instructed to place more than 2 ml of saliva into airtight containers. They were advised not to eat or brush their teeth for at least 30 minutes before providing a sample, and the samples were immediately stored in a freezer at -80°C. Viral RNA was extracted from respiratory specimens using a QIAamp viral RNA Mini kit (Qiagen Inc., Hilden, Germany) followed by manufacturer’s instruction. To detect subgenomic RNA, the reaction mixture included 0.1 μL of 200× enzyme mix, 4 μL of 5× master mix, 1000 nM of leader primer, 500 nM of each S and N gene reverse primer, 250 nM of S and N gene probes, and internal control primers and probe (Supplemental Table 1). In each mixture, 5 μL of extracted RNA or in vitro-synthesized control RNA were added. PCR amplification was performed with a LightCycler 96 system (Roche) in the following conditions: reverse transcription at 50°C for 10 min, initial denaturation at 95°C for 5 min, 45 cycles of 2-step amplification, denaturation at 95°C for 10 s and annealing and elongation at 60°C for 30 s, and final extension at 60°C for 5 min. Viral copy numbers were determined by plotting the Ct values against log copies/reaction. The decision of positive and measurement of viral loads were determined by N gene of SARS-CoV-2.The Omicron variant were detected using a PowerChek SARS-CoV-2 S-gene mutation detection kit Ver.3.0 (Kogenbiotech Co., Ltd).
Definitions
The Omicron variant was defined as detection of the N501Y, K417N, E484A substitutions in the spike protein. If the T547 substitution was also present, the strain was defined as the Omicron BA.2 subvariant. We categorized vaccinated patient into two subgroups: completion of the primary vaccine series (2-dose) and 3-dose vaccinated [6].
Statistical analysis
Kaplan Meier survival analysis was performed based on three different methods of detecting Omicron: genomic RNA PCR, subgenomic RNA PCR, and virus culture. The cut-off value for genomic RNA PCR and subgenomic RNA PCR was 2.6 virus copies/mL, the 95% limit of detection (LOD). For the sake of logical validity, samples identified as culture- or PCR-negative before obtaining a positive sample were neglected in the survival analysis, and the duration of virus shedding was taken to be the period up to the last day that a culture- or PCR-positive sample was obtained. All statistical analyses were done using R for statistics (version 4.1.1).
Results
Cohort 1: risk of transmission from HCWs who returned to work after a 5-day isolation
During the study period, 65 hospital units had 5 or more COVID-19-infected HCWs per week with a total of 736 HCWs with COVID-19. Their median age was 32 [IQR, 26 – 41] years, 550 [75%] were female, and the median days from symptom onset to diagnosis of SARS-CoV-2 infection in the 635 symptomatic HCWs was 0 [IQR, 0 – 1]). In addition, 248 of the HCWs were diagnosed with COVID-19 within 5 days of the return of previously-infected HCWs (median age of 32 [IQR, 27 – 42] years; 175 [71%] of them female). All the HCWs were asymptomatic or had mild illness, and none were immunocompromised. Of the 248 HCWs, 240 (97%) received 3-dose vaccination, and 18 (7%) had contact with returned HCWs from 1 to 5 days after their return. The study flowchart is shown in Supplemental Figure 1. Of these 18, 9 (4%) also had epidemiologic links other than with returned HCWs (exposure to other HCWs during their infectious period before diagnosis [n=3], household contacts [n=3], or exposure to patients with COVID-19 [n=3]), while 8 (3%) had contact with returned HCWs without any epidemiological link from the known COVID-19 patients without appropriate PPEs in the hospital or social & family contacts through interview or epidemiological investigations while wearing KF94 (FFP2-equivalent) masks, and the remaining HCWs (0.4%) ate a meal with a returned HCW. Detailed information concerning the exposure of the latter 9 HCWs is given in Table 1 .Table 1 Characteristics of healthcare workers (HCWs) exposed to returned HCWs between 1 and 5 days after their return (cohort 1)
Table 1Patient No. Day of contact with returned HCW (return day defined as day 0) Mask worn during contact Days from return day of index to day of diagnosis of exposed HCW
1 0, 1 + 3
2 0 + 1
3 1 + 1
4 0, 1 + 3
5 0, 1 + 2
6 0 + 3
7 0 + 3
8 1 + 3
9 0 - 1
Cohort 2: kinetics of shedding of the Omicron variant
A total of 32 HCWs were enrolled. All had received at least 2 doses of COVID-19 vaccines, and has no underlying illnesses. Baseline characteristics are shown in Table 2 . The median age was 28 years [IQR, 26 – 33] and 23 (72%) were female. More than half of the HCWs (62%) had mild COVID-19; the remaining 12 HCWs (38%) were asymptomatic.Table 2 Baseline characteristics of healthcare workers in cohort 2
Table 2Variable
Age, median years (IQR) 28 (26 – 33)
Female Gender 23 (72)
Initial Severity
Asymptomatic 12 (38)
Mild 20 (62)
Type of COVID-19 vaccine
2 doses of ChAdOx-nCoV-19 3 (9)
2 doses of mRNA-1273 1 (3)
2 doses of ChAdOx-nCoV-19, followed by BNT162b2 28 (88)
Vaccination status
2-dose 4 (12)
3-dose 28 (88)
Subvariants
Omicron BA.1. 14 (44)
Omicron BA.2. 18 (56)
Symptoms
Systemic 23 (77)
Gastrointestinal 13 (43)
Respiratory 30 (100)
Sensory 10 (33)
Time from last vaccination to infection, median days (IQR) 97 (82 – 108)a
NOTE. Data are presented as number of healthcare workers with SARS-CoV-2 infection (%) unless otherwise indicated.
a including 1 healthcare worker who did not remember the date of vaccination.
Viable virus was detected in 12 (10%) of the 116 samples (Figure 1 A, Figure 1B, Supplemental Table 2, and Supplemental Figures 2-3). The median time from symptom onset to negative conversion of viable virus was 4 days (95% CI, 3 to 5 days) (Figure 1C), while median times from symptom onset to negative conversion exceeded 8 days (95% CI. 7 to more than 9 days) and 5 days (95% CI, 5 to 7 days) for genomic RNA and subgenomic RNA, respectively (Figure 1C). Survival analysis showed that 16% of the HCWs shed viable virus on symptom onset day 6, and none on day 8 (Figure 1C, Supplemental Figure 2). Based on the date of diagnosis, the median clearance time for culture-based virus detection was 3 days (95% CI, 3 to 4 days), one day less than the duration of symptoms (Supplemental Figure 2). 19% (6/32) of the HCWs were culture-positive on day 5 from diagnosis (Supplemental Figure 2).Figure 1 Timing of presence or absence of viable SARS-CoV-2 on viral culture, and viral copy numbers for 116 serial samples obtained from 32 consecutive healthcare workers with COVID-19. Viral loads were determined copy numbers of the N gene of SARS-CoV-2. Each circle represents a sample obtained on the specified day. (A) Genomic RNA PCR viral copy number and viral culture results. (B) Subgenomic RNA PCR viral copy number and viral copy results. (C) Kaplan-Meier curve for viral clearance.
Figure 1
Discussion
In this study, we found that most of the HCWs who were diagnosed with COVID-19 within 5 days of the day another HCW returned to work had no history of contact with the returned HCW, and only about 4% had epidemiologic links with the returned HCW without any other epidemiologic link. In our study of viral kinetics, the median time from symptom onset to negative conversion of viable virus was median 4 days (95% CI, 3 to 5 days), and 16% of HCWs shed viable virus at symptom onset day 6, and none by day 8. In addition, the previous epidemiologic study [4] reporting that no contacts developed SARS-CoV-2 infection after exposure to a COVID-19 patient 5 days or more after the symptom onset support a 5-day isolation with additional 5 days of high quality mask wearing by CDC. In this contexture, our data build on the current evidences supporting the current CDC’s recommendation that HCWs should wear high-quality mask after their end of isolation and keep it until day 10, depending on the infection control practice and duty arrangement.
Replication-competent virus has been recovered up to 10 days after symptom onset in patients with mild illness [[7], [8], [9]]. However, the recovery of replication-competent virus does not always imply transmissibility, considering the dose required for successful infection. The concentration of SARS-CoV-2 RNA declines after onset of symptoms [10, 11]. Previous epidemiologic study has measured an attack rate of 1% (22 cases from 1818 contacts; 1.0%; 95% CI, 0.6% − 1.6%) among close contacts whose exposure to index cases started within 5 days of the index cases’ symptom onset, whereas for those who were exposed later it was zero (0 cases from 852 contacts; 0%; 95% CI, 0% −0.4%) [4]. Another cohort study also found a higher risk of transmission if exposure occurred between -2 and 3 days from symptom onset in the index patients [12]. During our study, which was performed during a huge Omicron outbreak, multiple exposures were possible, so it was difficult to identify those infections that occurred specifically via epidemiologic links between HCWs. Therefore, our figure for the risk of transmission after a 5-day isolation may be overestimated. In contrast, it is possible that the frequency of transmission through epidemiologic links between close contacts is an underestimate because of the higher frequency of airborne transmission of the Omicron variant than of historical strains [13]. Despite these limitations, our study adds to the evidence for a low risk of onward transmission by 6 days after index case symptom onset in the era dominated by Omicron.
To estimate the kinetics of viable viral shedding of the Omicron variants, we performed culture-based virus isolation and subgenomic RNA detections of spike (S) and nucleocapsid (N) proteins which are surrogate markers of viral infectivity. These two subgenomic genes were chosen because S protein is essential for viral entry to target cells and N protein is a RNA-binding protein critical for viral replication and the most abundant subgenomic RNA in infected cells [14, 15]. Recently, Chen et al. reported that ORF7b gene was the first undetectable subgenomic RNA, thus it may be the most sensitive surrogate marker for viral activity [15]. However, Chen et al. did not assess the relationship between the subgenomic RNA genes and virus culture-isolation. This study evaluated the risk of transmission of Omicron variants, which is highly associated with culturable virus. Phuphuakrat et al. reported that detection of subgenomic N RNA can be a marker for isolation period [16]. Also, our previous study demonstrated that the duration of positive subgenomic N or S RNA appeared to closely reflect the duration of positive culture-isolation [17].
We previously showed that the median duration of negative conversion of viral culture in young patients with the delta variant was 5 days (95% CI, 3 − 6) after symptom onset [18]. Therefore, viable viral shedding of the Omicron variant appears to last 1 day less than with the delta variant, although direct comparison between studies is difficult due to differences in vaccination status and symptom severity. At the time of this writing, Boucau J, et al. reported that the median time from diagnosis day to culture conversion was 5 days (IQR, 3 − 9) in the omicron group and 4 days (IQR, 3 − 5) in the delta group, although there was no significant difference in viable viral shedding between two groups [19]. However, it is difficult to draw a firm conclusion in this issue due to small sample size, so further studies are needed.
Our study has several limitations. First, it was a single center study over a short period, and the number of study participants was small. Second, we did not perform whole genome sequencing, and the epidemiologic investigation was based on self-report. Moreover the study was performed during a huge epidemic; therefore we could not rule out other sources of transmission in the HCWs in cohort 1 who were designated as having only epidemiologic associations with returned HCWs. Furthermore, epidemiologic investigations might miss the unrecognized contacts from multiple sources, so the source of infection might not be definite. Third, we evaluated how much the HCWs with wearing a high-quality mask who returned to work after a 5-day isolation contributed to the transmission to other HCWs in the workplace. So, we did not evaluate the transmission from the HCWs to the patients. Finally, in order to reduce public confusion over the isolation period in asymptomatic individuals with SARS-CoV-2 infection, the South Korean government has recommended that the reference day (day 1) for calculating the isolation period should be the day of diagnosis, because of the difficulty of defining the day of symptom onset. Most HCWs were diagnosed on the day of symptom onset or 1 day later, so the 5-day isolation period from diagnosis is essentially the same as the 5 days from symptom onset (with day 0 as symptom onset day), as currently recommended by the CDC.
In conclusion, shedding of viable Omicron virus lasted a median of 4 days from symptom onset, and less than 5% of the cases occurring over 5-day periods could have involved transmission from HCWs returning after 5 days of isolation. Therefore, it appears that the residual risk of transmission of Omicron by 5 days after diagnosis is low.
Funding
This work was supported by the Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI) funded by the Ministry of Health & Welfare, South Korea [grant number HW22C2045]; and the National Research Foundation of Korea (NRF) funded by the Ministry of Science and ICT, Republic of Korea [grant number NRF-2022M3A9I2017241].
Appendix A Supplementary data
The following is the Supplementary data to this article:
Acknowledgements
There are no conflicts of interest for any of the authors.
Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.jhin.2022.11.012.
==== Refs
References
1 Quarantine and Isolation. Centers for Disease Control and Prevention. Available at: https://www.cdc.gov/coronavirus/2019-ncov/your-health/quarantine-isolation.html. Accessed on April 29, 2022.
2 KCDA Press release. January 24, 2022. Available at: https://www.kdca.go.kr/board/board.es?mid=a20501020000&bid=0015&list_no=718415&cg_code=C01&act=view&nPage=25. Accessed on April 29, 2022.
3 Korean Hospital Association. Healthcare facility Business Continuity Plan (BCP) Guideline. Available at: https://www.kha.or.kr/board/dept/view. Accessed on April 29, 2022..
4 Cheng H.Y. Jian S.W. Liu D.P. Ng T.C. Huang W.T. Lin H.H. Contact tracing assessment of COVID-19 transmission dynamics in taiwan and risk at different exposure periods before and after symptom onset JAMA Intern Med 180 9 2020 1156 1163 32356867
5 KCDA Press release. April 12, 2022. Avialable at: https://www.kdca.go.kr/board/board.es?mid=a20501020000&bid=0015&list_no=719266&cg_code=C01&act=view&nPage=5. Accessed on April 29, 2022.
6 Pekosz A. Parvu V. Li M. Antigen-Based Testing but Not Real-Time Polymerase Chain Reaction Correlates With Severe Acute Respiratory Syndrome Coronavirus 2 Viral Culture Clin Infect Dis 73 2021 e2861 e2866 33479756
7 Rhee C. Kanjilal S. Baker M. Klompas M. Duration of Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) Infectivity: When Is It Safe to Discontinue Isolation? Clin Infect Dis 72 8 2021 1467 1474 33029620
8 Walsh K.A. Spillane S. Comber L. The duration of infectiousness of individuals infected with SARS-CoV-2 J Infect 81 6 2020 847 856 33049331
9 Kim M.C. Cui C. Shin K.R. Duration of Culturable SARS-CoV-2 in Hospitalized Patients with Covid-19 N Engl J Med 384 7 2021 671 673 33503337
10 Zou L. Ruan F. Huang M. SARS-CoV-2 Viral Load in Upper Respiratory Specimens of Infected Patients N Engl J Med 382 12 2020 1177 1179 32074444
11 Owusu D. Pomeroy M.A. Lewis L.M. Persistent SARS-CoV-2 RNA Shedding Without Evidence of Infectiousness: A Cohort Study of Individuals With COVID-19 J Infect Dis 224 8 2021 1362 1371 33649773
12 Ge Y. Martinez L. Sun S. COVID-19 Transmission Dynamics Among Close Contacts of Index Patients With COVID-19: A Population-Based Cohort Study in Zhejiang Province, China JAMA Intern Med 181 10 2021 1343 1350 34424260
13 Rowe B.R. Canosa A. Meslem A. Rowe F. Increased airborne transmission of COVID-19 with new variants, implications for health policies Build Environ 219 2022 109132
14 Cubuk J. Alston J.J. Incicco J.J. Singh S. Stuchell-Brereton M.D. Ward M.D. Zimmerman M.I. Vithani N. Griffith D. Wagoner J.A. Bowman G.R. Hall K.B. Soranno A. Holehouse A.S. The SARS-CoV-2 nucleocapsid protein is dynamic, disordered, and phase separates with RNA Nat Commun 12 2021 1936 33782395
15 Chen Z. Ng R.W.Y. Lui G. Ling L. Chow C. Yeung A.C.M. Boon S.S. Wang M.H. Chan K.C.C. Chan R.W.Y. Hui D.S.C. Chan P.K.S. Profiling of SARS-CoV-2 Subgenomic RNAs in Clinical Specimens Microbiol Spectr 10 2022 e0018222
16 Phuphuakrat A. Pasomsub E. Srichatrapimuk S. Kirdlarp S. Suksatu A. Srisaowakarn C. Manopwisedjaroen S. Ludowyke N. Purwono P.B. Priengprom T. Wongsa A. Thakkinstian A. Hongeng S. Malathum K. Thitithanyanont A. Tassaneetrithep B. Detectable Duration of Viable SARS-CoV-2, Total and Subgenomic SARS-CoV-2 RNA in Noncritically Ill COVID-19 Patients: a Prospective Cohort Study Microbiol Spectr 10 3 2022 Jun 29 e0050322
17 Kim J.Y. Bae J.Y. Bae S. Cha H.H. Kwon J.S. Suh M.H. Lee H.J. Jung J. Kim M.J. Cui C. Park H. Lee J. Park M.S. Kim S.H. Diagnostic usefulness of subgenomic RNA detection of viable SARS-CoV-2 in patients with COVID-19 Clin Microbiol Infect 28 1 2022 Jan 101 106 34400343
18 Park S. Lim S.Y. Kim J.Y. Clinical and virological characteristics of SARS-CoV-2 B.1.617.2 (Delta) variant: a prospective cohort study Clin Infect Dis 2022 E-pub ahead of print
19 Boucau J. Marino C. Regan J. Duration of Shedding of Culturable Virus in SARS-CoV-2 Omicron (BA.1) Infection N Engl J Med 2022 E-pub ahead of print
| 36460176 | PMC9705265 | NO-CC CODE | 2022-12-01 23:19:34 | no | J Hosp Infect. 2022 Nov 29; doi: 10.1016/j.jhin.2022.11.012 | utf-8 | J Hosp Infect | 2,022 | 10.1016/j.jhin.2022.11.012 | oa_other |
==== Front
Acta Trop
Acta Trop
Acta Tropica
0001-706X
1873-6254
Elsevier B.V.
S0001-706X(22)00472-7
10.1016/j.actatropica.2022.106781
106781
Article
Peptide microarray analysis of in-silico predicted B-cell epitopes in SARS-CoV-2 sero-positive healthcare workers in Bulawayo, Zimbabwe
Vengesai Arthur a⁎
Naicker Thajasvarie b
Midzi Herald c
Kasambala Maritha d
Muleya Victor a
Chipako Isaac e
Choto Emilia f
Moyo Praise g
Mduluza Takafira c
a Department of Biochemistry, Faculty of Medicine and Health Sciences, Midlands State University, Senga Road, Gweru, Zimbabwe
b Discipline of Optics and Imaging, Doris Duke Medical Research Institute, University of KwaZulu-Natal College of Health Sciences Durban, ZA
c Department of Biotechnology and Biochemistry, Faculty of Science, University of Zimbabwe, Harare, Zimbabwe
d Department of Biological Sciences and Ecology, Faculty of Science, University of Zimbabwe, Harare, Zimbabwe
e Aravas Pharmaceuticals Pvt LTD, Prospect Industrial Area, Harare, Zimbabwe
f Immunology Department, Simon Mazorodze School of Medical and Health Sciences, Great Zimbabwe University, Masvingo, Zimbabwe
g Department of Applied Biosciences and Biotechnology, Faculty of Science and Technology, Midlands State University, Senga Road, Gweru, Zimbabwe
⁎ Corresponding author: Dr. Arthur Vengesai, Midlands State University, Gweru, Zimbabwe
29 11 2022
29 11 2022
10678113 7 2022
31 10 2022
27 11 2022
© 2022 Elsevier B.V. All rights reserved.
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Immunogenic peptides that mimic linear B-cell epitopes coupled with immunoassay validation may improve serological tests for emerging diseases. This study reports a general approach for profiling linear B-cell epitopes derived from SARS-CoV-2 using an in-silico method and peptide microarray immunoassay, using healthcare workers’ SARS-CoV-2 sero-positive sera. SARS-CoV-2 was tested using rapid chromatographic immunoassays and real-time reverse-transcriptase polymerase chain reaction. Immunogenic peptides mimicking linear B-cell epitopes were predicted in-silico using ABCpred. Peptides with the lowest sequence identity with human protein and proteins from other human pathogens were selected using the NCBI Protein BLAST. IgG and IgM antibodies against the SARS-CoV-2 spike protein, membrane glycoprotein and nucleocapsid derived peptides were measured in sera using peptide microarray immunoassay. Fifty-three healthcare workers included in the study were RT-PCR negative for SARS-CoV-2. Using rapid chromatographic immunoassays, 10 were SARS-CoV-2 IgM sero-positive and 7 were SARS-CoV-2 IgG sero-positive. From a total of 10 SARS-CoV-2 peptides contained on the microarray, 3 (QTH34388.1-1-14, QTN64908.1-135-148, and QLL35955.1-22-35) showed reactivity against IgG. Three peptides (QSM17284.1-76-89, QTN64908.1-135-148 and QPK73947.1-8-21) also showed reactivity against IgM. Based on the results we predicted one peptide (QSM17284.1-76-89) that had an acceptable diagnostic performance. Peptide QSM17284.1-76-89 was able to detect IgM antibodies against SARS-CoV-2 with area under the curve (AUC) 0.781 when compared to commercial antibody tests. In conclusion in silico peptide prediction and peptide microarray technology may provide a platform for the development of serological tests for emerging infectious diseases such as COVID-19. However, we recommend using at least three in-silico peptide prediction tools to improve the sensitivity and specificity of B-cell epitope prediction, to predict peptides with excellent diagnostic performances.
Graphical abstract
Image, graphical abstract
Keywords
SARS-CoV-2
B-cell epitopes
epitope prediction
peptide microarrays
serological tests
==== Body
pmc1 Introduction
Diagnostic methods are crucial for the control of the ongoing coronavirus diseases of 2019 (COVID-19) pandemic, triggered by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) (Ferreira et al., 2021 Nov 30, Musicò et al., 2021 Jan 11). Though real-time reverse-transcriptase polymerase chain reaction (RT-PCR) is the most established method in detecting active or current SARS-CoV-2 infections (Rai et al., 2021 Jan 1) serological tests (or equivalently, serology, or antibody testing) are useful in resource-limited settings (Ong et al., 2021 Jul 1, Lagatie et al., 2017 Dec 1). Serological tests are primarily useful to determine whether people were previously infected by SARS-CoV-2 (West and Kobokovich, 2020). This is important at the population level to support surveillance studies, to determine the extent of exposure, case fatality rate, and to track changes in incidence and prevalence (Rai et al., 2021 Jan 1, West and Kobokovich, 2020, Ludolf et al., 2022 May 1).
With inadequate access to reagents and equipment, restrictive biosafety level facilities and technical sophistication (Javadi Mamaghani et al., 2021 Dec 1, Vengesai et al., 2021 Dec 1), serology offers an alternative method for screening SARS-CoV-2 (West et al., 2021 Mar 1). However, rapid antigen tests should be considered, as they can diagnose current viral infections unlike serological tests (Ong et al., 2021 Jul 1). Moreover, serological tests may be utilized when molecular or antigen tests results are inconclusive. Serological tests may be helpful in highly suspicious cases with negative molecular tests. In those circumstances, serologic tests may help explain clinical symptom (Ong et al., 2021 Jul 1). Some studies have reported that combining RT-PCR test and serological testing improve the overall sensitivity for detecting SARS-CoV-2 (Fokam et al., 2022, Sidiq et al., 2020 Dec 1).
SARS-CoV-2 infection history could be important for future medical management in cases of late-onset post-infection complications. Serological testing to identify past SARS-CoV-2 infection could be particularly helpful in paediatric patients with multisystem inflammatory syndrome (West et al., 2021 Mar 1). Furthermore, serology tests are often necessary to determine potential donors of convalescent plasma, a therapy for COVID-19 given to patients with active infections to boost their immune response (West et al., 2021 Mar 1, Lu et al., 2020 Dec 1).
Recombinant proteins, especially the S1 protein, are one of the major reagents for SARS-CoV-2 serological tests (Li et al., 2021 Mar 1). Despite recombinant protein based serological tests offering the advantages of (i) working without stringent biosafety requirements, (ii) easy assay standardization and (iii) availability in short periods of time, these tests have several shortcomings (Sidiq et al., 2020 Dec 1). In addition to high spike protein production costs and storage constraints, full length recombinant proteins have significant instability problems, they suffer from batch-to-batch variations and in some cases may cross-react with other coronaviruses (Musicò et al., 2021 Jan 11, Javadi Mamaghani et al., 2021 Dec 1, Li et al., 2021). The biggest challenge with cross-reactivity is that it leads to false positive results: a person previously exposed to SARS-CoV may test positive for SARS-CoV-2, although uninfected (Rai et al., 2021 Jan 1). Expressing S1 protein in the correct conformation is difficult, and in some cases, the antibodies that recognize the membrane spike protein are unable to bind recombinant S1 protein (Sidiq et al., 2020 Dec 1, Li et al., 2021 Mar 1). Additionally, currently available serological tests have sub-optimal diagnostic accuracies (EUA Authorized Serology Test Performance | FDA [Internet]). Sidiq and colleagues reported that several studies showed different reactivity of SARS-CoV-2 patient sera with spike protein, ranging from very low (13%) to 100%, depending on the test used (Sidiq et al., 2020 Dec 1).
Immunogenic peptides mimicking linear B-cell epitopes may improve SARS-CoV-2 serological testing (Javadi Mamaghani et al., 2021 Dec 1). Peptide-based serological tests are advantageous, compared to recombinant proteins based serological tests. The main reasons for this are that synthetic peptides are well defined, easily produced in large amounts when required, highly pure with almost no batch-to-batch variation and they are very stable. Moreover, synthetic peptide use is often cost-effective, compared to the production of the spike protein (Li et al., 2021, List et al., 2010 Aug 3, Vengesai et al., 2021 Jul, Qi et al., 2020 Sep 9).
We previously reported that six methods can be used for the identification and prediction of linear B-cell epitopes (peptides) (Vengesai et al., 2022 Jan). For comprehensive mapping of B-cell epitopes, experimental techniques including overlapping peptides and phage display library are time consuming and expensive even for SARS-CoV-2 which has relatively few genes (Vengesai et al., 2021 Mar 4). In contrast, in-silico B-cell epitope prediction bioinformatics techniques are manageable alternatives that allow for virtual cost-effective screening in the search for immuno-dominant epitopes with serological diagnostic potential (Van Regenmortel, 1989, 24, Giacò et al., 2012 Jul 25).
Several in-silico B-cell epitope prediction bioinformatics databases (Table 1 ) are available, in which computational strategies guide the selection of candidate epitopes for peptide microarray technology validation (Vengesai et al., 2022 Jan). For identification of effective vaccine candidates against SARS-CoV-2, Kumar et al,(2021) used various software and online bioinformatics tools including ABCpred, to predict B cell epitopes (Jena et al., 2021 Jan 23). Moreover, Singh et al., (2020) predicted SARS-CoV-2 continuous and discontinuous B-cell epitopes using BCpred 2.0 and Ellipro server, in-silico tools respectively (Singh et al., 2020). In this study we report on a general approach to discover immunogenic peptides mimicking linear B-cell epitopes derived from SARS-CoV-2 spike protein, membrane glycoprotein and nucleocapsid protein using ABCpred in-silico B-cell epitope prediction and peptide microarray immunoassay validation.Table 1 In-silico B-cell epitope prediction software (Vengesai et al., 2022 Jan).
Table 1:Software Server
APRANK https://github.com/trypanosomatics/aprank)
MLCE http://bioinf.uab.es/BEPPE
ABCpred http://www.imtech.res.in/raghava/abcpred/
BepiPred 1.0 www.cbs.dtu.dk/services/BepiPred/
Epitopia web server http://epitopia.tau.ac.il/
Antigenic http://www.bioinformatics.nl/cgi-bin/emboss/antigenic
BCPREDS http://ailab.ist.psu.edu/bcpred/
Bcepred http://crdd.osdd.net/raghava/bcepred/bcepred_instructions.html
2 Materials and Methods
2.1 Ethical approval and Study population
Ethical approval for the study was obtained from the Medical Research Council of Zimbabwe (MCRZ/A/2571/ and MRCZ/A2443/). Fifty-three healthcare workers (cleaners, security officers, nurses, administrators, and doctors) were recruited into the study from health facilities in Bulawayo, Zimbabwe (20.1457˚ S, 28.5873˚ E) in June 2020 prior to COVID-19 vaccination rollout. Prior to recruitment, the study objectives were fully explained to the healthcare workers who then gave their written consent to participate in the study. Negative sera were obtained from Schistosoma mansoni infected individuals collected in 2015 before the COVID-19 pandemic.
2.2 Antibody testing/ Serological test
Five millilitres of venous blood were collected from each worker. The blood was separated into serum samples within 24hrs of collection by centrifugation at 3000g for 15 minutes. The serum was used to detect SARS-CoV-2 antibodies (IgM and IgG) using two rapid immunoassay kits (Wuhan UNscience Biotechnology Companies UNICOV-40 test kit and Standard-Q Covid-19 IgM/IgG Duo antibody test from SD Biosensor) following manufacturer's guidelines .
Both test kits detect the presence of IgM and IgG antibodies directed against the nucleocapsid and the spike proteins of SARS-CoV-2 (Rusakaniko et al., 2021 Mar 1, SD BIOSENSOR | PRODUCTS [Internet]).
2.3 SARS-CoV-2 Real-Time reverse transcriptase (RT)-PCR-diagnosis
Clinicians collected nasopharyngeal swabs according to WHO and CDC protocols (https://www.who.int/docs/default source/coronaviruse/whoinhouseassays.pdf). RNA was then extracted from these swabs using the respiratory sample RNA isolation kit. Diagnosis of SARS-CoV-2 virus was performed using Real-Time reverse transcriptase (RT)-PCR as described by Rusakaniko et al (2021). The nucleocapsid protein gene and the virus open reading frame 1ab (ORF1ab) gene were amplified simultaneously as recommended by WHO. An internal control (RNasep) gene was used together with negative and positive samples in the assay (Rusakaniko et al., 2021 Mar 1).
2.4 Peptide selection
SARS-CoV-2 spike protein, nucleocapsid protein membrane glycoprotein sequences were obtained from the NCBI protein database (https://www.ncbi.nlm.nih.gov/). The bioinformatics tool ABCpred was used for the in silico prediction of the SARS-CoV-2 B-cell epitopes on the selected protein sequences. Peptides with the lowest sequence identity with human protein and proteins from other human pathogens were selected using the NCBI Protein BLAST (https://blast.ncbi.nlm.nih.gov/Blast.cgi select Protein BLAST) to minimize cross reactivity. Peptides that had the highest ABCpred rank (higher probability of being B-cell epitopes) and the lowest sequence identity with peptides from other human pathogens or proteins were selected for inclusion on the peptide microarrays.
2.5 Peptide microarray design and layout
2.5.1 Peptide microarray design
The peptide microarray was designed to contain 9aa-18aa peptides that were obtained from a variety of pathogens and printed in a laser-printer technique by PEPperPRINT GmbH (Heidelberg, Germany) (https://www.pepperprint.com/). Each sub-array on the peptide microarray contained 260 peptide positions. The microarray had 16 sub-arrays (copies) that were framed by flag anti-polio (KEVPALTAVETGAT, 3 spots) and flag anti-HA (hemagglutinin glycoprotein of influenza virus) (YPYDVPDYAG, 3 spots) as a quality control measurement. Ten SARS-CoV-2 peptides (14aa and 16aa) (details given in Table 2 ) were printed with random distribution across each sub-array.Table 2 ABCpred predicted B-cell linear epitopes
Table 2:Peptide name Source Protein Peptide Sequence Peptide length
PDB: 7KRQ_A-879-894 Chain A, spike glycoprotein AGTITSGWTFGAGAAL 16
PDB: 7KRQ_A-257-272 Chain, A spike glycoprotein GWTAGAAAYYVGYLQP 16
QPK73947.1-8-21 membrane glycoprotein ITVEELKKLLEQWN 14
PDB: 7LX5_B-686-701 Chain B, spike glycoprotein GVSVITPGTNTSNQVA 16
QSM17284.1-76-89 nucleocapsid protein TNSSPDDQIGYYRR 14
QLL35955.1-22-35 nucleocapsid protein DGKMKDLSPRWYFY 14
QTH34388.1-1-14 membrane glycoprotein MADSNGTITVEELK 14
QTN64908.1-135-148 membrane glycoprotein ESELVIGAVILRGH 14
QRU89900.1-41-54 nucleocapsid protein RPQGLPNNTASWFT 14
QTN64908.1-136-149 membrane glycoprotein SELVIGAVILRGHL 14
The peptide name consisted of the protein/antigen accession number followed by the amino acids’ positions of the peptide in the protein.
2.5.2 Peptide microarrays immunoassays
The immunoassays consisted of two steps on the same microarray: the pre-incubation step for identifying false positive signals by binding of the fluorescently labelled secondary antibody followed by the main incubation with serum and the secondary antibodies. Each step involved pre-swelling of the peptide microarray with washing buffer (PBS, pH 7.4 with 0.05% Tween 20) for 10 minutes, followed by incubation with a blocking buffer (Rockland blocking buffer MB-070) for 30 minutes. Initially, the peptide microarrays were incubated with secondary antibodies [Goat anti-human IgG (Fc) DyLight680 (0.1 µg/ml) and goat anti-human IgM (µ chain) DyLight800 (0.2 µg/ml)] and control antibodies [Mouse monoclonal anti-HA DyLight800 (0.5 µg/ml) and mouse monoclonal anti-polio DyLight800 (0.5 µg/ml] diluted in incubation buffer (washing buffer with 10% blocking buffer) at room temperature for 45 minutes. In the main step the microarrays were incubated with serum or plasma diluted 1:250 in incubation buffer for 16 h at 4°C and 140 rpm orbital shaking followed by incubation with the secondary antibodies. After each incubation step the microarrays were washed three times with washing buffer for 10 seconds. The microarrays were scanned using an LI-COR Odyssey Imaging System; scanning offset 0.65 mm, resolution 21 µm, scanning intensities of 7/7 (red = 680 nm / green = 800 nm). To ensure that all microarrays were responding correctly, all steps were repeated with the Cy3-conjugated anti-HA control antibody and Cy3-conjugated anti-polio control antibodies.
2.5.3 Image analysis and spot intensity quantification
Microarray image analysis was done using PepSlide® Analyzer (SICASYS Software GmbH; Heidelberg, Germany) and resulted in raw data CSV files for each sample (green = 800 nm = IgM staining, red = 680 nm = IgG staining). Quantification of spot intensities was based on 16-bit gray scale tiff files. Averaged median foreground intensities (foreground-background signal) and spot-to-spot deviations of spot duplicates were calculated using a PEPperPRINT software algorithm. For duplicate spots a maximum spot-to-spot deviation of 50% was tolerated, otherwise the corresponding intensity value was regarded as artefact and was zeroed.
2.6 Statistical analysis
The data set used for statistical analysis of the peptide microarray results were based on fluorescence intensity. Non-parametric statistical methods were used for data analysis with p-values < 0.05 considered statistically significant. Receiver operating characteristic (ROC) curves and area under the curve (AUC) were obtained using the web-based calculator for ROC curves, format 5 continuous rating scale. Bar graphs were created in Microsoft excel 2013.
2.7 Antibody reactivity and discrimination between the infected and uninfected groups by detection of immunodominant epitopes
The negative cut-off was determined by averaging the negative control readings (sera from 19 Schistosomiasis mansoni infected individuals collected in 2015 prior the COVID-19 pandemic) and adding 3 standard deviations. A positive response was defined as fluorescence intensity ≥ 500 FU (fluorescence intensity units) (Schwarz et al., 2021 Mar 23) and fluorescence intensity above the negative cut-off for each specific peptide for both IgG and IgM. Peptides not reactivity with at least one sero-positive sera were considered to be not reactive with SARS-CoV-2 antibodies. ROC curve analysis assessed diagnostic accuracy of the peptides and AUC assessed the overall diagnostic performance of the peptide.
3 Results
3.1 Demographic and clinical characteristics
The study population consisted of forty-nine health care workers [14.3% (Ludolf et al., 2022 May 1) males and 85.7% (Ceraolo and Giorgi, 2020 May 1) females] within the age range of 20-64 years (median age: 38.9; IQR: 29-23) from Bulawayo. The cohort of health workers were of two different health settings [87.8% (Carmona et al., 2012 Dec 14) hospital and 12.2% (West and Kobokovich, 2020) clinic] and comprised of 53.1% (Jena et al., 2021 Jan 23) nurses, 2% (Ferreira et al., 2021 Nov 30) doctor, 16.3% (Javadi Mamaghani et al., 2021 Dec 1) nurse aides, 16.3% (Javadi Mamaghani et al., 2021 Dec 1) student nurses, 8.2% (Ong et al., 2021 Jul 1) general and 4.1% (Musicò et al., 2021 Jan 11) clerks. Four workers who did not have demographic data were also included in the study. All individuals tested negative for SARS-CoV-2 using RT-PCR. Using rapid chromatographic immunoassays, 10 were SARS-CoV-2 IgM positive and 7 were SARS-CoV-2 IgG positive.
3.2 SARS-CoV-2 B-cell epitope profiling
Ten SARS-CoV-2 ABCpred in-silico predicted peptides were screened on a peptide microarray platform and five peptides were reactive (Fig 1 - 2 and supplementary file 1). For IgG, two were reactive; QTH34388.1-1-14 derived from the membrane glycoprotein (reactive with nine sero-negative sera and one sero-positive serum) and QLL35955.1-22-35 derived from nucleocapsid protein (reactive with one sero-positive serum). With respect to IgM, two peptides were also reactive; QSM17284.1-76-89 derived from nucleocapsid protein (reactive with one sero-negative serum and one sero-positive serum), QPK73947.1-8-21 derived from membrane glycoprotein (reactive with one sero-negative serum and one sero-positive serum). Peptide QTN64908.1-135-148 was reactive with both SARS-CoV-2 IgG (reactive with three sero-negative sera and three sero-positive sera) and IgM (reactive with five sero-negative sera and two sero-positive sera). None of the peptides that reacted with SARS-CoV-2 IgG were able to distinguish between SARS-CoV-2 sero-negative and sero-positive sera with AUC below 0.5 (Fig 3 ). Peptide QSM17284.1-76-89 was able to detect IgM antibodies against SARS-CoV-2 with an acceptable AUC of 0.781 when compared to commercial antibody tests (Fig 4 ).Fig. 1 IgM evaluation in serum and plasma samples derived SARS-CoV-2 sero-negative and sero-positive healthcare workers against SARS-CoV-2 predicted peptides.
Fig 1
Fig. 2 IgG evaluation in serum and plasma samples derived SARS-CoV-2 sero-negative and sero-positive healthcare workers against SARS-CoV-2 predicted peptides.
Fig 2
Fig. 3 Evaluation of the accuracy of SARS-CoV-2 reactive peptides against IgG. The diagnostic accuracy was determined by ROC curves. RED symbols and BLUE line indicate the fitted ROC curve. GRAY lines indicate the 95% confidence interval of the fitted ROC curve.
Fig 3
Fig. 4 Evaluation of the accuracy of SARS-CoV-2 reactive peptides against IgM. The diagnostic accuracy was determined by ROC curves. RED symbols and BLUE line indicate the fitted ROC curve. GRAY lines indicate the 95% confidence interval of the fitted ROC curve.
Fig 4
4 Discussion
Peptide microarray technology is an ideal tool to decipher epitope-specific B-cell immune responses toward the proteome of an emerging pathogen including SARS-CoV-2. The technology enables the simultaneous analysis of peptides in a fast and cost-effective way for applications, such as epitope discovery (Deeks et al., 2020). SARS-CoV-2 has relatively few numbers of proteins, classified as either structural or non-structural. The spike glycoprotein, nucleocapsid protein, membrane glycoproteins and envelope protein are the main structural proteins (Farrera-Soler et al., 2020 Sep 1). In light of this background, this study focused on the spike protein, nucleocapsid protein, membrane glycoproteins. Ten SARS-CoV-2 peptides were predicted in-silico with ABCpred, a widely used, freely available and user-friendly online tool. Following epitope prediction, peptide microarrays were generated in a laser-printer based approach by PEPperPRINT and evaluated with SARS-CoV-2 sero-positive and sero-negative sera.
Despite several studies identifying linear B-cell epitopes with good diagnostic performances in discriminating SARS-CoV-2 positive individuals from SARS-CoV-2 negative individuals (Musicò et al., 2021 Jan 11, Li et al., 2021, Schwarz et al., 2021 Mar 23, Holenya et al., 2021 Jul 1), our approach however identified one peptide (QSM17284.1-76-89) with an acceptable diagnostic performance. Tatjana and colleagues identified linear B-cell epitopes potentially applicable for early and or late COVID-19 diagnosis (Schwarz et al., 2021 Mar 23). Musico and colleagues (2021), identified an epitope of the nucleocapsid protein (region 155-71) with good diagnostic performance (92% sensitivity and 100%) in discriminating SARS-CoV-2 positive individuals from healthy individuals (Musicò et al., 2021 Jan 11). Yang Li and colleagues (2021) identified several spike protein-derived 12-mer peptides that have high diagnostic performance. One peptide (aa 1148–1159 or S2–78) in particular exhibited a sensitivity of 95.5%, 95% CI 93.7–96.9% and specificity of 96.7%, 95% CI 94.8–98.0% (Li et al., 2021).
However, it should be noted that the studies that identified linear B-cell epitopes with good diagnostic performances used whole-proteome peptide microarray analysis whilst in our current in-silico-based study only a few peptides were selected. The spike glycoprotein is transcribed into 1273 amino acid (aa), envelope protein into 76 aa, membrane protein into 220 aa to 260 aa, and nucleocapsid protein into 419 aa (Satarker and Nampoothiri, 2020). In the present study, only three 16 aa non-overlapping peptides covering approximately 4 % of protein sequence were selected for the spike protein, only three 14 aa non-overlapping peptides covering approximately 10 % of protein sequence were selected for the nucleocapsid protein and with respect to membrane glycoprotein, four 14 aa non-overlapping peptides covering approximately 20 % of protein sequence were selected. The implication for such selection is that potential immunogenic peptides may be missed and we recommend including all predicted peptides in prospective studies.
Several SARS-CoV-2 studies have reported antibody reactivity against the spike protein, nucleocapsid protein and membrane proteins with binding predominantly occurring on the spike protein and nucleocapsid protein, indicating that these two proteins are immunodominant (Holenya et al., 2020 Nov 27, Poh et al., 2020 Dec 1, Wang et al., 2020 Dec 23, Lopandić et al., 2021 May 1). However, we detected the nucleocapsid protein and membrane glycoproteins antibody reactivity, suggesting possible early infection as it has been postulated that antibody to the nucleocapsid protein is more sensitive than the spike protein antibody for detecting early SARS-CoV-2 infection (Burbelo et al., 2020 Apr 24).
One of the principal condition in antibody testing is to ensure that there is limited cross-reactivity of antibodies to other antigens (40). Antibody tests for SARS-CoV-2 infection are impeded by immunological cross-reactivity among the human coronaviruses. The SARS-CoV-1 and SARS-CoV-2 genomes are highly similar. SARS-CoV-2 has ∼30 kb positive-sense single-stranded RNA genome which shares ∼80% sequence identity with that of SARS-CoV-1 (Wang et al., 2020 Dec 23, Arya et al., 2021). Consequently, many of the proteins found in SARS-CoV-2 (NC_045512.2) are also found in SARS-CoV-1 (AY515512.1 or NC_004718.3) with 77.1% of the protein sequences shared in their proteomes (Ceraolo and Giorgi, 2020 May 1). To increase the probability of selecting SARS-CoV-2 unique peptides and mitigate the limitation posed by cross reactivity, peptides with the highest ABCpred rank and with the least cross-reactivity with proteins from other human pathogens (especially SARS-CoV-1 and Middle East respiratory syndrome-related coronavirus) were selected for inclusion in the peptide microarrays.
5 Limitations
In-silico prediction of B-cell epitopes is still an active biotechnology research field and a number of servers show improved performance, albeit with prediction accuracies that are still not satisfactory. ABCpred server predict B cell epitopes in an antigen sequence with 65.93 % accuracy using artificial recurrent neural network (machine based technique) (Saha S). Current B-cell epitope predictors are also based on epitopes derived from heterogeneous experimental conditions including many cases in which laboratory animals were immunized with relatively large doses of highly purified antigens. Unfortunately, it has been reported that humoral immune responses against the same antigen differ between species and members of the same species. Moreover, significant variability in individual B-cell epitope reactivity has been reported in tuberculosis and toxoplasmosis (Carmona et al., 2012 Dec 14).
6 Conclusion
Three peptides (QTH34388.1-1-14, QTN64908.1-135-148, and QLL35955.1-22-35) showed reactivity against SAR-CoV-2 IgG. Three peptides (QSM17284.1-76-89, QTN64908.1-135-148 and QPK73947.1-8-21) showed reactivity against SARS-CoV-2 IgM. The reactive peptides were derived from the membrane glycoprotein and nucleocapsid protein. Peptide QSM17284.1-76-89 was able to detect IgM antibodies against SARS-CoV-2 with area under the curve of 0.781 when compared to commercial serological tests. In conclusion in silico peptide prediction and peptide microarray technology provide a powerful platform for the development of new serological tests for emerging infectious diseases.
7 Data availability statement
All relevant data are within the manuscript and its Supporting Information files.
8 Author Agreement Statement
We the undersigned declare that this manuscript is original, has not been published before and is not currently being considered for publication elsewhere.
We confirm that the manuscript has been read and approved by all named authors and that there are no other persons who satisfied the criteria for authorship but are not listed. We further confirm that the order of authors listed in the manuscript has been approved by all of us.
We understand that the Corresponding Author is the sole contact for the Editorial process. He/she is responsible for communicating with the other authors about progress, submissions of revisions and final approval of proofs
Signed by all authors as follows:
Arthur Vengesai, PhD Thajasvarie Naicker, Professor Herald Midzi, Msc
Maritha Kasambala, MSC Victor Muleya, PhD Isaac Chipako, MSc
Emilia Choto, PhD
Praise Moyo, MSc
Takafira Mduluza, Professor
Uncited References
(La Marca et al., 2020)
Supporting information
Supplementary file 1: Demographic, parasitology, and peptide microarray immunoassay data set.
CRediT authorship contribution statement
Arthur Vengesai: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Writing – review & editing. Thajasvarie Naicker: Supervision, Writing – review & editing. Herald Midzi: Data curation, Writing – review & editing. Maritha Kasambala: Data curation, Writing – review & editing. Victor Muleya: Data curation, Formal analysis, Writing – review & editing. Isaac Chipako: Data curation, Writing – review & editing. Emilia Choto: Data curation, Writing – review & editing. Praise Moyo: Data curation, Writing – review & editing. Takafira Mduluza: Conceptualization, Funding acquisition, Investigation, Supervision, Writing – review & editing.
Declaration of Competing Interest
The authors declare that they have no competing interests.
Appendix Supplementary materials
Image, application 1
Data Availability
Data will be made available on request.
Funding
This research was commissioned by the National Institute for Health Research (NIHR) Global Health Research programme (16/136/33) using UK aid from the UK Government. The views expressed in this publication are those of the author(s) and not necessarily those of the NIHR or the Department of Health and Social Care.
Acknowledgments
The authors would like to thank Professor Francisca Mutapi funding acquisition. The authors would also like to acknowledge the valuable input of Professor Francisca Mutapi and Professor Simbarashe Rusakaniko. We would also like to acknowledge, Tackling Infections to Benefit Africa (TIBA) for training on the selection of B-cell epitopes.
Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.actatropica.2022.106781.
==== Refs
References
Ferreira CS Martins YC Souza RC Vasconcelos ATR. EpiCurator: an immunoinformatic workflow to predict and prioritize SARS-CoV-2 epitopes PeerJ [Internet] 9 2021 Nov 30 e12548 [cited 2022 Jun 23]Available from: https://peerj.com/articles/12548 34909278
Musicò A Frigerio R Mussida A Barzon L Sinigaglia A Riccetti S SARS-CoV-2 Epitope Mapping on Microarrays Highlights Strong Immune-Response to N Protein Region Vaccines 2021, Vol 9, Page 35 [Internet]. 9 1 2021 Jan 11 35 [cited 2022 Jun 23]Available from: https://www.mdpi.com/2076-393X/9/1/35/htm
Rai P Kumar BK Deekshit VK Karunasagar I Karunasagar I. Detection technologies and recent developments in the diagnosis of COVID-19 infection Appl Microbiol Biotechnol [Internet] 105 2 2021 Jan 1 441 455 [cited 2022 Oct 6]Available from: https://link.springer.com/article/10.1007/s00253-020-11061-5 33394144
Ong DSY Fragkou PC Schweitzer VA Chemaly RF Moschopoulos CD Skevaki C. How to interpret and use COVID-19 serology and immunology tests Clin Microbiol Infect [Internet] 27 7 2021 Jul 1 981 986 [cited 2022 Oct 6]Available from: https://www.clinicalmicrobiologyandinfection.com/article/S1198743X21002214/fulltext 33975005
Lagatie O Van Loy T Tritsmans L Stuyver LJ Stansfield SH Patel P No Title PLoS One [Internet] 8 1 2017 Dec 1 317 325 [cited 2020 Nov 22]Available from: https://pubmed-ncbi-nlm-nih-gov.ukzn.idm.oclc.org/25836992/
West R, Kobokovich A. Understanding the Accuracy of Diagnostic and Serology Tests: Sensitivity and Specificity Factsheet. 2020;
Ludolf F Ramos FF Bagno FF Oliveira-Da-Silva JA Reis TAR Christodoulides M Detecting anti–SARS-CoV-2 antibodies in urine samples: A noninvasive and sensitive way to assay COVID-19 immune conversion Sci Adv [Internet] 8 19 2022 May 1 [cited 2022 Oct 6]Available from: https://www.science.org/doi/10.1126/sciadv.abn7424
Javadi Mamaghani A, Arab-Mazar Z, Heidarzadeh S, Ranjbar MM, Molazadeh S, Rashidi S, et al. In-silico design of a multi-epitope for developing sero-diagnosis detection of SARS-CoV-2 using spike glycoprotein and nucleocapsid antigens. 2021 Dec 1 [cited 2022 Jun 11];10(1):1–15. Available from: https://pubmed.ncbi.nlm.nih.gov/34849326/
Vengesai A Midzi H Kasambala M Mutandadzi H Mduluza-Jokonya TL Rusakaniko S A systematic and meta-analysis review on the diagnostic accuracy of antibodies in the serological diagnosis of COVID-19 Syst Rev [Internet]. 10 1 2021 Dec 1 1 23 [cited 2022 Jun 14]Available from: https://systematicreviewsjournal.biomedcentral.com/articles/10.1186/s13643-021-01689-3 33388080
West R Kobokovich A Connell N Gronvall GK. COVID-19 Antibody Tests: A Valuable Public Health Tool with Limited Relevance to Individuals Trends Microbiol [Internet] 29 3 2021 Mar 1 214 223 [cited 2022 Oct 6]Available from: https://www.cell.com/article/S0966842X20302808/fulltext 33234439
Fokam J Alteri C Colagrossi L Genevieve A-M Siré Takou D Ndjolo A Diagnostic performance of molecular and serological tests of SARS-CoV-2 on well-characterised specimens from COVID-19 individuals: The EDCTP ‘PERFECT-study’ protocol (RIA2020EF-3000) PLoS One [Internet] 17 9 2022 e0273818 [cited 2022 Oct 6]Available from https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0273818
Sidiq Z Hanif M Dwivedi KK Chopra KK. Benefits and limitations of serological assays in COVID-19 infection Indian J Tuberc [Internet] 67 4 2020 Dec 1 S163 [cited 2022 Oct 6]Available from: /pmc/articles/PMC7409828/ 33308664
Lu L Zhang H Zhan M Jiang J Yin H Dauphars DJ Antibody response and therapy in COVID-19 patients: what can be learned for vaccine development? Sci China Life Sci 2020 6312 [Internet] 63 12 2020 Dec 1 1833 1849 [cited 2022 Oct 6]Available from: https://link.springer.com/article/10.1007/s11427-020-1859-y
Li Y yun Lai D Lei Q wei Xu Z Wang F Hou H Systematic evaluation of IgG responses to SARS-CoV-2 spike protein-derived peptides for monitoring COVID-19 patients Cell Mol Immunol [Internet] 18 3 2021 Mar 1 621 631 [cited 2022 Oct 6]Available from: https://pubmed.ncbi.nlm.nih.gov/33483707/ 33483707
Li Y, Lai D yun, Lei Q, Xu Z wei, Wang F, Hou H, et al. Systematic evaluation of IgG responses to SARS-CoV-2 spike protein-derived peptides for monitoring COVID-19 patients. 2021 [cited 2022 Jun 11];18(3). Available from: https://pubmed.ncbi.nlm.nih.gov/33483707/
Li Y yun Lai D Lei Q wei Xu Z Wang F Hou H Systematic evaluation of IgG responses to SARS-CoV-2 spike protein-derived peptides for monitoring COVID-19 patients Cell Mol Immunol [Internet] 18 3 2021 Mar 1 621 631 [cited 2022 Oct 31]Available from: https://pubmed.ncbi.nlm.nih.gov/33483707/ 33483707
EUA Authorized Serology Test Performance | FDA [Internet]. [cited 2022 Jun 23]. Available from: https://www.fda.gov/medical-devices/coronavirus-disease-2019-covid-19-emergency-use-authorizations-medical-devices/eua-authorized-serology-test-performance
List C Qi W Maag E Gottstein B Müller N Felger I. Serodiagnosis of Echinococcus spp. infection: Explorative selection of diagnostic antigens by peptide microarray PLoS Negl Trop Dis [Internet] 4 8 2010 Aug 3 e771 [cited 2020 Feb 26]Available from: https://pubmed.ncbi.nlm.nih.gov/20689813/ 20689813
Vengesai A, Naicker T, Kasambala M, Midzi H, Mduluza-Jokonya T, Rusakaniko S, et al. Clinical utility of peptide microarrays in the serodiagnosis of neglected tropical diseases in sub-Saharan Africa: protocol for a diagnostic test accuracy systematic review. 2021 Jul [cited 2022 Jan 9];11(7):e042279. Available from: https://pubmed.ncbi.nlm.nih.gov/34330850/
Qi H Ma M-L Jiang H-W Ling J-Y Chen L-Y Zhang H-N Systematic profiling of SARS-CoV-2 specific IgG epitopes at single amino acid resolution medRxiv [Internet] 2020 Sep 9 10.1101/2020.09.08.20190496 [cited 2020 Dec 11]2020.09.08.20190496Available from:
Vengesai A Kasambala M Mutandadzi H Mduluza-Jokonyaid TL Mduluzaid T Naicker T Scoping review of the applications of peptide microarrays on the fight against human infections PLoS One [Internet] 17 1 2022 Jan e0248666 [cited 2022 Feb 13]Available from: https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0248666
Vengesai A Kasambala M Mutandadzi H Mduluza-5 Jokonya TL Mduluza T Naicker T. Scoping review of the applications of peptide microarrays on the fight against human Short title: Peptide microarrays application 4 bioRxiv [Internet] 2021 Mar 4 10.1101/2021.03.04.433859 [cited 2021 Jun 14]2021.03.04.433859Available from:
Van Regenmortel MHV. Structural and functional approaches to the study of protein antigenicity [Internet] 10 1989 266 272 Immunology Today. Immunol Today[cited 2021 Jun 14]Available from: https://pubmed.ncbi.nlm.nih.gov/2478146/
Saha S, Raghava GPSS. Prediction of continuous B-cell epitopes in an antigen using recurrent neural network. 2006 [cited 2021 Feb 9];65(1):40–8. Available from: https://doi.wiley.com/10.1002/prot.21078
Giacò L Amicosante M Fraziano M Gherardini PF Ausiello G Helmer-Citterich M B-Pred, a structure based B-cell epitopes prediction server Adv Appl Bioinforma Chem [Internet] 5 1 2012 Jul 25 11 21 10.2147/AABC.S30620 [cited 2021 Jun 14]Available from:
Jena M, Kumar V, Kancharla S, Kolli P. Reverse vaccinology approach towards the in-silico multiepitope vaccine development against SARS-CoV-2. F1000Research 2021 1044 [Internet]. 2021 Jan 23 [cited 2022 Feb 18];10:44. Available from: https://f1000research.com/articles/10-44
Singh A, Thakur M, Sharma LK, Chandra K. Designing a multi-epitope peptide based vaccine against SARS-CoV-2. 2020 [cited 2022 Feb 18];10(1):1–12. Available from: https://www.nature.com/articles/s41598-020-73371-y
Rusakaniko S Sibanda EN Mduluza T Tagwireyi P Dhlamini Z Ndhlovu CE SARS-CoV-2 Serological testing in frontline health workers in Zimbabwe PLoS Negl Trop Dis [Internet] 15 3 2021 Mar 1 e0009254 10.1371/journal.pntd.0009254 [cited 2021 Jun 14]Available from:
SD BIOSENSOR | PRODUCTS [Internet]. [cited 2022 Oct 6]. Available from: https://www.sdbiosensor.com/product/product_view?product_no=239
Schwarz T Heiss K Mahendran Y Casilag F Kurth F Sander LE SARS-CoV-2 Proteome-Wide Analysis Revealed Significant Epitope Signatures in COVID-19 Patients Front Immunol [Internet]. 12 2021 Mar 23 765 [cited 2021 May 24]Available from: www.frontiersin.org
Deeks JJ Dinnes J Takwoingi Y Davenport C Spijker R Taylor-Phillips S Antibody tests for identification of current and past infection with SARS-CoV-2 [Internet] Cochrane Database of Systematic Reviews 2020 2020 [cited 2021 Mar 19]Available from: https://www.cochranelibrary.com/cdsr/doi/10.1002/14651858.CD013652/full
Farrera-Soler L Daguer JP Barluenga S Vadas O Cohen P Pagano S Identification of immunodominant linear epitopes from SARS-CoV-2 patient plasma PLoS One [Internet] 15 9 2020 Sep 1 e0238089 10.1371/journal.pone.0238089 [cited 2021 Jun 14]septemberAvailable from:
Holenya P Lange PJ Reimer U Woltersdorf W Panterodt T Glas M Peptide microarray-based analysis of antibody responses to SARS-CoV-2 identifies unique epitopes with potential for diagnostic test development Eur J Immunol [Internet] 51 7 2021 Jul 1 1839 1849 [cited 2022 Jun 24]Available from: https://onlinelibrary.wiley.com/doi/full/10.1002/eji.202049101 33772767
Satarker S Nampoothiri M. Structural Proteins in Severe Acute Respiratory Syndrome Coronavirus-2 Archives of Medical Research. Elsevier Inc. 51 2020 482 491
Holenya P Lange PJ Reimer U Woltersdorf W Pan-Terodt T Glas M Peptide microarray based analysis of antibody responses to SARS-CoV-2 identifies unique epitopes with potential for diagnostic test development medRxiv [Internet] 2020 Nov 27 10.1101/2020.11.24.20216663 [cited 2020 Dec 11]2020.11.24.20216663Available from:
Poh CM Carissimo G Wang B Amrun SN Lee CYP Chee RSL Two linear epitopes on the SARS-CoV-2 spike protein that elicit neutralising antibodies in COVID-19 patients Nat Commun [Internet] 11 1 2020 Dec 1 1 7 10.1038/s41467-020-16638-2 [cited 2021 Jun 14]Available from: 31911652
Wang H Wu X Zhang X Hou X Liang T Wang D SARS-CoV-2 Proteome Microarray for Mapping COVID-19 Antibody Interactions at Amino Acid Resolution ACS Cent Sci [Internet]. 6 12 2020 Dec 23 2238 2249 10.1021/acscentsci.0c00742 [cited 2021 Jun 14]Available from: 33372199
Lopandić Z Protić-Rosić I Todorović A Glamočlija S Gnjatović M Ćujic D Igm and igg immunoreactivity of sars-cov-2 recombinant m protein Int J Mol Sci [Internet] 22 9 2021 May 1 4951 10.3390/ijms22094951 [cited 2021 Jun 14]Available from: 34066920
Burbelo PD Riedo FX Morishima C Rawlings S Smith D Das S Detection of Nucleocapsid Antibody to SARS-CoV-2 is More Sensitive than Antibody to Spike Protein in COVID-19 Patients medRxiv 2020 Apr 24 2020.04.20.20071423
La Marca A Capuzzo M Paglia T Roli L Trenti T Nelson SM Testing for SARS-CoV-2 (COVID-19): a systematic review and clinical guide to molecular and serological in-vitro diagnostic assays Reproductive BioMedicine Online 41 2020 483 499 32651106
Arya R Kumari S Pandey B Mistry H Bihani SC Das A Structural insights into SARS-CoV-2 proteins Journal of Molecular Biology 433 2021 166725
Ceraolo C Giorgi FM. Genomic variance of the 2019-nCoV coronavirus J Med Virol [Internet] 92 5 2020 May 1 522 528 [cited 2021 Jun 14]Available from: https://onlinelibrary.wiley.com/doi/full/10.1002/jmv.25700 32027036
Carmona SJ Sartor PA Leguizamón MS Campetella OE Agüero F. Diagnostic Peptide Discovery: Prioritization of Pathogen Diagnostic Markers Using Multiple Features PLoS One [Internet] 7 12 2012 Dec 14 50748 [cited 2020 Nov 22]Available from: www.plosone.org
| 36460093 | PMC9705268 | NO-CC CODE | 2022-12-10 23:14:57 | no | Acta Trop. 2023 Feb 29; 238:106781 | utf-8 | Acta Trop | 2,022 | 10.1016/j.actatropica.2022.106781 | oa_other |
==== Front
Am Behav Sci
Am Behav Sci
ABS
spabs
The American Behavioral Scientist
0002-7642
1552-3381
SAGE Publications Sage CA: Los Angeles, CA
10.1177/00027642221138279
10.1177_00027642221138279
Article
Are People Hesitating—Or Just Postponing—to Get the Covid-19 Vaccine? Vaccine Outreach in Marginalized Urban Communities
Kogen Lauren 1
Cai Deborai A. 1
Pitts Cornelius 1
Imms Patricia 2
Perkins Mitch 1
Reeves Kathleen 1
1 Temple University, Philadelphia, PA, USA
2 Miriam Medical Clinics, Philadelphia, PA, USA
Lauren Kogen, Klein College of Media and Communication, Temple University, 2020 N. 13th St., Philadelphia, PA 19122, USA. Email: [email protected]
25 11 2022
25 11 2022
00027642221138279© 2022 SAGE Publications
2022
SAGE Publications
This article is made available via the PMC Open Access Subset for unrestricted re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the COVID-19 pandemic or until permissions are revoked in writing. Upon expiration of these permissions, PMC is granted a perpetual license to make this article available via PMC and Europe PMC, consistent with existing copyright protections.
Across Philadelphia, approximately 80% of adults are fully vaccinated against Covid-19. However, many zip codes in the city remain far below the city-wide vaccination rate. These zip codes correspond to marginalized sections of the city and to neighborhoods with a high proportion of residents of color and high levels of poverty. In-depth interviews were conducted with representatives from 15 community-based organizations (CBOs) that serve such communities in the city to (1) learn why people are not yet vaccinated and (2) evaluate methods for encouraging vaccination. A qualitative thematic analysis of interview transcripts was conducted to evaluate why people are not getting vaccinated. Together, the findings suggest that distrust toward the vaccine, the government, and the healthcare system, combined with a host of matters considered by residents to be more urgent—such as missing work, cost concerns, and concerns around presenting identification—result in what might be better described as vaccine postponement rather than vaccine refusal. For many, vaccination is simply not a priority. The findings from this analysis illuminate some of the lesser discussed reasons for vaccination delay and provide insights into how to promote vaccinations both for the current Covid pandemic and for future vaccination efforts.
Covid-19 pandemic
Covid-19 vaccine
vaccine hesitancy
vaccine refusal
community involvement
urban health services
Federal Emergency Management Agency (FEMA) City of Philadelphia’s Department of Public Health (PDPH) edited-statecorrected-proof
typesetterts1
==== Body
pmcThe global Covid-19 pandemic has already claimed over six million lives globally (BBC News, 2022) and over one million lives in the United States (Centers for Disease Control, 2022). Vaccinations have been shown to be highly effective in curbing the spread of the disease, including variants, reducing the likelihood of severe disease effects, hospitalization, and death (Centers for Disease Control, 2021; Iuliano et al., 2022).
However, despite the overall success of vaccines in reducing severe complications from Covid-19, many people living in the United States remain hesitant to get vaccinated (Laughlin & Lai, 2021). The reasons for hesitation are varied and complex; they include skepticism due to historical injustices within the American healthcare system (Lozano, 2020), distrust of current political institutions (Lerer, 2021), as well as a variety of beliefs regarding the speed in which the vaccine was developed, the efficacy of the vaccine, the risk of side effects, and the perceived threat of Covid-19 itself.
Vaccination rates in the city of Philadelphia have been greatly affected by vaccine hesitancy. Many zip codes in the city are far below the city-wide vaccination rate of 74% for those aged 12 and older (City of Philadelphia, 2022). These zip codes correspond to marginalized sections of the city and to neighborhoods with a high proportion of residents of color and high levels of poverty. People living within many zip codes still have vaccination rates under 60% and more than half of the residents in many of these regions live in poverty (compared to the city-wide poverty level of 23%) (U.S. Census Bureau, 2019).
To support its effort to increase vaccination rates in these neighborhoods, The Lewis Katz School of Medicine at Temple University received a grant from the City of Philadelphia’s Department of Public Health (PDPH) with Federal Emergency Management Agency (FEMA) funds from the U.S. government beginning April 2021. The School of Medicine later recruited faculty members from the Klein College of Media and Communication at Temple University to help with outreach efforts and to design tailored campaigns to reach diverse communities across Philadelphia with vaccination clinics.
This study will describe the results from in-depth interviews conducted among North Philadelphia community partners for two purposes: understanding hesitancy and developing plans for vaccination campaigns. These results are compelling because, although data collection relates to particular neighborhoods within Philadelphia, themes derived from this research suggest that the lessons learned are useful for developing strategies elsewhere that recognize and address societal issues fundamental to present and future vaccination hesitancy.
Overall, recommendations resulting from these community interviews focus on vaccine postponement, safety concerns, and a lack of trust. First, although current media narratives around vaccine hesitancy focus on “anti-vaxxers” who refuse vaccination, our findings suggest that many people are not completely opposed to vaccination but rather delay vaccination for several reasons. In marginalized communities, higher priority may be placed on other daily challenges people face. Second, although narratives around vaccine hesitancy emphasize a general distrust of government, our findings add nuance to this generalization, including an appreciation for the legitimate fear of government expressed by many.
Background and Theoretical Framework
Vaccine hesitancy has been a popular topic in the U.S. news media since the Covid-19 vaccines were first approved. However, the term vaccine hesitancy has evolved to have two quite distinct meanings. This term has, at times, referred specifically to concerns around vaccine safety, such as how quickly the vaccines were developed, or risks of potential side effects, or beliefs that the dangers of the Covid-19 virus are exaggerated (Karni & Kanno-Youngs, 2021).
One reason for vaccine hesitancy is associated with vaccine safety: This aspect of vaccine hesitancy is a general distrust of the medical establishment. The most often cited historical episode linked with this distrust is the Tuskegee Experiment (Aslan & Wanamaker, 2016). This event and others have had a major negative impact on the relationship communities of color have with healthcare and governmental systems.
A second understanding of vaccine hesitancy is rooted in material concerns. This aspect has received much less attention in the press: vaccination delay may result from lack of access to vaccine clinics, concerns about missing work, and concerns about having to provide identification (Feldman, 2021). These barriers have little relation to vaccine safety but are consequential to the marginalization of communities without adequate resources.
These two narratives have been used to explain low vaccination rates in high-poverty regions or neighborhoods with high proportions of residents of color. In Philadelphia—the location of the current project—the five zip codes with the lowest vaccination rates1 (Laughlin & Lai, 2021) have high poverty (four of the five zip codes have median household incomes between $30,633 and $37,345) and a high proportion of Black residents (80–91%) (U.S. Census Bureau, 2019). Within these types of neighborhoods, the distrust narrative has been widely used to explain why residents might be reluctant to get the vaccine (Laughlin & Lai, 2021), especially referencing historical mistreatment within the U.S. healthcare system.
To be sure, distrust regarding the safety of the vaccine and the healthcare institutions that administer them exist. But other types of barriers have received far less attention. Tom Frieden (2021), former director of the U.S. Centers for Disease Control, posited that “most of the people who are not yet vaccinated aren’t strongly opposed to being vaccinated, they just haven’t had the vaccine be as convenient as it should be.” Frieden argued that the focus on vaccine hesitancy, defined as distrust often fueled by misinformation, has been featured prominently in the news media because of how politicized the vaccine has become in the United States.
These two understandings of vaccine hesitancy—one regarding vaccine safety and one regarding material barriers to vaccination–align with two aspects of Fishbein and Cappella’s (2006) integrative model of behavior, and each suggests different pathways toward vaccination decision making. Fishbein and Cappella’s model posits that attitudes and beliefs regarding a health behavior (e.g., unfavorable attitudes toward the vaccine) influence an individual’s decision about whether to carry out the behavior, but that even if someone has decided to take on the behavior, environmental factors, or other barriers (e.g., lack of proper identification at a vaccination site) may prevent or delay action (p. S2).
These two pathways that lead away from getting vaccinated—one primarily influenced by unfavorable attitudes toward the vaccine and one primarily influenced by environmental barriers that make vaccination difficult—have received uneven attention in both the media and in Covid-19 research to date. Current research provides abundant evidence that the first of these pathways, in particular, beliefs about vaccine safety and efficacy, beliefs about vulnerability, and attitudes toward health and government institutions, is indeed a valid explanation for why some remain unvaccinated (Bass et al., 2021; Latkin et al., 2021; Rane et al., 2022). Less research has been devoted to the second pathway, although some studies have briefly mentioned the need for more convenient vaccination locations and timings (Forman et al., 2021; Peters, 2022; Rutten et al., 2021). Our project seeks to elucidate how these pathways have played out in the North Philadelphia context.
Method
The vaccination effort described here began with investigators from the School of Medicine who are researchers and healthcare professionals. These researchers conducted vaccination clinics at the locations of several community-based organizations (CBOs) that work with populations observed to be under-vaccinated at the start of the project. These CBOs included churches, advocacy organizations, and schools.
Although the program, which started in February 2021, had an early period of success vaccinating hundreds of residents per week, by weeks 10 and 11 (June 2021) vaccinations had declined to fewer than 30 per week. At this time, faculty researchers from the College of Media and Communication were asked to assist in the effort to improve outreach to the communities that the organizations served. Figure 1 displays the vaccination trend across these locations over the course of this study.
Figure 1. Weekly vaccination rates across community-based organizations in Philadelphia.
Between June 2021 and January 2022, [Communication School] researchers conducted in-depth key informant interviews with 15 of the CBOs brought onto the project. All interviewees except two were members of African American, Caribbean, or Latinx communities. Demographic data, such as gender and age of the interviewees, were not collected. Interviewees were leaders within the CBOs, worked directly with community members, and were experts in their community’s needs. Thus, these individuals were able to serve as primary sources of information about the community’s beliefs and barriers to vaccination. All the organizations serve populations within the targeted low vaccination zip codes (see Table 1).
Table 1. Community-Based Organizations Interviewed.
Organization Type of organization Primary populations served (self-described)
CBO1 (Bryant Baptist Church) Baptist Church African American
CBO2 (Second Antioch Baptist Church) Baptist Church African American
CBO3 (Bright Hope Baptist Church) Baptist Church African American
CBO4a Baptist Church African American
CBO5a Supportive services Hispanic
CBO6a Supportive services Latino
CBO7 (Caribbean Community in Philadelphia) Cultural Caribbean
CBO8 (New Courtland) Senior services African American
CBO9a Health care African American
CBO10a Health care African American
CBO11a Health care Diverse
CBO12 (AHARI) Low-income housing African American
CBO13 (Mary McLeod Bethune Elementary School) Public school Diverse
CBO14a Public school Diverse
CBO15a Public school Diverse
Note. CBO = community-based organization.
a Organization that did not grant permission to use its name.
The goal of these interviews was to understand why these communities continued to have exceedingly low vaccination rates. We made no assumptions about what was causing low vaccination rates and approached the interviews under the belief that no two communities were exactly alike. The main purpose of the interviews was to understand what the interviewees thought were the reasons for low vaccination rates within their community and what strategies they thought may be effective in their community to promote vaccinations.2 The interviews with the community partner interviewees lasted between 30 and 60 minutes and were conducted via Zoom.
Data Analysis
A qualitative thematic analysis (Braun & Clarke, 2006; Joffe, 2012) of the interview transcripts was conducted. Interviews were transcribed and coded for themes using ATLAS.ti. Since we did not enter the project with a priori expectations about attitudes toward vaccination, the transcripts were coded inductively; in other words, we coded comments related to vaccinations as they emerged rather than comparing transcripts against a preexisting hypothesis or coding scheme. The interviews were also coded iteratively: They were revised and recoded until a consistent set of themes regarding reasons for not getting vaccinated and pathways toward vaccination emerged (Bradley et al., 2007).
Qualitative and inductive coding processes are useful for phenomena in which the researchers seek to understand how or why complex phenomena are occurring without a priori assumption about what is driving human behavior (Braun & Clarke, 2006). This type of data analysis has been shown to be useful in the social sciences, in general, and in the field of health in particular, for understanding complex health-related behaviors (e.g., Joffe and Haarhoff’s 2002 examination of audience understanding regarding Ebola in the 1990s).
Results
Themes that emerged from the analysis fall within two broad categories: (1) reasons for delaying vaccination and (2) recommendations for encouraging vaccination. Together, the findings suggest that attitudes toward the vaccine, the government, and the healthcare system, combined with a host of more urgent matters, including environmental barriers, result in what might be better described as vaccine postponement rather than vaccine refusal or hesitancy. According to the representatives we interviewed, many people across the communities are delaying getting vaccinated because they are wary of the vaccine, vaccination is simply not a priority, or both (see Table 2).
Table 2. Themes and Sub-Themes From Interviews.
Themes and sub-themes # of interviews supporting theme
Themes related to delaying vaccination
Theme 1: Vaccination is not a priority 10
Theme 2: Distrust, anger, and resentment 13
Distrust due to fear and lack of relevant health information 10
Themes related to promoting vaccinations
Theme 3: Build trust 15
Treat people as intelligent and rational 12
Make clinics warm, welcoming, and family friendly 8
Theme 4: Make vaccinations more convenient 13
Provide convenient times and locations 12
Reduce language barriers 6
Theme 5: Hyper-local communication outreach with trusted sources 11
Theme 6: Provide appropriate incentives 13
Themes of Postponement
One of the aims of this analysis was to shed light on competing media narratives regarding whether the main impediment to vaccinations is distrust (attitudes) or inconvenience (environmental factors). Our analysis suggests that both barriers were present, but there were not nearly as many vaccine refusers as media narratives suggest.
Instead, many people across the communities were delaying getting vaccinated because it was one of several competing life obstacles that ranked low on the list of priorities. Compared to community safety, personal health, work, and family commitments, low urgency was applied to being vaccinated. Interviewees mostly described this population of “postponers” as being in their 20s, 30s, and 40s, and as relatively healthy people who believed that should they contract the virus, they would probably not have severe illness.
Theme 1: Vaccination is Not a Priority
Two examples of commentary on low priority follow:CBO7: People are really consumed by this constant daily struggle. And you see a lot of things past Covid where we had to deal with it every day. Covid is just an added layer, but it’s nothing different than what our communities are used to seeing. . . There’s already a lot of death in our communities. Philadelphia’s homicide rate is extremely high. When you think about things like that, the last thing on that list will be Covid. Because especially at our age, we think, “Okay, we can fight it. I can’t fight a bullet.” So that’s [the types of] conversations I’m having when I talk to people my age.
CBO6: [The last time we went door-to-door] I had one person who was shut down, not getting it, but everyone else who I spoke to who hadn’t gotten it was not opposed, they just hadn’t found a time yet to either take themselves or to take someone else. . . I think the actual going and getting it, the scheduling of it, when you have a really busy lifestyle, and you’re working. . . you’re picking up your kid from childcare, you’re a mother making dinner then, and it’s just difficult to then go out of the way and get one. . . You could potentially be sick for 24 hours. And I think for some people, it’s just a deterrent. . .
The theme of low priority was expressed in nine interviews. Employment and childcare responsibilities were the most common priorities cited by residents as reasons being vaccinated was postponed. The risk of side effects, or having to take time off from work, were presented as major factors to avoid vaccination.
Many residents also believed they had to pay for the vaccination or show proof of insurance. Showing identification at the clinics was also interpreted as a risk or potential cost to some residents, particularly those who were undocumented.
Theme 2: Distrust, Anger, and Resentment
Distrust, anger, and resentment toward a variety of U.S. institutions were frequently expressed views in a large portion of the interviews. Some of this distrust seemed to dissolve into extreme skepticism of the government’s intentions, fear, and a desire for reassurance of safety of the vaccine. Our interviewees convincingly emphasized the legitimacy of feelings of distrust and anger, given the well documented historical governmental mistreatment of minority citizens.
CBO6: The presentation of the vaccine as this thing that’s going to help communities come out of this really challenging year isn’t always [going to be trusted]. Because I think for many, they already felt they were in this on their own and handling it. . . Because the community has been there for them in ways that I think they don’t necessarily see the Philadelphia government being there for them.
CBO12: You got to understand when you come to an underserved community. . . There’s trust issues because they have been slighted on a lot of different things. Literally bamboozled. They do not trust.
CBO13: [There’s a] skeptical mindset about vaccinations and government. Most of the parents I’ve spoken to about vaccines. . . they’re just really wary about the fact that it’s being pushed so heavily in our community.
While historical injustices such as the Tuskegee Experiment and birth control testing on Puerto Rican women came up frequently, there was also a sense that the attitude of elite institutions toward marginalized populations—that they are expendable—still exists. Residents do not feel the government has their best interests at heart. This skepticism about the government’s intentions came up in seven of the interviews. In many cases this distrust led to outright fear of government institutions, to be discussed below.
Distrust is due to fear and lack of relevant health information
Ten CBOs cited misinformation as a problem. Many are considering getting the vaccine, but due to distrust and pervasive misinformation they do not know who to believe and are genuinely scared:CBO7: They have a healthy distrust, first of all, of government, of Western medicine. And they have their reasons, and that creates a good sense of fear. . . A lot of individuals I talk to aren’t totally opposed, they’re more just scared that we’re going to hurt them than anything else. And I’ve had people say, like, “Well, will my children be okay?” Because she wants to be sure that she can live to raise them, and she’s heard that the vaccine might hurt her. And those conversations usually happen from people. . . who are often victimized by Western medicine, and the abuses of the Western world. And that coming into your ear is very scary.
CBO1: I’ve had to go to the clinic with them to get the vaccination and hold their hands. . . This young lady just called me yesterday because she had to go for her second shot. And she’s like, “Well, can you meet me there? I’m really still afraid.” It’s not that personable when you go to the pharmacies. Most people want somebody that they know is going to be there for them if they get sick, that they can call and say, “Hey, I’m having this reaction. Is this normal?” You know?. . . Because people have a hard time. . . trusting the people that are administering it.
Many of the residents in these communities genuinely want to understand better and educate themselves. But this is difficult given the complex (and often conflicting) information coming from the government. Three CBOs also mentioned language barriers as a reason residents found it difficult to access the vaccine information they wanted.
Recommendations for Encouraging Vaccination
The findings from interviews suggest that, although efforts to address misinformation are important, they should not be the primary aim of vaccination campaigns. Providing data on the safety of vaccines and on the risks of Covid-19 may be helpful, but this approach misses some key reasons people are not getting vaccinated.
Theme 3: Build Trust
As noted in the previous section, distrust is pervasive within the communities we targeted. This section addresses ways CBOs felt levels of trust, therefore willingness to get vaccinated, could be promoted.
Treat people like intelligent human beings capable of making their own health decisions
This theme arose in 12 of the interviews. Residents in these communities were resentful of “we-know-better-than-you” attitudes. They care about their health and often need someone who can provide information without judgment. Communication should focus on answering people’s questions honestly so that they can make informed decisions for themselves, not on trying to convince them to get vaccinated. This was best summed up in the following interview excerpt:CBO13: There needs to be education, but not from a condescending place. . . They’ll listen as long as you’re respectful. Treat them as the humans they are: intelligent beings capable of making sound choices. That’s been the pushback really. . . [We] don’t like it when people talk down to us.
Along the same lines, interviewees felt more of a focus ought to be placed on helping individuals in underserved communities with their health needs no matter what they are, not only vaccinating them against Covid. For many marginalized communities Covid is one among a variety of everyday hurdles, and not always the top priority.
Make clinics warm, welcoming, and family friendly
Eight of the representatives we spoke with mentioned that vaccination clinics needed to be fun, warm, and welcoming to build comfort and trust levels. They suggested offering vaccination clinics at family-oriented community events, including music, balloons, child-friendly activities or treats, science demonstrations, and magic shows.
Vaccination clinics can be seen as an opportunity to have questions answered about the vaccine (even if they do not choose to get it that day), and the potential to dispel distrust in a healthcare system they don’t believe cares about them.
Theme 4: Make Vaccinations More Convenient
Provide convenient times and locations
For people who are not prioritizing vaccination, one persuasion strategy is to make getting vaccinated as easy as possible by bringing the vaccines to them when it is most convenient. People are more likely to get vaccinated when the clinic is within a few blocks of their home or when they are already somewhere for another reason and can be persuaded to get the vaccine while they are there.
As noted in the previous section, many remained unvaccinated because they did not want to miss work. It is therefore also important to have clinics at various times to accommodate work schedules. Saturdays seem to be particularly useful in that (1) clinics can be attached to family-friendly events that are already happening and (2) they give residents (who work traditional 9–5 schedules) an extra weekend day to recover from potential side effects, thus minimizing the possibility of missing work.
Additionally, CBOs who serve large English as a Second Language populations stated that an ideal clinic would have (1) vaccine information available in multiple languages; (2) bilingual vaccinators (ideally from the same region of origin as patients), and (3) translators who are not vaccinating and can devote their time to helping residents communicate with staff and helping residents feel comfortable and welcome.
Finally, outreach should make clear that vaccines are free, require no identification, and require no insurance.
Tag onto other events where people already are
These were straightforward and easy-to-implement suggestions from CBOs. These “events” may include mobile clinics in residential neighborhoods; organized community events such as fairs or festivals; or more regular weekend activities such as Saturday morning trips to the grocery store. As one representative stated:CBO10: I think, like, having other resources at this event [is helpful.] People come in, they get warm, they feel the energy and then they might go over, ask questions, go back to the [other tables] and think about it. So I think [if it’s] just a Covid vaccine [clinic], if they’re not vaccinated, they don’t want to come. . . But they could go to the event, get something to eat, think about it, call somebody, come back, ask more questions. . . It’s more of a back and forth as they’re making their mind up.
The overall sentiment was that clinics needed to be arranged in such a way that people could attend without the explicit purpose of getting vaccinated, could think about it, ask questions, and could decide on the spot whether to receive the vaccine. Interviewees felt that this would allow someone on the fence to feel that vaccination would likely never be more convenient than at that moment, and so this was the time to do it.
Theme 5: Hyper-Local Communication Outreach with Trusted Sources
For marginalized and under-vaccinated communities, clinics ought to be hyperlocal so that residents do not need to travel outside their community. This recommendation means communication outreach needs to be hyperlocal as well. For the CBOs we spoke with, this approach includes phone calls, word of mouth, posters and flyers around the community, and door-to-door knocking—sometimes during an event and sometimes prior. CBOs strongly feel that this type of direct outreach is what is needed.
CBO5: [And even when we did it locally] we realized that people didn’t know. No matter how many times we posted on social media, no matter how many times radio and TV mentioned it, they didn’t know that the clinic was so close to their homes. And here. . . there’s a huge internet disparity. . . and we decided the only way they’re going to know that we’re actually here, the only way they’re going to be able to take advantage, is if we actually go to their doorstep, and knock, and talk to them, and do flyers.
Theme 6: Provide Appropriate Incentives
There were mixed feelings among the interviewees regarding the value of, and ethics behind, providing cash incentives for vaccinations. Overall, the findings indicate that incentives can be beneficial if they do not feel like bribes. Large incentives, particularly cash, make residents trust the vaccine less, feel that they are being used, or insulted (if the implication is that they will go against their beliefs just for a prize).
CBO10: I think right now people are like, “If I get it, I get it,” not. . . “if you offer me $25 for a shot I’m going to get the shot.” . . . I don’t think it will draw people to get it, but it’s like, “We appreciate you for coming to this event.” So it shows some type of appreciation, but I don’t think it makes somebody’s mind up.
But respondents were generally supportive of small incentives that felt more like a celebration (e.g., food, snacks, and balloons for kids) or that addressed local needs (e.g., a grocery store gift card) or giveaways that draw people to an event (and which attendees would receive whether they were vaccinated).
CBO6: It can feel a little like. . . “Really? You think that just for a popsicle I would come out and get vaccinated?” So, I think if it’s more just part of our celebration of you getting vaccinated. . . like, it’s something nice. . . in exchange for whatever terror you went through leading up to it, to memorialize the fact that you did something courageous and scary. . . I think it was nice to have. . . to make the experience really positive, which I do think is important, because. . . that means that we have more people then who are more likely to talk to their family members and other people and say, “I had a positive experience, it was fine. You should also get vaxxed.”
In sum, our interviews suggest that by focusing on building trust; making vaccinations more convenient; engaging in hyperlocal, direct communication outreach; and providing appropriate incentives, those who are on the fence or postponing their vaccination might be more likely to “get the jab.”
Discussion
As discussed above, Fishbein and Cappella’s (2006) integrative model of behavior posits that attitudes are an important influence on behavior decisions, but environmental factors ought not to be ignored. Indeed, we found that unfavorable attitudes toward vaccination, particularly with respect to distrust, anger, and resentment toward medical institutions, with respect to lack of information about vaccine safety, influenced decisions to refuse or postpone Covid-19 vaccinations. This conclusion is in line with current research (e.g., Bass et al., 2021; Latkin et al., 2021; Rane et al., 2022) and with the dominant media discourse surrounding vaccine hesitancy.
However, this explanation is insufficient to fully understand vaccination postponement. Interviews revealed that daily challenges such as community violence, employment and childcare demands, and concerns around cost and identification requirements, are among some of the environmental factors that may influence individuals to postpone vaccination, especially for those who are already concerned about vaccine safety or do not believe they are at high risk for Covid complications. In other words, for many individuals, vaccination is simply not high enough of a priority to overcome either unfavorable beliefs or environmental barriers.
Limitations and Directions for Future Research
The rapidly developing case of the pandemic did not allow us to speak to a large number of unvaccinated community members to ask them their reasons for delay. Future research should ask community members themselves of what environmental barriers, if any, result in vaccination delay. This research is crucial to be prepared for the next pandemic.
It is also the case that, although many barriers presented here (e.g., time, childcare, and cost constraints) seem applicable to residents living in other low-income communities, future research should seek to confirm these findings. Likewise, some of the reasons we found (e.g., distrust and resentment toward government) may not be as salient for other underserved groups.
Conclusion
Our findings suggest vaccination campaigns must do more to address both unfavorable attitudes and environmental barriers that make vaccinations a low priority for some. Public health entities must acknowledge competing needs people from marginalized neighborhoods experience during a pandemic. Even as trusted healthcare providers establish easily accessible and compassionate venues for vaccinations to be administered, issues of violence, food and housing insecurity, and economic burdens prevail. These are important pieces of a complex puzzle that may constitute more of a threat to one’s well-being than a pandemic. These interviews often express elements of sustained trauma, exclusion, and community neglect in which residents, by necessity, must be attentive to immediate priorities of survival. Vaccination strategies must therefore recognize this communal attitude and work with residents to build community and personal safety.
Often, in situations of vaccination (and general health care) postponement, healthcare organizations explore ways to enhance persuasion toward immediate acceptance. Yet these anecdotes point to deeply held priorities and convictions that one could argue and reflect on basic societal injustices. Thus, vaccination delay could be characterized as symptomatic of these injustices. If we want to increase vaccination acceptance, a close examination of societal inequity is required.
Lee and Kotler (2008) have argued that, when it comes to persuading people to take on public health behaviors, early and late adopters of innovations (Rogers, 1983) may be better described as the “show me,” “help me,” and “make me” groups. Those in the “show me” group are, with minimal direction, willing take on requested public health behaviors for the public. Those in the “help me” group are willing to comply but need help, such as extra assistance, instruction, or incentives. Those in the “make me” group will only comply when the law requires it.
Our study reflects this perspective and provides evidence that more resources must focus on helping this group, where there is room for movement. People who feel overwhelmed and overburdened by daily circumstances may need help to incorporate vaccination into their lives. Theories of persuasion for altering healthcare behaviors, coupled with a recognition of, and sensitivity toward the factors that create community marginalization, need to successfully address the challenges posed by Covid-19 and other vaccination campaigns.
The authors would like to sincerely thank the people at the community-based organizations that spoke with us as part of this project. Without their generosity and honesty, this research would have not been possible.
Author Biographies
Lauren Kogen is an Assistant Professor in the Department of Media Studies and Production at the Klein College of Media and Communication at Temple University. Her research focuses on communication for social change (how media and communication can promote positive social change for marginalized or underserved communities) and communication about social change (how information about social change work is communicated to policy makers and the public).
Deboraih A. Cai is professor and senior associate dean in the Klein College of Media and Communication at Temple University, and a faculty member in the Media and Communication doctoral program. Deborah is an international researcher with expertise in intercultural communication, persuasion, negotiation and conflict management. Deborah is a Fellow in the International Academy of Intercultural Research and a Fellow in the International Association for Conflict Management.
Cornelius Pitts is Director of the COVID-19 Vaccination Project at the Temple University, Lewis Katz School of Medicine and a founding director of Miriam Medical Clinics. He is an Assistant Professor at the Katz School has taught Global Health at the Philadelphia College of Pharmacy. His research centers on methods for increasing healthcare access for underserved populations
Patricia Imms is the Nursing Director at Miriam Medical Clinics in Philadelphia, and coordinated the Covid-19 Vaccination Project at the Lewis Katz School of Medicine at Temple University. She is a psychiatric nurse and has served as Clinical Research Coordinator at the University of Pennsylvania Center for the Treatment and Study of Anxiety. Her research has centered on traumatic stress.
Mitch Perkins is a doctoral student at Temple University. Their primary research focuses on alternative media for grassroots political communication and social movements. Their research also critically examines United States journalism’s relationship to grassroot politics and social movements.
Kathleen Reeves is the Chair of the Department of Urban Health and Population Sciences. A faculty leader since 2003, Dr. Reeves is also Professor of Clinical Pediatrics and founding Director of the Center for Urban Bioethics. Her research centers on trauma informed approaches to k-12 education, violence prevention, substance use disorder, and food insecurity.
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding: The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by a grant from the City of Philadelphia’s Department of Public Health (PDPH) with Federal Emergency Management Agency (FEMA) funds from the U.S. government.
1. As of August 2021: 19131, 19139, 19141, 19142, 19151
2. This research was approved as exempt from Human Subjects Research review by the Institutional Review Board at Temple University.
==== Refs
References
Aslan M. Wanamaker M. (2016). Tuskegee and the health of Black men. The Quarterly Journal of Economics, 133 (1 ), 407–455. 10.1093/qje/qjx029
Bass S. B. Wilson-Genderson M. Garcia D. T. Akinkugbe A. A. Mosavel M. (2021). SARS-CoV-2 vaccine hesitancy in a sample of US adults: Role of perceived satisfaction with health, access to healthcare, and attention to COVID-19 news. Frontiers in Public Health, 9 (665724 ), 1–5. 10.3389/fpubh.2021.665724
BBC News (2022). Covid map: Coronavirus cases, deaths, vaccinations by county. BBC [Internet]. https://www.bbc.com/news/world-51235105
Bradley E. H. Curry L. A. Devers K. J. (2007). Qualitative data analysis for health services research: Developing taxonomy, themes, and theory. Health Services Research, 42 (4 ), 1758–1772. 10.1111/j.1475-6773.2006.00684.x 17286625
Braun V. Clarke V. (2006). Using thematic coding in psychology. Qualitative Research in Psychology, 3 (2 ), 77–101. 10.1191/1478088706qp063oa
Centers for Disease Control and Prevention (2021, August 26). Delta variant: What we know about the science. Retrieved February 14, 2022, from https://www.cdc.gov/coronavirus/2019-ncov/variants/delta-variant.html?s_cid=11608:which%20vaccine%20is%20better%20for%20delta%20variant:sem.ga:p:RG:GM:gen:PTN.Grants:FY22
Centers for Disease Control and Prevention (2022). Covid data tracker weekly review. Retrieved July 6, 2022, from https://www.cdc.gov/coronavirus/2019-ncov/Covid-data/Covidview/index.html
City of Philadelphia (2022). Covid-19 vaccine data. Retrieved February 14, 2022, from https://www.phila.gov/programs/coronavirus-disease-2019-covid-19/vaccines/vaccine-data/
Feldman N. (2021, April 6). Why Black and Latino people still lag on Covid vaccines – and how to fix it. NPR. https://www.npr.org/sections/health-shots/2021/04/26/989962041/why-black-and-latino-people-still-lag-on-covid-vaccines-and-how-to-fix-it
Fishbein M. Capella J. N. (2006). The role of theory in developing effective health communications. Journal of Communication, 56 (1 ), S1–S17. 10.1111/j.1460-2466.2006.00280.x
Forman R. Shah S. Jeurissen P. Jit M. Mossialos E. (2021). COVID-19 vaccine challenges: What have we learned so far and what remains to be done? Health Policy, 125 (5 ), 553–567. 10.1016/j.healthpol.2021.03.013 33820678
Frieden T. (2021, May 10). Dr. Tom Frieden addresses vaccine hesitancy, and role of politics. YouTube: American Medical Association. https://www.youtube.com/watch?v=zjSv33vlqOM
Iuliano A. D. Brunkard J. M. Boehmer T. K. Peterson E. Adjei S. Binder A. M. Cobb S. Graff P. Hidalgo P. Panaggio M. J. Rainey J. J. Rao P. Soetebier K. Wacaster S. Ai C. Gupta V. Molinari N. M. Ritchey M. D. (2022). Trends in disease severity and health care utilization during the early Omicron variant period compared with previous SARS-COV-2 high transmission periods—United States, December 2020–January 2022. Centers for Disease Control and Prevention Morbidity and Mortality Weekly Report, 71 (4 ), 146–152. 10.15585/mmwr.mm7104e4
Joffe H. (2012). Thematic analysis. In Harper D. Thompson A.R. (Eds.), Qualitative research methods in mental health and psychotherapy: A guide for students and practitioners (pp. 209–223). John Wiley & Sons.
Joffe H. Haarhoff G. (2002). Representations of far-flung illnesses: The case of Ebola in Britain. Social Science & Medicine, 54 (6 ), 955–969. 10.1016/s0277-9536(01)00068-5 11996028
Karni A. Kanno-Youngs Z. (2021, March 15). As Biden confronts vaccine hesitancy, Republicans are a particular challenge. The New York Times. https://www.nytimes.com/2021/03/15/us/politics/joe-biden-vaccine-republicans.html
Latkin C. Dayton L. Yi G. Jaleel A. Nwosu C. Limaye R. (2021). COVID-19 vaccine delay: An examination of United States residents’ intention to delay vaccine uptake. Human vaccines and immunotherapeutics, 17 (9 ), 2903–2913. 10.1080/21645515.2021.1917234 34014130
Laughlin J. Lai J. (2021, August 10). In Philly’s least-vaccinated zip codes, fighting the Delta variant is a game of catch-up. The Philadelphia Inquirer. https://www.inquirer.com/health/coronavirus/vaccine-zip-code-philadelphia-rates-20210810.html
Lee N. Kotler P. (2008). Social marketing: Influencing behaviors for good. Sage.
Lerer L. (2021, July 17). How Republican vaccine opposition got to this point. The New York Times. https://www.nytimes.com/2021/07/17/us/politics/coronavirus-vaccines-republicans.html
Lozano A. V. (2020). Racial disparities create obstacles for Covid-19 vaccine rollout. NBC News Online. https://www.nbcnews.com/news/us-news/racial-disparities-create-obstacles-Covid-19-vaccine-rollout-n1249627
Peters M. D. J. (2022). Addressing vaccine hesitancy and resistance for COVID-19 vaccines. International Journal of Nursing Studies, 131 , 1–8. 10.1016/j.ijnurstu.2022.104241
Rane M. S. Kochhar S. Poehlein E. You W. Robertson M. M. Zimba R. Westmoreland D. A. Romo M. L. Kulkarni S. G. Chang M. Berry A. Parcesepe A. M. Maroko A. R. Grov C. Nash D. (2022). Determinants and trends of COVID-19 vaccine hesitancy and vaccine uptake in a national cohort of US adults: A longitudinal study. American Journal of Epidemiology, 191 (4 ), 570–583. 10.1093/aje/kwab293 34999751
Rogers E. M. (1983). Diffusion of innovations. Free Press.
Rutten L. J. F. Zhu X. Leppin A. L. Ridgeway J. L. Swift M. D. Griffin J. M. St Sauver J. L. Virk A. Jacobson R. M. (2021). Evidence-based strategies for clinical organizations to address Covid-19 vaccine hesitancy. Mayo Clinic Proceedings, 96 (3 ), 699–707. 10.1016/j.mayocp.2020.12.024 33673921
U.S. Census Bureau (2019). American Community Survey 5-year estimates. Retrieved December 3, 2021, from https://censusreporter.org/profiles/16000US4260000-philadelphia-pa/
| 0 | PMC9705503 | NO-CC CODE | 2022-12-01 23:19:37 | no | Am Behav Sci. 2022 Nov 25;:00027642221138279 | utf-8 | Am Behav Sci | 2,022 | 10.1177/00027642221138279 | oa_other |
==== Front
Ann Otol Rhinol Laryngol
Ann Otol Rhinol Laryngol
AOR
spaor
The Annals of Otology, Rhinology, and Laryngology
0003-4894
1943-572X
SAGE Publications Sage CA: Los Angeles, CA
36433692
10.1177/00034894221137273
10.1177_00034894221137273
Original Article
Otolaryngology Patient Satisfaction with In-Office Appointments and Virtual Visits Due to COVID-19
https://orcid.org/0000-0002-1686-6963
Arrighi-Allisan Annie E. BA 1
Wong Anni MD 1
Gidumal Sunder MD 1
Shah Janki MD 1
Filip Peter MD 1
Omorogbe Aisosa BS 1
Rosenberg Joshua MD 1
Govindaraj Satish MD 1
Iloreta Alfred-Marc MD 1
1 Department of Otolaryngology, Icahn School of Medicine at Mount Sinai, New York, NY, USA
Annie E. Arrighi-Allisan, BA, Department of Otolaryngology – Head and Neck Surgery, Icahn School of Medicine at Mount Sinai, 3400 Spruce St, New York, NY 19104, USA. Email: [email protected]
25 11 2022
25 11 2022
00034894221137273© The Author(s) 2022
2022
SAGE Publications
This article is made available via the PMC Open Access Subset for unrestricted re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the COVID-19 pandemic or until permissions are revoked in writing. Upon expiration of these permissions, PMC is granted a perpetual license to make this article available via PMC and Europe PMC, consistent with existing copyright protections.
Background:
The COVID-19 pandemic forced otolaryngologists to seek new methods of providing patient care in a remote setting. The effect of this paradigm shift on patient satisfaction, however, remains unelucidated. This study compares patient satisfaction with telehealth visits during the COVID-19 pandemic to that with in-office visits during the same period in 2019.
Methods:
Press Ganey survey responses of patients seen by otolaryngologists within a large, academic, multicenter hospital system were gathered. Responses were included in analyses if they corresponded with a visit that occurred either in clinic March to December 2019 or via telehealth March to December 2020. Chi-Square Test of Independence and Fisher’s Exact Test were employed to detect differences between years. Binary logistic regressions were performed to detect the factors most predictive of positive telehealth experiences.
Results:
Patient overall satisfaction with in-office and telehealth visits did not differ significantly (76.4% in 2019 vs 78.0% in 2020 rated visit overall as “very good,” P = .09). Patients seen by a Head and Neck (odds ratio 4.13, 95% confidence interval 1.52-11.26, P = .005), Laryngology (OR 5.96, 95% CI 1.51-23.50, P = .01), or Rhinology (OR 4.02, 95% CI 1.55-10.43, P = .004) provider were significantly more likely to report a positive telehealth experience.
Conclusions:
Patients seen via telehealth during COVID-19 reported levels of satisfaction similar to those seen in-office the year prior. These telehealth satisfaction levels, however, are contextualized within the expected confines of a pandemic. Further research is required to determine whether satisfaction remains consistent as telemedicine becomes a ubiquitous component of medical practice.
quality of life
patient reported outcomes
otolaryngology
telemedicine
COVID-19
edited-statecorrected-proof
typesetterts1
==== Body
pmcIntroduction
The widespread adoption of telehealth technologies allowed many patients to seek or continue care during the COVID-19 pandemic. While a paltry 0.1% of Medicare primary care visits were conducted through telehealth in February 2020, that number rose to 43.5% just 2 months later.1 As a precipitous change to remote interactions in early spring proved enduring throughout the summer and fall, providers augmented efforts to provide routine care in a remote setting. While initially sought as a temporary remedy, the expectation of a majority of stakeholders in the American healthcare system is that telehealth will become an accepted and ubiquitous component of medical practice in the United States.2-5 Some experts estimate that up to $250 billion dollars in healthcare expenditures could be feasibly and effectively translated to a virtual platform.6
Prior to the pandemic, providers had the luxury of cherry-picking the types of visits occurring remotely. Rehabilitation, radiology, and genetic counseling services gained easy patient acceptance as the substance of these visits was well-suited to a remote setting,7-10 but telemedicine visits overall remained a stark minority due to limited reimbursement rates and patient and provider reticence.11,12 Patient acceptance and perceived value are integral to the successful widespread adoption of telehealth in American medical practice.
Though the expansion of telemedicine is in its nascent stages, several studies have begun to explore the impact of expanded telehealth on patient satisfaction. Early reports in internal medicine,13 family medicine,14 gastroenterology,15 and oncology16 have evidenced relatively high levels of overall patient satisfaction. There remains a paucity of research, however, exploring this dynamic within surgical fields.17,18 A small number of studies within otolaryngology have reported promising results, but the interpretation of satisfaction is innately relative, and these studies fail to contextualize their findings within preexisting patient satisfaction levels prior to the pandemic.19-22 The authors hypothesize that patient satisfaction with otolaryngology telehealth visits during the COVID-19 pandemic will be comparable to that with in-office visits during the same period in 2019. We examine these satisfaction levels at the patient, divisional, and departmental levels to elucidate the factors most predictive of a positive telehealth experience.
Materials and Methods
Participants
Press Ganey survey responses of patients seen by an otolaryngologist within a large, academic, multicenter hospital system were gathered. Telehealth visits were universally offered to patients seeking an initial or follow-up visit. Though the proportion of new to return visits varied by month, new visits comprised slightly more than one-third of all visits in 2020, which was a slightly lower proportion than the year prior (33.93% in 2020 vs 38.34%). Providers were required to offer these visits. Responses were included in analyses if they corresponded with a visit that occurred either in clinic from March to December 2019 or via telehealth from March to December 2020.
Study Design
This study was approved by the Mount Sinai Program for the Protection of Human Subjects. Patient satisfaction was assessed by both quantitative and qualitative measures. Numerical responses to Press Ganey patient satisfaction surveys were grouped and compared by both year (2019 vs 2020) and subspecialty. In-office and telehealth patient satisfaction surveys from 2019 and 2020, respectively, asked patients to rate different aspects of their visit. Questions not shared by both surveys (ie, cleanliness of clinic, nurse or assistant friendliness, etc.) were excluded from analyses. A composite telehealth satisfaction score was calculated by averaging each individual’s numerical responses to telehealth-related questions. A binary variable was then created to indicate whether an individual had assigned the telehealth component of their visit a perfect rating, which constituted a “strongly positive” telehealth experience. Press Ganey satisfaction survey comments were analyzed and assigned a valence of positive, neutral, mixed, or negative (+1, 0, 0, and −1, respectively) based on pre-coded keywords. For patients who commented on more than one aspect of their visit, aggregate scores were calculated by averaging all available values. Blank comment sections were counted as neutral and assigned a valence of 0.
Statistical Analyses
Descriptive statistics were employed to examine baseline patient characteristics. Chi-Square Test of Independence and Fisher’s Exact Test were utilized to compare numerical ratings and comments from both years. Binary logistic regressions, controlling for patient age, race/ethnicity, and sex were performed to detect factors predictive of strongly positive telehealth experiences. All statistical analyses were performed using IBM’s Statistical Product and Service Solutions software (SPSS; IBM, Chicago, Illinois). Statistical significance (P value) was set at less than .05.
Results
A number of 4713 (4434 in 2019, 279 in 2020) numerical survey responses were received and included in analyses. The 2020 telehealth response cohort skewed significantly younger than those responding to in-office visits in 2020 (62.7% below age 65 in 2019 vs 69.5% in 2020, P = .001). The vast majority of responses in both years were from patients whose primary language was English (98.7% vs 98.9% English speakers responding in 2019 and 2020, respectively, P = .91). Type of otolaryngology provider seen varied considerably between cohorts, with significantly fewer patients responding to general otolaryngology versus subspecialty visits in 2020 when compared with 2019 (19.0% vs 31.7% general, respectively, P < .0001). Demographics of the 2019 and 2020 response cohorts can be found in Table 1.
Table 1. Demographics of Patients Seen In-Office in 2019 and via Telehealth in 2020.
Demographic 2019 (n = 4434) 2020 (n = 279) P value
Sex (Female) 50.4% (2236) 52.0% (145) .62
Age, y .001
0-17 6.2% (273) 2.9% (8)
18-34 10.3% (458) 15.4% (43)
35-49 14.8% (658) 19.7% (55)
50-64 31.4% (1394) 31.5% (88)
65-79 30.3% (1344) 26.5% (74)
80+ 6.9% (307) 3.9% (11)
Language .91
English 98.7% (4377) 98.9% (276)
Spanish 1.2% (55) 1.1% (3)
Race/Ethnicity <.0001
White 34.1% (1512) 49.5% (138)
Black or African American 3.9% (175) 5.4% (15)
Hispanic or Latino 14.3% (635) 0% (0)
Asian 1.4% (62) 1.8% (5)
Unknown 46.1% (2044) 42.7% (119)
Provider subspecialty <.0001
General 31.7% (1404) 19.0% (53)
Facial plastics 5.8% (255) 3.9% (11)
Head and neck surgery 10.2% (454) 26.5% (74)
Laryngology 10.6% (472) 12.2% (34)
Otology 6.9% (305) 4.3% (12)
Pediatric ENT 5.1% (227) 2.5% (7)
Rhinology 20.3% (898) 22.9% (64)
Sleep medicine 9.4% (419) 8.6% (24)
Table 2 compares patient satisfaction with a multitude of visit elements between 2019 and 2020. Patients reported comparable overall satisfaction with 2019 in-office and 2020 virtual visits (76.4% vs 78.0%, respectively, rated visit overall as “very good,” P = .09). Patient satisfaction with the provider overall was slightly higher in 2020, though not to a significant level (83.3% in 2019 vs 87.5% in 2020 rated provider as “very good,” P = .46). Satisfaction with office and provider accessibility in 2020 exhibited a bimodal distribution when compared to 2019; though a greater overall number of patients in 2020 found the clinic much easier to contact (69.9% in 2020 vs 64.9% in 2019), patients were also more likely to rate accessibility poorly than were their 2019 counterparts (14.7% rated contact ease as “very poor” to “fair” in 2020 vs 11.0% in 2019, P = .0004 for both). Respondents in 2019 were more likely to recommend the practice to others than were those seen via telehealth in 2020 (95.6% vs 90.4% likely or very likely to recommend, P = .002). No other significant differences emerged between 2019 and 2020 satisfaction elements (Table 2).
Table 2. Comparison of In-Office and Telehealth Patient Satisfaction Ratings.
Element of visit 2019 2020
Very poor Poor Fair Good Very good Very poor Poor Fair Good Very good P value
Visit overall 0.9% (41) 1.1% (47) 3.9% (177) 17.8% (800) 76.4% (3440) 1.0% (3) 2.4% (7) 5.1% (15) 13.9% (41) 78.0% (231) .09
Ease of scheduling appointments 1.0% (45) 1.9% (84) 6.6% (293) 21.7% (969) 68.8% (3066) 1.4% (4) 2.0% (6) 7.5% (22) 16.0% (47) 73.1% (215) .16
Ease of contacting 1.6% (65) 1.8% (76) 7.6% (319) 24.1% (1012) 64.9% (2722) 0.7% (2) 4.1% (12) 9.9% (29) 15.4% (45) 69.9% (204) .0004
Ease of video visit instructions 1.4% (4) 3.4% (10) 6.2% (18) 17.9% (52) 71.1% (207)
Provider overall 0.9% (39) 0.8% (35) 2.4% (106) 12.7% (566) 83.3% (3713) 0.3% (1) 0.3% (1) 1.7% (5) 9.8% (29) 87.5% (258) .46
Explanations of prob/condition 0.6% (25) 0.8% (35) 2.6% (113) 12.7% (559) 83.4% (3686) 0.3% (1) 0.7% (2) 1.0% (3) 10.6% (31) 87.3% (255) .40
Concern for questions/worries 0.7% (29) 0.7% (29) 2.3% (102) 13.6% (604) 82.8% (3667) 0.3% (1) 0.3% (1) 1.7% (5) 9.8% (29) 87.8% (259) .33
Efforts to include in decisions 0.6% (27) 0.9% (38) 2.6% (113) 13.4% (585) 82.5% (3592) 0.3% (1) 0.3% (1) 2.1% (6) 9.2% (27) 88.0% (257) .22
Likelihood of recommending 1.6% (72) 0.8% (36) 1.9% (83) 10.3% (451) 85.4% (3750) 0.3% (1) 0.3% (1) 2.4% (7) 8.8% (26) 88.1% (259) .31
Discussion of treatments 0.9% (39) 0.9% (36) 2.5% (106) 13.3% (565) 82.4% (3503) 0.3% (1) 0.3% (1) 1.7% (5) 11.3% (33) 86.3% (253) .59
Staff took care for you 0.5% (20) 0.6% (27) 2.8% (120) 17.3% (756) 78.8% (3437) 1.0% (3) 0.7% (2) 1.7% (5) 15.1% (44) 81.5% (238) .34
Likelihood of recommending practice 1.3% (55) 0.8% (33) 2.4% (105) 12.2% (534) 83.4% (3663) 1.4% (4) 2.4% (7) 5.8% (17) 15.1% (44) 75.3% (220) .002
Telemedicine overall 1.7% (5) 4.4% (13) 7.1% (21) 15.6% (46) 71.1% (209)
Ease of talking over video 0.7% (2) 3.7% (11) 6.1% (18) 14.6% (43) 74.8% (220)
Video connection 2.0% (6) 4.4% (13) 7.2% (21) 14.7% (43) 71.7% (210)
Audio connection 1.4% (4) 4.1% (12) 8.6% (25) 16.8% (49) 69.1% (201)
Ease logging in 1.0% (6) 5.5% (16) 6.8% (20) 17.1% (50) 68.6% (201)
A number of 905 (747 in 2019, 158 in 2020) free-text survey responses were received, assigned valences (strongly positive, positive, neutral, negative, or strongly negative), and included in analyses (Table 3). Similar levels of satisfaction in 2019 and 2020 were elicited in the department overall (P = .63), and across all 8 otolaryngology subspecialty practices: General, Facial Plastics, Head and Neck, Laryngology, Otology, Pediatrics, Rhinology, and Sleep Medicine (P = .07, .77, .51, .69, .71, 1, .52, and .35, respectively). A full report of comment valence analyses can be found in Table 3.
Table 3. Comparison of 2019 and 2020 Comment Valence Analyses.
All divisions
Year Strongly positive Positive Neutral Negative Strongly negative P value
2019 36.9% (276) 5.0% (37) 41.9% (313) 3.2% (24) 13.0% (97)
2020 36.1% (57) 3.8% (6) 46.8% (74) 3.8% (6) 9.5% (15) .63
2019 2020
Subspecialty Strongly positive Positive Neutral Negative Strongly negative Strongly positive Positive Neutral Negative Strongly Negative P value
General 34.2% (80) 11.1% (26) 39.7% (93) 3.8% (9) 11.1% (26) 22.7% (5) 0.0% (0) 45.5% (10) 13.6% (3) 18.2% (4) .07
Facial plastics 41.9% (13) 0.0% (0) 45.2% (14) 3.2% (1) 9.7% (3) 28.6% (2) 0.0% (0) 57.1% (4) 0.0% (0) 14.3% (1) .77
Head and neck 40.0% (40) 10% (10) 38.0% (38) 2.0% (2) 10.0% (10) 31.7% (13) 4.9% (2) 53.7% (22) 2.4% (1) 7.3% (3) .51
Laryngology 38.3% (44) 8.7% (10) 39.1% (45) 2.6% (3) 11.3% (13) 25.0% (4) 12.5% (2) 56.3% (9) 0.0% (0) 6.3% (1) .69
Otology 11.1% (1) 0.0% (0) 44.4% (4) 0.0% (0) 44.4% (4) 30.0% (3) 0.0% (0) 30.0% (3) 0.0% (0) 40.0% (4) .71
Pediatric ENT 36.7% (18) 2.0% (1) 53.1% (26) 0.0% (0) 8.2% (4) 50.0% (1) 0.0% (0) 50.0% (1) 0.0% (0) 0.0% (0) 1
Rhinology 32.7% (35) 1.9% (2) 50.5% (54) 3.7% (4) 11.2% (12) 37.5% (12) 3.1% (1) 56.3% (18) 0.0% (0) 3.1% (1) .52
Sleep medicine 38.5% (35) 4.4% (4) 37.4% (34) 2.2% (2) 17.6% (16) 55.0% (11) 5.0% (1) 30.0% (6) 5.0% (1) 5.0% (1) .35
A multivariable logistic regression was performed (Table 4), controlling for patient sex, age, and race/ethnicity, to examine whether otolaryngology division correlated with a patient’s likelihood of having a strongly positive telehealth experience. Patients seen by a provider within Head and Neck (odds ratio 4.13, 95% CI 1.52-11.26, P = .005), Laryngology (OR 5.96, 95% CI 1.51-23.50, P = .01), or Rhinology (OR 4.02, 95% CI 1.55-10.43, P = .004) were approximately 4, 6, and 4 times as likely, respectively, to have a strongly positive telehealth experience than were patients seen by generalists.
Table 4. Predictors of Strongly Positive Telehealth Experiences.
Factor Odds ratio 95% confidence interval P value
Sex
Female Ref
Male 1.15 0.59-2.24 .69
Age .69
0-17 Ref
18-34 - 0 1
35-49 - 0 1
50-64 - 0 1
65-79 - 0 1
80+ - 0 1
Race/Ethnicity 1
White Ref
Black or African American 1.01 0.25-4.36 .99
Hispanic or Latino 189 526 561.69 0 1
Asian 0.76 0.07-8.52 .82
Unknown 1.02 0.51-2.05 .95
Subspecialty seen .02
General Ref
Facial Plastics 1.00 0.24-4.19 1.00
Head and Neck 4.13 1.52-11.26 .005
Laryngology 5.96 1.51-23.50 .01
Otology 1.17 0.31-4.51 .82
Pediatric ENT 5.41 0 1
Rhinology 4.02 1.55-10.43 .004
Sleep Medicine 1.45 0.48-4.43 .51
Discussion
The COVID-19 pandemic has compelled otolaryngologists to seek new methods of providing patient care in a remote setting. Our study demonstrates that otolaryngology patient satisfaction with telehealth visits in 2020, assessed by both quantitative and qualitative means, was high and comparable across numerous measures to that with in-person visits prior to the pandemic. Our study is the largest, in both patient sample size and time period examined, to explore telehealth satisfaction within otolaryngology, and just the second to compare otolaryngology patient satisfaction with virtual and in-person formats.
The cohort of patients who completed the telehealth satisfaction survey were significantly younger than those whose survey answers corresponded to an in-person visit (Table 1). Age has been found, in many settings, to inversely correlate with technological proficiency.15,23,24 A smaller proportion of older patients therefore may have opted for remote visits, yielding a correspondingly lower number of survey responses from adults over age 65. An additional source of this discrepancy may lie in the format in which the surveys were presented. Patients seen in 2019 were given the option of completing their survey in person; this option, in contrast, was unavailable to those seen remotely, perhaps precluding patients with lower technological familiarity from responding.
The present analyses found overall patient satisfaction levels with in-person and virtual visits to be equivalent (Table 2). Just one extant study has compared patient satisfaction levels with otolaryngology visits before and during the pandemic. In contrast to our findings, Itamura et al19 reported that otolaryngology patients consistently rated their telehealth visits more poorly than in-person ones. Perceived provider listening and effective conveyance of information were most notably reduced. There are several potential etiologies for these differences in perception. First, Itamura et al’s analysis included responses to telehealth visits for just the months of March and April 2020, 2 months at the very inception of the pandemic. Our data spans March through December of 2020, during which providers underwent additional telemedicine training and practices optimized their virtual workflows. Our augmented time period also allowed our study a threefold increase in sample size. Patient demographics and visit details, such as patient age, gender, or subspecialty seen, were not included in Itamura et al’s analyses, which precludes us from drawing any conclusions between patient population composition and resulting telehealth satisfaction levels. Nonetheless, the aforementioned differences in setting may have played a significant role.
The present study revealed that patients found providers and practices significantly easier to contact during the pandemic (Table 2). Office staff who have transitioned to a work-from-home format are more readily available to staff phonelines and respond to patient queries. Patients seen virtually do not have to be transported from waiting to examination rooms, improving the likelihood that providers remain on schedule and reducing a patient’s overall wait time. These findings are supported by numerous studies of patient satisfaction, both within and beyond otolaryngology, that laud decreased scheduling and day-of wait times as significant advantages of telemedicine.22,25-29 Though not explicitly recorded in our analyses, it follows logically that telemedicine eliminates significant commute times for patients, which can be up to several hours in each direction.27,30
Despite comparable satisfaction levels with providers and visits overall, patients seen virtually were significantly less likely to recommend the practice and format to others than were patients who had been seen in person the previous year. Telehealth patient satisfaction studies in other fields have reported similar dichotomous findings.18,21 The vast majority of patients who scheduled telehealth visits with otolaryngologists in 2020, however, did so because it was a necessity, and not necessarily because they perceived telemedicine as materially benefiting their quality of care under all circumstances. Accordingly, a patient’s satisfaction with their individual provider and gratitude for care received may not be commensurate with a desire for telehealth permanence.
Providers and patients agree that telemedicine, in its current form, is unable to entirely replace in-person interactions.12,22,31-34 While the classic model of patient care was defined by the in-person history and physical exam, little more than a cursory physical examination is possible during most telehealth visits. This limitation is magnified in a field such as otolaryngology, where endoscopic and microscopic exams are routine elements of a patient visit.19,22 One survey study performed last year found that nearly one-third of rhinology patients seen via telehealth felt that something was missed or left unaddressed in the absence of an in-person examination.22
This practice gap has been met with the rapid expansion of remote examination technologies. Smartphone attachments are now capable of enabling otoscopy or anterior rhinoscopy and capturing oropharyngeal images.6 Mobile technology has proven efficacious in tracking and improving functional outcomes of patients following septoplasty, functional endoscopic sinus surgery, and allergic rhinitis.35,36 A handheld device now allows otoscopic visualization and records heart and lung sounds at higher image and sound qualities than those produced by traditional otoscopes and stethoscopes.37 Though the daily utility and affordability of these technologies in a remote otolaryngology setting remains unelucidated, these new tools have garnered auspicious initial reviews.
Multivariate analysis revealed that provider type correlated significantly with positive telehealth experiences. Specifically, when controlling for age, sex, and race/ethnicity, patients seen by Rhinology, Head and Neck, and Laryngology providers were significantly more likely than those seen by generalists to report a strongly positive telehealth experience. This may be partially explained by the degree to which these visits routinely incorporate interventions, and the rapidity with which specialists responded to a shifting paradigm. A deluge of patients over the last year have sought rhinology consultation for anosmia resulting from COVID-19, the purported treatments for which may all be explained or prescribed virtually.5 Patients with allergic rhinitis or chronic sinusitis may be evaluated virtually for initial or worsening symptoms; this approach is particularly feasible for patients familiar to the provider with previously documented nasal endoscopies or sinus CTs.38 By April 2020, an estimated 96.2% of rhinologists were already incorporating telemedicine into their practices.39
Patients with previously identified vocal pathologies, returning for voice therapy or routine follow-up, are particularly well-suited for a virtual platform.40 In response to the COVID-19 pandemic, the American Speech-Language-Hearing Association (ASHA) loosened treatment prerequisites for those awaiting objective laryngologic assessment, allowing new patients to initiate voice therapy prior to stroboscopic evaluation.41 These factors may all contribute to a laryngology patient’s improved telehealth experience.
The link between telehealth and improved patient experience in head and neck surgery has been previously described.5 One randomized control trial found that head and neck cancer patients who were monitored via telehealth reported significantly improved symptom burdens and quality of life scores when compared to those monitored through standard-of-care in-person appointments.42 Head and neck cancer often requires a multidisciplinary approach with consultations from numerous specialties. Allowing many of these to proceed via virtual means, in addition to a shorter scheduling latency, may provide substantial comfort to patients feeling overwhelmed by their diagnosis.
These particular findings should be taken as preliminary, however, as they may be limited in their generalizability. Some subspecialized otolaryngologists still see a small portion of patients whose primary otolaryngologic symptoms lie outside the confines of their subspecialty. Lower numbers of Press Ganey responses for individual providers may additionally skew results. Regardless, our data illustrate that a greater proportion of virtual visits in 2020 occurred with subspecialists than in 2019, perhaps suggesting that patients with specific and complex problems requiring subspecialty care are more likely to readily engage with telehealth. Further investigation into the nuances of telehealth preferences across otolaryngology subspecialties is warranted.
Our study is limited by several factors. First, data are based on voluntary patient responses, and are therefore subject to sampling bias, heightened by low response rates (approximately 4% in 2019 and 1% in 2020). In addition, the 2020 telehealth satisfaction survey administered by the department did not ask whether respondents identified as Hispanic or Latino, which precluded us from assessing those specifically identifying with this ethnic group in 2020. Nonetheless, the combined race/ethnicity variable was found to be highly uncorrelated with patient responses, suggesting that this difference did not meaningfully impact results. Due in large part to lower patient volume during COVID’s peak, our sample for 2020 telehealth visits was significantly smaller than for in-person visits in 2019. Additionally, patients seen in 2019 were given the option of completing their survey in person, while those seen remotely were only offered an emailed survey. This difference may also have contributed to a lower capture rate of responses in 2020 when compared to 2019. This likely means that our ability to comment on divisional differences was underpowered. As patient satisfaction data was deidentified, we were unable to retroactively link patient responses with insurance type, a variable often used as a proxy for socioeconomic status.43 This demographic variable may modulate patient satisfaction levels, perhaps to a greater degree in a telehealth setting, as computer and internet speed significantly affect the quality and quantity of virtual visits.43
Conclusion
The present analyses demonstrated that patients seen via telehealth during COVID-19 experienced levels of satisfaction similar to those seen in-office the year prior. Telehealth satisfaction levels, however, are contextualized within the expected confines of a pandemic. Future studies should address and seek to mitigate barriers to telehealth accessibility such as socioeconomic status, age, or health literacy, as well as those exclusive to specific otolaryngology subspecialties. Sophisticated remote examination technology represents the next frontier and must be incorporated in order for telemedicine’s full value within otolaryngology to be realized.
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding: The author(s) received no financial support for the research, authorship, and/or publication of this article.
ORCID iD: Annie E. Arrighi-Allisan https://orcid.org/0000-0002-1686-6963
==== Refs
References
1 Bosworth ARJ Samson LW Sheingold S Taplin C Tarazi W Zuckerman R . Medicare Beneficiary Use of Telehealth Visits: Early Data from the Start of COVID-19 Pandemic. Washington, DC: Office of the Assistant Secretary for Planning and Evaluation, U.S. Department of Health and Human Services; 2020.
2 Svider PF Setzen M Ow R Folbe AJ Eloy JA Johnson AP. Incorporation of telemedicine by rhinologists: the COVID-19 pandemic and beyond. Am J Otolaryngol. 2020;41 (6 ):102567.32920475
3 Abdel-Wahab M Rosenblatt E Prajogi B Zubizarretta E Mikhail M. Opportunities in telemedicine, lessons learned after COVID-19 and the way into the future. Int J Radiat Oncol Biol Phys. 2020;108 (2 ):438-443.32890528
4 Olson CA Thomas JF. Telehealth: no longer an idea for the future. Adv Pediatr. 2017;64 (1 ):347-370.28688597
5 Samarrai R Riccardi AC Tessema B Setzen M Brown SM. Continuation of telemedicine in otolaryngology post-COVID-19: applications by subspecialty. Am J Otolaryngol. 2021;42 (3 ):102928.33545447
6 Bestsennyy O GG Harris A Rost J. Telehealth: a quarter-trillion-dollar post-COVID-19 reality? May 29, 2020. https://www.mckinsey.com/industries/healthcare-systems-and-services/our-insights/telehealth-a-quarter-trillion-dollar-post-covid-19-reality. Accessed April 23, 2021.
7 Bradbury A Patrick-Miller L Harris D , et al . Utilizing remote real-time videoconferencing to expand access to cancer genetic services in community practices: a multicenter feasibility study. J Med Internet Res. 2016;18 (2 ):e23.
8 Jacobs JJWM Ekkelboom R Jacobs JPAM van der Molen T Sanderman R . Patient satisfaction with a teleradiology service in general practice. BMC Fam Pract. 2016;17 :17.26864118
9 Hoaas H Andreassen HK Lien LA Hjalmarsen A Zanaboni P. Adherence and factors affecting satisfaction in long-term telerehabilitation for patients with chronic obstructive pulmonary disease: a mixed methods study. BMC Med Inform Decis Mak. 2016;16 :26.26911326
10 Dias AE Limongi JC Barbosa ER Hsing WT. Voice telerehabilitation in Parkinson’s disease. CoDAS. 2016;28 (2 ):176-181.27191882
11 Knierim K Palmer C Kramer ES , et al . Lessons learned during COVID-19 that can move telehealth in primary care forward. J Am Board Fam Med. 2021;34 (Suppl ):S196-s202.
12 Ohlstein JF Garner J Takashima M. Telemedicine in otolaryngology in the COVID-19 era: initial lessons learned. Laryngoscope. 2020;130 :2568-2573.32740925
13 Bate NJ Xu SC Pacilli M Roberts LJ Kimber C Nataraja RM. Effect of the COVID-19 induced phase of massive telehealth uptake on end-user satisfaction. Intern Med J. 2021;51 (2 ):206-214.33631844
14 Volcy J Smith W Mills K , et al . Assessment of patient and provider satisfaction with the change to telehealth from in-person visits at an academic safety net institution during the COVID-19 Pandemic. J Am Board Fam Med. 2021;34 (Suppl ):S71-s76.
15 Lahat A Shatz Z. Telemedicine in clinical gastroenterology practice: what do patients prefer? Therap Adv Gastroenterol. 2021;14 :1756284821989178.
16 Sathiyaraj A Lopez H Surapaneni R. Patient satisfaction with telemedicine for prechemotherapy evaluation during the COVID-19 pandemic. Future Oncol. 2021;17 :1593-1600.33631995
17 Kirby DJ Fried JW Buchalter DB , et al . Patient and physician satisfaction with telehealth during the COVID-19 pandemic: sports medicine perspective. Telemed J E Health. 2021;27 :1151-1159.33512302
18 Sabbagh R Shah N Jenkins S , et al . The COVID-19 pandemic and follow-up for shoulder surgery: the impact of a shift toward telemedicine on validated patient-reported outcomes. J Telemed Telecare. 2021;2021 :1357633x21990997.
19 Itamura K Tang DM Higgins TS , et al . Comparison of patient satisfaction between virtual visits during the COVID-19 pandemic and In-person visits pre-pandemic. Ann Otol Rhinol Laryngol. 2021;130 :810-817.33251849
20 Riley PE Fischer JL Nagy RE , et al . Patient and provider satisfaction with telemedicine in otolaryngology. OTO Open. 2021;5 (1 ):2473974x20981838.
21 Lechien JR Radulesco T Distinguin L , et al . Patient and otolaryngologist perceptions of telemedicine during COVID-19 pandemic. Eur Arch Otorhinolaryngol. 2021;278 :4101.33523283
22 Hentati F Cabrera CI D’Anza B Rodriguez K. Patient satisfaction with telemedicine in rhinology during the COVID-19 pandemic. Am J Otolaryngol. 2021;42 (3 ):102921.33508592
23 Liu L Goodarzi Z Jones A Posno R Straus SE Watt JA. Factors associated with virtual care access in older adults: a cross-sectional study. Age Ageing. 2021;50 :1412-1415.33625475
24 Roberts A Johnston GHF Landells C. COVID-19: pivoting from in-person to virtual orthopedic surgical evaluation. Can J Surg. 2021;64 (1 ):E101-e102.
25 Hofstetter PJ Kokesh J Ferguson AS Hood LJ. The impact of telehealth on wait time for ENT specialty care. Telemed J E Health. 2010;16 (5 ):551-556.20575722
26 Triantafillou V Layfield E Prasad A , et al . Patient perceptions of head and neck ambulatory telemedicine visits: a qualitative study. Otolaryngol Head Neck Surg. 2021;164 :923-931.32662749
27 Rimmer RA Christopher V Falck A , et al . Telemedicine in otolaryngology outpatient setting-single center head and neck surgery experience. Laryngoscope. 2018;128 (9 ):2072-2075.29446446
28 Yeroushalmi S Millan SH Nelson K Sparks A Friedman AJ. Patient perceptions and satisfaction with teledermatology during the COVID-19 pandemic: a survey-based study. J Drugs Dermatol. 2021;20 (2 ):178-183.33538563
29 Sclafani AP Shomorony A Stewart MG Grunstein E Overdevest JB. Telemedicine lessons learned during the COVID-19 pandemic: the augmented outpatient otolaryngology teleconsultation. Am J Otolaryngol. 2021;42 (4 ):102960.33610921
30 Beswick DM Vashi A Song Y , et al . Consultation via telemedicine and access to operative care for patients with head and neck cancer in a Veterans Health Administration population. Head Neck. 2016;38 (6 ):925-929.26899939
31 Silver SL Lewis MN Ledford CJW . A stepwise transition to telemedicine in response to COVID-19. J Am Board Fam Med. 2021;34 (Suppl ):S152-s161.
32 Petersen W Karpinski K Backhaus L Bierke S Häner M. A systematic review about telemedicine in orthopedics. Arch Orthop Trauma Surg. 2021;141 :1731-1739.33635399
33 Robiony M Bocin E Sembronio S Costa F Arboit L Tel A. Working in the era of COVID-19: an organization model for maxillofacial surgery based on telemedicine and video consultation. J Craniomaxillofac Surg. 2021;49 (4 ):323-328. doi:10.1016/j.jcms.2021.01.027 33581957
34 Powell RE Henstenburg JM Cooper G Hollander JE Rising KL. Patient perceptions of telehealth primary care video visits. Ann Fam Med. 2017;15 (3 ):225-229.28483887
35 Bousquet J Hellings PW Agache I , et al . Allergic rhinitis and its impact on asthma (ARIA) Phase 4 (2018): change management in allergic rhinitis and asthma multimorbidity using mobile technology. J Allergy Clin Immunol. 2019;143 (3 ):864-879.30273709
36 Khanwalkar AR Shen J Kern RC , et al . Utilization of a novel interactive mobile health platform to evaluate functional outcomes and pain following septoplasty and functional endoscopic sinus surgery. Int Forum Allergy Rhinol. 2019;9 (4 ):345-351.30586227
37 McDaniel NL Novicoff W Gunnell B Cattell Gordon D. Comparison of a novel handheld telehealth device with stand-alone examination tools in a clinic setting. Telemed J E Health. 2019;25 (12 ):1225-1230.30561284
38 Seim NB Philips RHW Matrka LA , et al . Developing a synchronous otolaryngology telemedicine clinic: prospective study to assess fidelity and diagnostic concordance. Laryngoscope. 2018;128 (5 ):1068-1074.29076534
39 Setzen M Svider PF Setzen S Setzen G Eloy JA Johnson AP. The novel corona virus and rhinology: Impact on practice patterns and future directions. Am J Otolaryngol. 2020;41 (6 ):102569. doi:10.1016/j.amjoto.2020.102569
40 Doarn CR Zacharias S Keck CS Tabangin M DeAlarcon A Kelchner L. Design and implementation of an interactive website for pediatric voice Therapy-The concept of In-Between Care: a telehealth model. Telemed J E Health. 2019;25 (5 ):415-422.29870314
41 Becker DR Gillespie AI. In the zoom where it happened: Telepractice and the voice clinic in 2020. Semin Speech Lang. 2021;42 (1 ):64-72.33596605
42 Pfeifer MP Keeney C Bumpous J , et al . Impact of a telehealth intervention on quality of life and symptom distress in patients with head and neck cancer. J Community Support Oncol. 2015;13 (1 ):14-21.25839061
43 Darrat I Tam S Boulis M Williams AM. Socioeconomic disparities in patient use of telehealth during the coronavirus disease 2019 Surge. JAMA Otolaryngol Head Neck Surg. 2021;147 (3 ):287-295. doi:10.1001/jamaoto.2020.5161 33443539
| 36433692 | PMC9705504 | NO-CC CODE | 2022-12-01 23:19:37 | no | Ann Otol Rhinol Laryngol. 2022 Nov 25;:00034894221137273 | utf-8 | Ann Otol Rhinol Laryngol | 2,022 | 10.1177/00034894221137273 | oa_other |
==== Front
Sociology
Sociology
SOC
spsoc
Sociology
0038-0385
1469-8684
SAGE Publications Sage UK: London, England
10.1177/00380385221136035
10.1177_00380385221136035
Article
Me? A Hero? Gendered Work and Attributions of Heroism among Volunteers during the COVID-19 Pandemic
Leap Braden Mississippi State University, USA
Kelly Kimberly Mississippi State University, USA
Stalp Marybeth C University of Northern Iowa, USA
Braden Leap, Mississippi State University, 207 Bowen Hall, 456 Hardy Rd, PO Box C, Mississippi State, MS 39762, USA. Email: [email protected]
25 11 2022
25 11 2022
003803852211360354 2022
10 2022
© The Author(s) 2022
2022
BSA Publications Ltd.
This article is made available via the PMC Open Access Subset for unrestricted re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the COVID-19 pandemic or until permissions are revoked in writing. Upon expiration of these permissions, PMC is granted a perpetual license to make this article available via PMC and Europe PMC, consistent with existing copyright protections.
The gendered features of adults’ attributions of heroism to themselves and others has received substantially less scholarly attention than the gendered dynamics of media representations of (super)heroes. Utilizing 78 interviews and 569 self-administered questionnaires completed by adults in the United States who were voluntarily making personal protective equipment during the COVID-19 pandemic, we show how respondents effectively deployed popularized assessments of the relative value of gendered labour in the private and public spheres to shift attributions of heroism from themselves to others. Though media portrayals at the outset of the pandemic depicted these volunteers working in their homes as heroes, respondents insisted that the real heroes were those working in the public sphere. Even if media representations increasingly frame women as heroes, these results suggest that the long-standing associations between men and heroism will likely remain in place if feminized labour associated with the private sphere of households remains devalued.
COVID-19
gender
heroes
heroism
labour
work
edited-statecorrected-proof
typesetterts1
==== Body
pmcIntroduction
Heroes are especially important to communities and societies because they enhance the well-being of others and personify who and what are considered worthy of emulation (Kinsella et al., 2015a). Like ‘cultural constellations’ (Dyson, 1996), those labelled heroes provide individuals with ‘moral beacons’ (Porpora, 1996: 210) for how they should look, act and relate to others. In western contexts, heroism is generally associated with voluntarily acting to benefit others despite the physical, emotional and/or social risks incurred by taking such actions (Franco et al., 2018; Kinsella et al., 2015b). In addition to prominent public figures, ordinary individuals are fully capable of being heroes (Franco and Zimbardo, 2006). Heroism can entail dramatic, death-defying feats, yet heroism also regularly involves caring for others and generosity (Becker and Eagly, 2004).
Heroism can also present drawbacks for individuals, communities and societies (Frisk, 2019). Prominent associations between heroism, men and masculinities facilitate the reproduction of inequitable gender orders by justifying men’s power over supposedly weak and vulnerable women (Cocca, 2016; Cree, 2020; Lorber, 2002). Although the gendered features of media representations of (super)heroes have received substantial scholarly attention, significant work remains to be done in examining the links between gender and heroism (Frisk, 2019; Kinsella et al., 2015b). Regarding attributions of heroism, Kinsella et al. (2017: 9) provocatively ask, ‘Do women have to achieve more to be recognized as heroes?’
We answer this question with an emphatic yes, but with a twist. We argue that women and men must achieve more to be recognized as heroes if their heroism is based on feminized labour associated with the private sphere of households. The ideology of separate spheres cleaves space into two distinct categories differentially associated with women and men. The private sphere of homes, care work and unpaid labour is associated with women and femininities. The public sphere of businesses, politics and paid labour is associated with men and masculinities. Following its initial development in the 19th century, how the spheres interface in practice has regularly been reorganized in response to shifting political-economic conditions and policies (Hochschild and Machung, 1989; Laslett and Brenner, 1989). Although women increasingly entered the public sphere in the United States following the passage of civil rights legislation in the 1960s, the ideology of separate spheres is still centrally important to how gendered labour is accomplished and understood in both spheres. Paid work in the public sphere is still strongly associated with men and masculinities, for example, but women in the United States complete more unpaid labour on behalf of volunteer organizations outside their households at least in part because this unpaid work is associated with caring dispositions and skillsets commonly linked to women and feminized labour in the private sphere (Gerstel, 2000; Leap, 2019).
By analysing 78 semi-structured interviews and 569 self-administered questionnaires completed by those who voluntarily made personal protective equipment (PPE) at the outset of the COVID-19 pandemic, we show how individuals who were labelled heroes at the outset of the pandemic shifted attributions of heroism from themselves to others by repeatedly drawing on the ideology of separate spheres. Although our respondents, whom we refer to as ‘makers’, were voluntarily subjecting themselves to a range of risks by producing and distributing PPE to help others during the pandemic, they overwhelmingly rejected the idea that they were heroes even though they were being lauded as such in the news media (e.g. Congresswoman Abigail Spanberger, 2020; Heloise, 2020). Drawing on understandings of heroism that associate it with masculinized work in the public sphere (Featherstone, 1992), makers generally insisted there was nothing heroic about the sewing and 3D printing they were doing in their homes to protect others from a deadly virus. They emphasized that the hero label should be reserved for businesses and individuals working in the public sphere during the pandemic. Echoing popular assessments of gendered labour that have repeatedly elevated the value of masculinized work in the public sphere over feminized work in the private sphere (Daniels, 1987; DeVault, 1991), men and women respondents lionized others’ work in the public sphere while deploying devalued assessments of work completed in their households to shift attributions of heroism from themselves to others.
Other scholars contend the lack of representations of women acting heroically in news and popular media is a key reason why heroism is more commonly attributed to men (Becker and Eagly, 2004; Cocca, 2016; Rankin and Eagly, 2008). We argue that the link between men and heroism is also informed by gendered assessments of labour that obscure the importance and risks of feminized work in the private sphere. Even when the news media were portraying men and women who were voluntarily making and distributing PPE as heroes, those doing this work drew from devalued depictions of feminized work to insist they were not heroes. Disrupting the cultural links between men, masculinities and heroism, we show, can require more than just increased public attention to women acting heroically. It also requires acknowledging the complexities and significance of feminized work completed in the private sphere so that it can be recognized for what it often is – heroism that involves voluntarily placing one’s self at risk to enhance others’ well-being.
We utilize this analysis to synthesize and extend the literatures on pandemic heroism and the gendered contours of heroism. Although scholars have critically engaged with attributions of heroism at the outset of the pandemic (Halberg et al., 2021; Kinsella et al., 2022), this scholarship has not considered either the gendered contours of heroism or heroes beyond the public sphere. Further, although there is an extensive collection of content analyses examining the gendered dynamics of representations of heroes, fewer studies focus on the gendered dynamics of attributing heroism to others. This is especially true in respect to adults’ assessments of heroism (Kinsella et al., 2015b). This deserves closer analysis because attributions of heroism are closely coupled with the reproduction, and potential subversion, of hegemonic gender orders that empower some men at the expense of women and other men (Cocca, 2016; Cree, 2020; Lorber, 2002).
Gendered Representations of Heroes and Gendered Inequalities
Media representations of heroes inform how individuals and groups do gender by legitimating certain behaviours as appropriate for men or women (Coyne et al., 2014; Pennell and Behm-Morawitz, 2015). Representations of heroes have long provided powerful symbolic models against which actual men and women were judged as worthy of status and respect (Connell, 1995; Cree, 2020). Although individuals can challenge the gendered dynamics of such representations (Connell and Messerschmidt, 2005; Dallacqua and Low, 2021), others often still hold them accountable to expectations legitimated by these representations (Marsh, 2000; Moeller, 2011). By engaging with these representations, individuals and groups situate themselves within inequitable gender orders (Cocca, 2016; Dyson, 1996).
Although evidence suggests women have been more likely to act heroically in some situations (Becker and Eagly, 2004), across a range of national contexts heroism is generally associated with men and masculinities (Cocca, 2016; Danilova and Kolpinskaya, 2020; Lorber, 2002). Even the fictional worlds of superheroes, which provide opportunities to celebrate ‘ultimate androgyny’ (Taylor, 2007: 346), regularly feature powerful men rescuing vulnerable damsels in distress (Burch and Johnsen, 2020; Scott, 2013). Female characters such as Wonder Woman provide exceptions to this trend, but they can still facilitate the reproduction of a hegemonic gender order because these heroic women are often dependent on men to validate their (hetero)sexual desirability and vanquish evil (Avery-Natale, 2013; Cocca, 2016; Gilpatric, 2010; Magoulik, 2006). Consequently, such depictions can promote a patriarchal gender order closely linked to heterosexuality even though female heroes may kick some ass (Cocca, 2016; Magoulik, 2006).
Representations of heroes also legitimate hierarchies among men and masculinities. Particular men are represented as heroes worthy of status because of their unique bravery, selflessness or abilities to utilize exceptional skills – often to protect women from men portrayed as exceptionally evil (Cree, 2020; Kelly, 2012; Lorber, 2002). Even when not fighting bad guys, depictions of heroic men equate certain practices and body types with status and respectability (Avery-Natale, 2013; Burch and Johnsen, 2020). In short, representations of heroes have regularly provided exemplary hegemonic masculinities that justify hierarchal gender relations between, and among, men and women (Connell, 1995; Connell and Messerschmidt, 2005).
Gendered Perceptions of Heroes and Work
Although Porpora (1996) encouraged greater attention to adults’ attributions of heroism over two decades ago, studies examining individuals’ assessments of heroism have continued to focus on children and adolescents (Estrada et al., 2015; Gash and Bajd, 2005; Holub et al., 2008). This is especially true in respect to analyses of the gendered features of hero attribution (Danilova and Kolpinskaya, 2020; Kinsella et al., 2015b). The limited studies focusing on adults diverge from and parallel studies on youth. In contrast to children, adults regularly indicate that they either do not have a hero or express reservations about identifying one (Danilova and Kolpinskaya, 2020; Porpora, 1996; Yair et al., 2014). Nevertheless, laboratory studies show that popular depictions of heroes can influence adults’ understandings of themselves (Pennell and Behm-Morawitz, 2015) and whether men or women can be heroic (Rankin and Eagly, 2008). Echoing research on children, women are more likely than men to indicate that a woman is their hero (Danilova and Kolpinskaya, 2020; Donoghue and Tranter, 2018; Rankin and Eagly, 2008). Nevertheless, overall, adults are more likely to identify men as heroes (Danilova and Kolpinskaya, 2020; Donoghue and Tranter, 2018; Rankin and Eagly, 2008) or associate masculinized occupations such as ‘fireman’ with heroism (Keczer et al., 2016). Men and women also sometimes associate different characteristics with heroism. Kinsella et al. (2015b) found that men were more likely than women to associate fearlessness, strength and saving others with heroism.
The gendered dynamics of hero attribution are related to whether and under what circumstances men and women are depicted as heroes in news and popular media (Cocca, 2016). Becker and Eagly (2004) contend that heroism tends to be associated with masculinity because men are generally overrepresented in occupations in the public sphere whose death-defying rescues of people in distress receive consistent public recognition. Consequently, although women also routinely risk their well-being to help others, men are more strongly associated with heroism because men’s heroism gains more public recognition (Becker and Eagly, 2004). A follow-up study by Rankin and Eagly (2008) supports this contention. Respondents were more likely to name men when asked to identify public figures who were heroes. However, when respondents were asked to identify heroes whom they personally knew, they were equally likely to identify men and women as heroes. When presented with a hypothetical rescue scenario, male and female respondents were equally likely to consider the fictional rescuer as heroic whether the rescuer were depicted as a man or woman (Rankin and Eagly, 2008).
Becker and Eagly (2004) are right that men are overrepresented in occupations in the public sphere that are often lauded for heroic work, yet what is unclear is why work commonly associated with women in the private sphere is not considered heroic. Featherstone (1992), for example, explicitly frames feminized work in households as the antithesis of heroism. Presumably, if attributions of heroism are contingent on recognizing that individuals’ actions benefitted others at the risk of harming their own well-being, work associated with the private sphere and femininities must be understood as either failing to benefit others or being risk free. Given that work associated with femininity and the private sphere is widely considered central to caring for others (Laslett and Brenner, 1989), failing to consider this work heroic must be at least partially because it is not considered risky. Featherstone (1992: 165) seemingly confirms this when he notes, ‘A basic contrast then, is that the heroic life is the sphere of danger, violence and the courting of risk whereas everyday life is the sphere of women, reproduction and care.’
Because the meanings associated with heroism, work and gender are social constructs, there is nothing inevitable about the conceptual links between heroism and work in the public sphere associated with men and masculinities. Seale’s (1995, 2002) analyses of terminal illness are illustrative. Seale (1995: 598) frames those who care for those dying from such illnesses as engaging in ‘specifically female heroics’. Further, newspaper coverage of cancer patients deployed gendered representations to frame women as heroes who were uniquely capable of utilizing emotional labour to effectively cope with devastating cancer diagnoses (Seale, 2002). In both cases, Seale (1995, 2002) portrays women as heroic due to their skilled utilization of caring and emotional labour commonly associated with the feminized private sphere.
Heroic Work during the COVID-19 Pandemic
Contrary to the general trend of focusing on the positive features of heroism (Frisk, 2019), initial analyses of the COVID-19 pandemic have emphasized its potential drawbacks (Kinsella et al., 2022). Emphasizing how heroic acts are undertaken within institutions that inequitably distribute risks and social obligations to address them (Halberg et al., 2021), neoliberal policies and ideologies shaped the need for and portrayal of pandemic heroes (Lohmeyer and Taylor, 2021). Labelling first responders, healthcare workers and other ‘essential’ personnel heroes positioned them as individually responsible for addressing problems exacerbated by decades of neoliberal policies (Lohmeyer and Taylor, 2021). Meanwhile, the culpability of state institutions in helping to create conditions of suffering and any responsibility to effectively address them was minimized (Cox, 2020; Halberg et al., 2021; Kinsella and Sumner, 2022; Lohmeyer and Taylor, 2021).
These portrayals also threatened to undermine the well-being of workers, as attributions of heroism made it more difficult to acknowledge workers’ needs for institutional supports, personal protective equipment, adequate pay and boundaries delineating how much they should sacrifice for others who often refused to participate in collective efforts to blunt transmission rates (Cox, 2020). Analyses of frontline workers’ responses to public efforts to label them heroes are especially notable. Danish nurses reported that their experiences did not align with their understandings of heroism (Halberg et al., 2021). They reported feeling overwhelmed, afraid and largely unprepared to work with COVID patients. Like a variety of frontline workers in the UK and Ireland (Kinsella et al., 2022), Danish nurses also reported that being labelled a hero facilitated the expectation that they should endlessly sacrifice their own well-being on behalf of others.
Cox (2020) advocates ceasing labelling healthcare workers heroes, but others conclude that the label can still be useful for describing responses to the pandemic. Instead of portraying pandemic heroes as endless wells of sacrifice and bravery, it is necessary to acknowledge the social institutions that worked to facilitate the need for heroic acts, the negative consequences that can stem from sacrificing on behalf of others and how attributions of heroism can create unrealistic, potentially harmful expectations (Halberg et al., 2021; Kinsella and Sumner, 2022). In short, heroes must be contextualized within institutions and considered fully human.
Methods and Data
This analysis utilizes 569 self-administered online questionnaires and 78 semi-structured phone interviews completed by US adults between July 2020 and January 2021. Both data generation techniques were approved by the authors’ institutional review boards. Inviting PPE makers to complete either a questionnaire or a telephone interview accommodated their time constraints and comfort with different technologies. There was a greater risk of social desirability bias during interviews because respondents were required to talk with a person. When asked whether they considered themselves heroes for making PPE, for example, interview respondents could have been more inclined to indicate that they did not because they wanted to avoid seeming narcissistic to the interviewer. However, responses across data generation techniques were consistent in content on this and all other questions utilized in this analysis. Consequently, we do not suspect social desirability bias had a more substantive impact on interview responses. On average, respondents completed questionnaires in 36 minutes and interviews in 53 minutes.
We used professional, personal and virtual networks to distribute flyers and invitations to participate in the study. Paralleling prior analyses of disaster responses (Penta et al., 2020), social media platforms such as Facebook, Instagram and Reddit were especially useful for identifying potential participants because makers often used social media to organize PPE production. Because individuals who acquired raw materials and organized distribution networks were just as important as makers in getting PPE into the hands of users, they were included in the study.
As Table 1 illustrates, women, whites and relatively well-educated individuals are overrepresented in the sample as compared with the population of the United States. Other analyses that distributed online calls to participate during the pandemic reported similarly skewed samples (Craig and Churchill, 2021; Friedman et al., 2021). The dynamics of making and volunteering in the USA could also help explain the sample composition. Sewing and 3D printing requires specialized equipment that presents classed barriers to participation (Stalp, 2015), and the gendered and racialized dynamics of volunteerism in the USA have often encouraged white women to engage in volunteering at higher rates than other groups (Pham, 2020). Although we purposefully contacted maker groups comprised primarily of people of colour to try to further diversify our sample, non-response bias could have also skewed the sample.
Table 1. Demographic composition of interview and questionnaire respondents.
Interviews Questionnaires
Age (Mean)a 49.48 50 to 59
Gender (%)
Man 20.51 8.44
Woman 79.49 90.69
Other 0 0.88
Race (%)b
American Indian 1.28 1.05
Asian 3.85 3.69
Black or African American 1.28 1.05
White 88.46 95.08
Other 7.69 1.76
Highest educational attainment (%)c
Grade or middle school 0.18
High school 12.48
Associate’s degree 12.3
Bachelor’s degree 33.39
Master’s degree 26.19
Professional degree 6.15
Doctorate degree 9.31
Hispanic or Latino, any race (%) 6.41 2.81
US census regions (%)
Midwest 32.05 23.73
Northeast 7.69 13.18
South 41.03 37.79
West 19.23 25.31
N 78 569
Notes: aAge was structured as an ordinal variable on questionnaires and an interval variable during interviews.
b Race was a choose all that apply item on questionnaires and an open question in interviews.
c Educational attainment data were only obtained from questionnaire respondents.
We cannot generalize our findings beyond our sample, but our dataset does have some notable strengths. The regional composition of questionnaire respondents is within +/– 5% of population estimates of the four Census Bureau regions of the USA (US Census Bureau, 2022). Our data are also comprised solely of individuals’ first-hand assessments of their responses to the pandemic as it unfolded. As a result, it is an emotionally evocative dataset with detailed descriptions of makers’ efforts to survive a worsening disaster.
Following analyses of hero attribution that advocate utilizing open-ended questions to assess individuals’ attributions of heroism (Danilova and Kolpinskaya, 2020; Donoghue and Tranter, 2018), we focus primarily on data generated through open questions. In addition to asking makers to detail the who, what, where, why and how of PPE production, we asked respondents whether they believed they or anyone else was a hero for their responses to the pandemic. Following Charmaz (2003), we utilized an open, iterative approach to data analysis. We first read and re-read responses to identify themes respondents regularly invoked. After identifying and analysing key themes such as ‘risks incurred through making’ and ‘attributing heroism to PPE group organizers’ in greater detail, we determined that the ideology of separate spheres was centrally important to makers’ attributions of heroism because it effectively integrated the themes that emerged during the initial stages of data analysis.
Respondents are identified by their gender and region in the ensuing analysis. Each quote comes from a different maker. We begin by focusing on makers’ descriptions of fabricating PPE to highlight that their actions were heroic. We then illustrate how makers overwhelmingly rejected the suggestion that their work was heroic even though they incurred a multitude of risks to enhance others’ safety. Finally, we illustrate how makers deployed the gendered ideology of separate spheres to shift attributions of heroism from their work in the private sphere to other individuals and entities working in the public sphere.
Heroic PPE Production: Voluntarily Incurring Risks to Bolster Others’ Safety
Supplies of PPE were quickly exhausted in the USA in early 2020. After decades of neoliberal reforms had weakened public health infrastructure and encouraged for-profit production facilities to be moved out of the United States, neither state institutions nor private businesses could provide adequate PPE supplies (Leap et al., 2022a, 2022b). Individuals and civic organizations were encouraged to voluntarily produce and distribute substantial amounts of masks, face shields, hand sanitizer and other protective equipment. Across the USA, makers who answered this call were publicly lauded as heroes by news media and state officials. A June 2020 article in the Washington Post portrayed makers as heroes (Heloise, 2020) and US House Representative Abigail Spanberger officially recognized makers as heroes (Congresswoman Abigail Spanberger, 2020), for example.
Nevertheless, it may not be readily apparent that makers met the definition of heroes who were voluntarily risking their well-being to assist others. Pham (2020), for example, echoes Featherstone (1992) when she contrasts the working conditions of low-paid garment workers with the supposedly safe working conditions enjoyed by volunteer PPE makers. She notes, ‘[Garment workers are] not making masks in the safety and comfort of their homes’ (Pham, 2020: 322). However, makers often exposed themselves to a multitude of physical, emotional, financial and social risks to get PPE to those who desperately needed it.
Repeatedly, makers indicated that they incurred aches, pains and repetitive motion injuries from spending hours on end producing PPE. One interviewee had even gone to the hospital after her legs became swollen following an especially intense mask production session. Makers also placed themselves at risk of infection as they sought out materials for production and then distributed finished PPE. In response to a questionnaire item that asked makers to detail any drawbacks associated with producing PPE, a Midwestern woman remarked, ‘Some people do not realize the time it takes and the risks I took with my own health when I stood in line for hours to get fabric. The assumptions that it was easy made me feel undervalued.’
Beyond physical risks, makers regularly indicated that it was emotionally overwhelming to know that others might die if they did not produce enough PPE. This caused considerable emotional turmoil for a man in the Western United States who was helping organize a network of makers that was producing tens of thousands of pieces of PPE. During his interview, he explained: For several months, it was just making. That’s all I did. I woke up at 6 every morning, started working on making stuff, trying to organize, call people, try and track down materials to get stuff done. I would work until 10 o’clock at night and got to a point where I was like, super stressed out, I was having mental breaks. I was losing hair, and I ended up having to kind of step away for a little bit. And it was, it was tough. But it’s one of those things where, you know, you’re helping people, you know, it’s really, really important. You know, it’s not like you’re shipping teddy bears, you know, we’re shipping protective equipment to help people. It’s really difficult to walk away.
Makers also subjected themselves to financial risks. In materials alone, nearly two-thirds (65.4%) of questionnaire respondents reported spending at least US$100 for supplies to fabricate PPE. More than a quarter (26.9%) of respondents reported spending over US$350, and it was not uncommon for makers to indicate that they had spent well over US$1000 on materials. Makers were sometimes reimbursed for some of their costs, but, most commonly, they were not. Of those that reported spending over US$350 on materials, 62.1% indicated that they had not received any reimbursements. Makers also ran the risk of getting stuck with substantial bills when plans for reimbursement unexpectedly failed. One Southern man reported losing nearly US$1000 after not being reimbursed. In some cases, these expenses undermined makers’ financial well-being. Responding to the questionnaire item concerning drawbacks associated with PPE production, a Northeastern woman noted: It was a financial strain. I now have a lot of materials that I will continue to use, of course, but I definitely spent more out of pocket than I was reimbursed for. A lot of these purchases were made using my unemployment cheques.
Finally, makers routinely indicated that fabricating PPE facilitated interpersonal conflicts in their households and communities. Numerous women noted tensions with others in their households who did not appreciate the amount of space PPE production required or that they had shifted some of their time from domestic chores and childcare to produce PPE. Makers also risked social standing in their communities. Respondents reported being judged negatively by others who did not like the aesthetics of their PPE. Others reported being ‘ostracized’ from collective group efforts. Some even experienced abuse from those opposed to mask wearing. Detailing the drawbacks of PPE production on the questionnaire, one Southern woman explained: ‘Anti-maskers are quite nasty and rude and have been going out of their way to make my life difficult. We had to anonymously coordinate deliveries and donations because people were so rude to our drivers and seamstresses.’ Similarly, a Southern man remarked, ‘The Karens and anti-maskers are vocal and abundant. It’s disheartening.’
Voluntary PPE production has been characterized as largely risk free (Pham, 2020), but makers provided far more complex characterizations of their work. Although our respondents were relatively privileged in terms of socio-economic status and race, they described a multitude of physical, emotional, financial and social risks associated with PPE production. Makers made significant sacrifices and incurred a multitude of risks to help others during the pandemic.
I’m Not a Hero: Makers’ Denials of Their Heroism
Makers were rightly recognized as heroes for their work at the outset of the pandemic, but did makers consider themselves heroes? Among questionnaire respondents, exceptionally few men (6.3%) and women (5.8%) indicated that they considered themselves heroes for producing and distributing PPE. During interviews, makers also overwhelmingly rejected the suggestion that they were heroes. Some even laughed when asked if they believed they were heroes, as if equating heroism with their efforts to voluntarily save others’ lives verged on humorous absurdity. One Southern woman’s reaction is notable. After being asked if she considered herself a hero, she immediately guffawed before exclaiming, ‘Nooo. No, that’s ridiculous.’
Makers provided a variety of reasons for rejecting the hero label. Some did not want to seem narcissistic. A Midwestern woman noted, ‘I never want to glorify myself, but I am willing to call others heroes when they are willing to sacrifice for others.’ Reflective of how incurring risks and assisting others are both centrally important to attributions of heroism (Kinsella et al., 2015b), others indicated that they had not incurred enough risks or assisted enough people to be considered heroes. After being asked during her interview whether she considered herself a hero for fabricating and donating over 600 masks, one Southern woman explained, ‘To me, [heroes are] someone who is like literally putting their life in danger for other people. I’m sitting at my sewing machine getting a little carpal tunnel.’ Echoing the association between pandemic heroism and limitless sacrifice (Halberg et al., 2021), makers routinely set exceedingly lofty thresholds of assistance rendered to others for a maker to be considered heroic. Describing when he considered other makers heroes, one Northeastern man noted, ‘People who have been able or willing to devote every waking moment to it.’ In response to the same questionnaire item, a Southern woman noted, ‘I have known individuals who have donated 100% of their time to making masks nonstop.’
Reflective of some adults’ hesitancy to venerate others through attributions of heroism (Yair et al., 2014), makers also sometimes expressed ambivalence about labelling anyone a hero because lauding individuals distracted from the need for collective responses to the pandemic. A Midwestern man’s sentiment is notable: ‘Heroes’ is a hard word for me. It feels very ego-focused. We need heroes because we need to point to *individuals* who have gone above and beyond, who have made heroic personal sacrifices, done brave and helpful things. But . . . We all have a role to play in dealing with this pandemic. Some of us have already established roles that make their contributions more risky. But ultimately, we are all at risk. We all have to make sacrifices. We all have to care for others. And ‘hero’ misses the point.
A Northeastern woman echoed this while linking the need for pandemic heroism to failed state responses. Reminiscent of Danish nurses’ assessments of heroism (Halberg et al., 2021), while answering the questionnaire item regarding who she considered heroes, she wrote: The term hero is very complicated. The hero rhetoric makes martyrs out of [frontline] workers when it would not have to come to that if the government ensured everyone had the proper protective gear and/or a liveable income under safe conditions.
Finally, and echoing sentiments expressed by frontline workers in the UK and Ireland (Kinsella et al., 2022), makers often indicated that they were not heroes because they were unremarkable and simply doing what they should. After being asked whether she considered herself a hero for making and distributing 400 masks, one Southern woman replied: I don’t think so. I don’t know. Now, I guess I see it as, this is, you know, like people in WWII or WWI or the Revolutionary War. They were born for those moments, and they lived up to them. And this is just me living up to what my time and day is asking of me. It’s just stepping up and using talents that I know God has given me to serve and to help. Don’t think I’d say hero, no. (laughing) But just someone who saw a need and stepped up.
When asked in a follow-up question whether she considered soldiers, nurses, doctors and teachers who stepped up during the pandemic heroes, she replied, simply, ‘Yes.’ As we illustrate in the following section, attributing heroism to those working in the public sphere was a recurring way that makers denied their own heroism while emphasizing the heroism of others.
They’re the Heroes: Makers’ Descriptions of Others’ Heroism
Although very few makers personally accepted the hero label, roughly 75% of men and women questionnaire respondents indicated that at least some of the other makers fabricating PPE were heroes. Further, over 90% of men and women indicated that there were individuals beyond makers who should be considered heroes due to their responses to the pandemic. Unlike studies that find men and women tend to associate different characteristics and occupations with heroism (Danilova and Kolpinskaya, 2020; Kinsella et al., 2015b), men and women in our sample generally utilized the same standards to assess heroism. Nevertheless, gender was still centrally important to how makers attributed heroism because both men and women consistently drew on the gendered ideology of separate spheres to delineate who was a hero. By drawing on popularized assessments of gendered labour in the public and private spheres, makers shifted attributions of heroism from themselves to others.
Businesses that transitioned into PPE production, makers who prioritized PPE over paid employment, those involved in community organizing and ‘frontline’ or ‘essential’ workers all featured prominently in makers’ delineations of heroism. These depictions of heroism were sometimes deployed in isolation, yet they were also regularly interlaced to associate the public sphere with risks, sacrifice and, subsequently, heroism. In contrast, and as is common in respect to feminized work associated with the private sphere (Daminger, 2019; Daniels, 1987; DeVault, 1991), the complexities and significance of work completed by makers in their homes was dismissed.
Businesses operating in the public sphere that transitioned into PPE production were routinely lauded as heroic. When describing whom she considered a hero, a Southern woman wrote, ‘The bourbon industry in KY stopped making bourbon and made hand sanitizer! General Motors and Ford shut down to make ventilators. Tech Schools started making face shields with their 3D printers. Those people are heroes.’ Respondents regularly stressed that businesses were heroic because they appeared to be jeopardizing profits by engaging in PPE production. A woman in the Northeast noted, ‘Companies that quickly switched to supply and distribute what was needed. Many of them risked their profit that allows them to exist as a business.’ Likewise, a Midwestern woman wrote, ‘Businesses that are sacrificing profit for the greater good.’ Paralleling this focus on businesses risking profits, respondents were sometimes only willing to attribute heroism to other makers when they risked their paid occupations in the public sphere by devoting time to PPE production. Explaining when he considered other makers heroes, a Southern man explained, ‘People giving up jobs to organize the large donation drives and coordinating directly with the hospitals/organizations.’
As the preceding man’s response alludes to, community organizing in the public sphere also featured prominently in makers’ attributions of heroism. Describing when she considered other makers heroes, a Midwestern woman wrote: The [Midwestern] Mask drive organizers, they have been on the cutting edge of making this happen in our community and are continuing with not only making masks but setting up stations around the community for donation and receiving (free) masks.
Similarly, a Western woman noted: The folks who are organizing the collection and distribution of the items. Those of us making them are doing what we can, which is really just doing what we should. I don’t have the skill set to do community organization – those that do this are the heroes.
In some cases, makers linked community organizing to enhanced risks. In response to the same questionnaire item, a Northeastern woman wrote, ‘Those who go out and interact with the community and put themselves at greater risk.’
Paralleling public attributions of heroism during the initial phases of the pandemic that often focused on workers in the public sphere (Cox, 2020; Lohmeyer and Taylor, 2021), makers also repeatedly referenced ‘frontline’ and ‘essential’ workers when describing whom they considered heroes. Reflective of the pandemic hierarchy of heroes that placed medical professionals at the very top (Kinsella et al., 2022), ‘healthcare workers’ were referenced by an overwhelming majority of makers in questionnaires and interviews. A Western woman who produced and distributed 11,000 masks is illustrative. After being asked whether makers could be considered heroes, she shifted the conversation directly to healthcare workers: You know, my thought was that the real heroes are the healthcare workers. They are the ones who really sacrificed themselves. And to me, you know, making masks is part of the effort to be worthy of their service. You know, to deserve their sacrifice. Yeah, the real heroes are really the healthcare people.
Less common, yet still noted by a sizable share of makers, were delivery drivers, first responders, grocery store staff, janitors, local governmental officials and teachers.
Although makers repeatedly linked heroism to labour in the public sphere, makers generally did not link heroism to labour in the private sphere when describing other pandemic heroes. References to risks were often centrally important to makers’ attributions of heroism, but makers did not associate risks with the private sphere when describing pandemic heroes. Makers did not indicate that they considered other makers heroes because producing masks risked physical injury or intra-household tensions, for example. Further, in only a handful of instances did respondents portray work associated with the private sphere as involving significant sacrifices worthy of commendation. One Southern woman provided an exception. When describing makers she considered heroes, she wrote, ‘Possibly people who have made PPE full time – turning their home or workspace into more of a factory.’ However, even her invocation of ‘possibly’ seemed to indicate some trepidation with describing such makers as heroes. Further, linking heroism to homes transformed ‘into more of a factory’ still emphasized the links between heroism and the public sphere because only those homes that were made more reminiscent of a workplace in the public sphere were indicative of heroism.
The substantial risks and sacrifices that warranted attributions of heroism were overwhelmingly associated with work in the public sphere. Not work in the private sphere, which was often feminized and portrayed as relatively insignificant. The following passages are illustrative. When describing whom she considered a pandemic hero, a Southern woman replied, ‘People working the actual front lines are heroes. Not those of us sewing at a sewing machine safely in our homes.’ In response to the same question, a different Southern woman remarked, ‘Those that were able to invent things to help. Organize mass groups to make PPE. I was just a wife at home making as many as I could to help.’
Discussion and Conclusion: Gendered Work and Attributions of Heroism
Paralleling previous considerations of heroism (Kinsella et al., 2015b), makers repeatedly referenced risks and providing benefits to others when assessing their and others’ heroism. In contrast to prior analyses (Danilova and Kolpinskaya, 2020; Kinsella et al., 2015b), makers generally deployed the same criteria in their assessments of heroism regardless of whether they were men or women. Nevertheless, gender was still centrally important to respondents’ attributions of heroism. Makers repeatedly associated heroism with work in the masculinized public sphere. Work in the feminized private sphere was not heroic, according to respondents, because it was unremarkable work that did not involve significant risks or sacrifices.
Whether and how men and women are represented as heroes in news and popular media is especially important to attributions of heroism (Cocca, 2016). Becker and Eagly (2004) and Rankin and Eagly (2008) contend that men are more readily associated with heroism because they are overrepresented in occupations that consistently receive public recognition for heroic acts. Seemingly, the association between men and heroism is largely due to the relative visibility of men’s heroism in the public sphere and invisibility of women’s heroism in the private sphere. Our findings indicate that even when men and women are publicly commended for heroic acts in the private sphere, popularized assessments of gendered labour that devalue feminized labour in the private sphere can be deployed to delineate whether those publicly lauded for such heroism are actually worthy of such accolades.
In the case of PPE makers, even when individuals were intimately aware of the benefits and risks associated with their work, they still overwhelmingly rejected portrayals celebrating their heroism. By downplaying the risks and significance of voluntary work to produce PPE in the feminized private sphere, they shifted attributions of heroism from themselves to those working in the public sphere during the pandemic. This was not simply because respondents wanted to avoid appearing narcissistic. Among the hundreds of responses we received, in only a very limited handful of cases did makers associate others’ heroism with labour in the feminized private sphere. If makers were just trying to avoid appearing vain and the relative value of gendered labour in the public and private spheres was not central to their attributions of heroism, respondents would have been just as likely to associate others’ heroism with work in the public and private spheres.
Work completed in the private sphere by both men and women is not simple, insignificant or risk free. Although this work is regularly taken for granted, it requires skilled physical, emotional and cognitive labour (Daminger, 2019; DeVault, 1991; Leap et al., 2022a). As makers’ descriptions of their work emphasize, this work is also associated with a range of risks. Nevertheless, so long as the significance and risks of work associated with the private sphere continue to be downplayed, it seems reasonable to conclude that the heroism of those doing this work will continue to be obscured even when it gains public recognition.
What is required to complicate the long-standing link between men, masculinities and heroism is both an expansion of the public recognition of heroic women and an expansion of what work is considered worthy of respect and commendation. Only then can the true scope of heroism be acknowledged. Without acknowledging the true value and complexities of work associated with the private sphere, it seems unlikely we will ever fully recognize the heroic men and women who voluntarily risk their own well-being to help others through such work.
Like heroism in the public sphere (Halberg et al., 2021; Kinsella et al., 2022), heroism in the private sphere cannot be disassociated from social institutions or the personal costs that are often incurred by heroes. Labour in the private sphere is also informed by state, market and cultural institutions that encourage some to sacrifice their well-being on behalf of others (Hochschild and Machung, 1989; Laslett and Brenner, 1989; Leap et al., 2022a). PPE makers would not have needed to place their health, finances and social standing at risk if PPE stockpiles and manufacturing facilities in the United States had not been hallowed out by decades of neoliberal political-economic restructuring (Leap et al., 2022a, 2022b), for example. Feminized work in the private sphere can be heroic, and this heroism is just as political as heroism in the public sphere.
We want to thank all of the makers who graciously agreed to participate in this study while also voluntarily fabricating PPE to heroically assist their families, friends and communities amid a global pandemic. We are also very appreciative for the thoughtful feedback provided by the reviewers and editors at Sociology.
Braden Leap is Associate Professor of Sociology at Mississippi State University. He studies the gendered dynamics of socio-ecological transformations, disruptions and disasters.
Kimberly Kelly is Associate Professor of Sociology and Director of Gender Studies at Mississippi State University. Kelly studies gender and collective behaviour, with an emphasis on religion, politics and abortion.
Marybeth C Stalp is Professor of Sociology at Northern Iowa University. Stalp studies how and where gender, culture and leisure intersect using qualitative methods, with a particular focus on women’s creativity and leisure pursuits through the life course.
Funding: The authors received no financial support for the research, authorship and/or publication of this article.
==== Refs
References
Avery-Natale E (2013) An analysis of embodiment among six superheroes in DC comics. Social Thought and Research 32 : 71–106.
Becker SW Eagly AH (2004) The heroism of women and men. American Psychologist 59 (3 ): 163–178.15222859
Burch RL Johnsen L (2020) Captain Dorito and the bombshell: Supernormal stimuli in comics and film. Evolutionary Behavioral Sciences 14 (2 ): 115–131.
Charmaz K (2003) Grounded theory: Objectivist and constructivist methods. In: Denzin NK Lincoln YS (eds) Strategies in Qualitative Inquiry. Los Angeles, CA: SAGE, 249–291.
Cocca C (2016) Superwomen: Gender, Power, and Representation. New York, NY: Bloomsbury.
Congresswoman Abigail Spanberger (2020) Spanberger Honors dedication, generosity, and service of six more ‘VA-07 Heroes’. Available at: https://spanberger.house.gov/posts/spanberger-honors-dedication-generosity-and-service-of-six-more-va-07-heroes (accessed 5 October 2022).
Connell R (1995) Masculinities. Berkeley, CA: University of California Press.
Connell R Messerschmidt JW (2005) Hegemonic masculinity: Rethinking the concept. Gender & Society 19 (6 ): 829–859.
Cox CL (2020) ‘Healthcare heroes’: Problems with media focus on heroism from healthcare workers during the COVID-19 pandemic. Journal of Medical Ethics 46 (8 ): 510–513.32546658
Coyne SM Linder JR Rasmussen EE , et al (2014) It’s a bird! It’s a plane! It’s a gender stereotype! Longitudinal associations between superhero viewing and gender stereotyped play. Sex Roles 70 : 416–430.
Craig L Churchill B (2021) Unpaid work and care during COVID-19: Subjective experiences of same-sex couples and single mothers in Australia. Gender & Society 35 (2 ): 233–243.
Cree A (2020) People want to see tears’: Military heroes and the ‘constant Penelope’ of the military wives choir. Gender, Place & Culture 27 (2 ): 218–238.
Dallacqua AK Low DE (2021) Cupcakes and beefcakes: Students’ readings of gender in superhero texts. Gender and Education 33 (1 ): 68–85.
Daminger A (2019) The cognitive dimension of household labor. American Sociological Review 84 (4 ): 609–633.
Daniels AK (1987) Invisible work. Social Problems 34 (5 ): 403–415.
Danilova N Kolpinskaya E (2020) The politics of heroes through the prism of popular heroism. British Politics 15 (2 ): 178–200.
DeVault ML (1991) Feeding the Family. Chicago, IL: University of Chicago Press.
Donoghue J Tranter B (2018) On Bradman’s bat: Australian sporting heroes. National Identities 20 (2 ): 143–156.
Dyson AH (1996) Cultural constellations and childhood identities: On Greek gods, cartoon heroes, and the social lives of schoolchildren. Harvard Educational Review 66 (3 ): 471–495.
Estrada JD Garcia-Ael C Mortelli JL (2015) Gender differences in adolescents’ choice of heroes and admired adults in five countries. Gender and Education 27 (1 ): 69–87.
Featherstone M (1992) The heroic life and everyday life. Theory, Culture & Society 9 : 159–182.
Franco ZE Zimbardo PG (2006) The banality of heroism. Greater Good 3 (2 ): 30–35.
Franco ZE Allison ST Kinsella EL , et al (2018) Heroism research: A review of theories, methods, challenges, and trends. Journal of Humanistic Psychology 58 (4 ): 382–396.
Friedman M Lichtfuss KK Martingnetti L , et al (2021) ‘It feels a bit like drowning’: Expectations and experiences of motherhood during COVID-19. Atlantis 42 (1 ): 47–57.
Frisk K (2019) What makes a hero? Theorising the social structuring of heroism. Sociology 53 (1 ): 87–103.
Gash H Bajd B (2005) Young people’s heroes in Ireland and Slovenia. The Irish Journal of Psychology 26 (3–4 ): 137–148.
Gerstel N (2000) The third shift: Gender and care work outside the home. Qualitative Sociology 23 (4 ): 467–483.
Gilpatric K (2010) Violent female action characters in contemporary American cinema. Sex Roles 62 : 734–746.
Halberg N Jensen PS Larsen TS (2021) We are not heroes – the flipside of the hero narrative amidst the COVID-19 pandemic: A Danish hospital ethnography. Journal of Advanced Nursing 77 (5 ): 2429–2436.33616210
Heloise (2020) Hints from Heloise: Unsung heroes of the coronavirus pandemic. Washington Post, 1 June. Available at: washingtonpost.com (accessed 5 October 2022).
Hochschild A Machung A (1989) The Second Shift. New York, NY: Viking Penguin.
Holub SC Tisak MS Mullins D (2008) Gender differences in children’s hero attributions: Personal hero choices and evaluations of typical male and female heroes. Sex Roles 58 (7 ): 567–578.
Keczer Z File B Orosz G , et al (2016) Social representations of hero and everyday hero: A network study from representative samples. PLoS One 11 (8 ): e0159354.
Kelly J (2012) Popular culture, sport and the ‘hero’-fication of British militarism. Sociology 47 (4 ): 722–738.
Kinsella EL Hughes S Lemon S , et al (2022) ‘We shouldn’t waste a good crisis’: The lived experience of working on the frontline through the first surge (and beyond) of COVID-19 in the UK and Ireland. Psychology & Health 37 (2 ): 151–177.34187244
Kinsella EL Ritchie TD Igou ER (2015a) Lay perspectives on the social and psychological functions of heroes. Frontiers in Psychology 6 : 130.25741302
Kinsella EL Ritchie TD Igou ER (2015b) Zeroing in on heroes: A Prototype analysis of hero features. Journal of Personality and Social Psychology 108 (1 ): 114–127.25603370
Kinsella EL Ritchie TD Igou ER (2017) On the bravery and courage of heroes: Considering gender. Heroism Science 2 (1 ): 1–14.
Kinsella EL Sumner R (2022) High ideals: The misappropriation and reappropriation of the heroic label in the midst of a global pandemic. Journal of Medical Ethics 48 (3 ): 198–199.33593874
Laslett B Brenner J (1989) Gender and social reproduction: Historical perspectives. Annual Review of Sociology 15 : 381–404.
Leap BT (2019) Gone Goose: The Remaking of an American Town in the Age of Climate Change. Philadelphia, PA: Temple University Press.
Leap B Kelly K Stalp MC (2022a) Choreographing social reproduction: Making personal protective equipment and gender during a neoliberal pandemic. Gender, Work & Organization 29 (3 ): 758–777.
Leap B Stalp MC Kelly K (2022b) Raging against the ‘Neoliberal Hellscape’: Anger, pride, and ambivalence in civil society responses to the COVID-19 Pandemic in the USA. Antipode 54 (4 ): 1166–1187.35602931
Lohmeyer BA Taylor N (2021) War, heroes and sacrifice: Masking neoliberal violence during the COVID-19 pandemic. Critical Sociology 47 (4–5 ): 625–639.
Lorber J (2002) Heroes, Warriors, and Burqas: A feminist sociologist’s reflections on September 11. Sociological Forum 17 (3 ): 377–396.
Magoulik M (2006) Frustrating female heroism: Mixed messages in Xena, Nikita, and Buffy. The Journal of Popular Culture 39 (5 ): 729–755.
Marsh J (2000) But I want to fly too! Girls and superhero play in the infant classroom. Gender and Education 12 (2 ): 209–220.
Moeller RA (2011) ‘Aren’t these boy books?’: High school students’ reading of gender in graphic novels. Journal of Adolescent and Adult Literacy 54 (7 ): 476–484.
Pennell H Behm-Morawitz E (2015) The empowering (super) heroine? The effects of sexualized female characters in superhero films on women. Sex Roles 72 : 211–220.
Penta A Wachtendorf T Nelan M (2020) Disaster relief as social action: A Weberian look at postdisaster donation behavior. Sociological Forum 35 (1 ): 145–166.
Pham M (2020) ‘How to make a mask’: Quarantine feminism and global supply chains. Feminist Studies 46 (2 ): 316–326.
Porpora DV (1996) Personal heroes, religion, and transcendental metanarratives. Sociological Forum 11 (2 ): 209–299.
Rankin LE Eagly AH (2008) His heroism hailed and hers hidden? Women, men, and the social construction of heroism. Psychology of Women Quarterly 32 : 414–422.
Scott S (2013) Fangirls in refrigerators: The politics of (in)visibility in comic book culture. Transformative Works and Cultures 13 : 1–20.
Seale C (1995) Heroic death. Sociology 29 (4 ): 597–613.
Seale C (2002) Cancer heroics: A study of news reports with particular reference to gender. Sociology 36 (1 ): 107–126.
Stalp MC (2015) Girls just want to have fun (too): Complicating the study of femininity and women’s leisure. Sociology Compass 9 (4 ): 261–271.
Taylor A (2007) ‘He’s gotta be strong, and he’s gotta be fast, and he’s gotta be larger than life’: Investigating the engendered superhero body. The Journal of Popular Culture 40 (2 ): 344–360.
US Census Bureau (2022) U.S. Census Bureau United States population growth by region. Available at: https://www.census.gov/popclock/print.php?component=growth&image=//www.census.gov/popclock/share/images/growth_1561939200.png (accessed 5 October 2022).
Yair G Girsh Y Alayan S , et al (2014) ‘We don’t need another hero’: Heroes and role models in Germany and Israel. Comparative Education Review 58 (2 ): 269–295.
| 0 | PMC9705506 | NO-CC CODE | 2022-12-01 23:19:37 | no | Sociology. 2022 Nov 25;:00380385221136035 | utf-8 | Sociology | 2,022 | 10.1177/00380385221136035 | oa_other |
==== Front
J Psychiatr Res
J Psychiatr Res
Journal of Psychiatric Research
0022-3956
1879-1379
Elsevier Ltd.
S0022-3956(22)00661-6
10.1016/j.jpsychires.2022.11.031
Article
Is there an association between depression, anxiety disorders and COVID-19 severity and mortality? A multicenter retrospective cohort study conducted in 50 hospitals in Germany
Kostev Karel a∗1
Hagemann-Goebel Marion b1
Gessler Nele cde
Wohlmuth Peter d
Feldhege Johannes d
Arnold Dirk f
Jacob Louis gh
Gunawardene Melanie c
Hölting Thomas i
Koyanagi Ai gjk
Schreiber Ruediger l
Smith Lee m
Sheikhzadeh Sara n
Wollmer Marc Axel eo
a Epidemiology, IQVIA, Frankfurt, Germany
b Department of Behavioral Medicine, Asklepios Hospital Nord-Heidberg, Hamburg, Germany
c Department of Cardiology and Internal Intensive Care Medicine, Asklepios Hospital St. Georg, Hamburg, Germany
d Asklepios Proresearch, Research Institute, Hamburg, Germany
e Faculty of Medicine, Semmelweis University, Budapest, Hungary
f Department of Hematology, Oncology, Palliative Care Medicine and Rheumatology, Asklepios Hospital Altona, Hamburg, Germany
g Research and Development Unit, Parc Sanitari Sant Joan de Déu, Dr. Antoni Pujadas, 42, Sant Boi de Llobregat, Barcelona, Spain
h Faculty of Medicine, University of Versailles Saint-Quentin-en-Yvelines, Montigny-le-Bretonneux, France
i Department of Internal Medicine and Cardiology, Asklepios Hospital Wandsbek, Hamburg, Germany
j Centro de Investigación Biomédica en Red de Salud Mental (CIBERSAM), ISCIII, Madrid, Spain
k ICREA, Pg. Lluis Companys 23, 08010, Barcelona, Spain
l Department of Anesthesiology and Intensive care medicine, Asklepios Hospital West, Hamburg, Germany
m Centre for Health, Performance, and Wellbeing, Anglia Ruskin University, Cambridge, CB1 1PT, UK
n Asklepios hospitals GmbH & Co. KGaA, Hamburg, Germany
o Asklepios Klinik Nord Ochsenzoll, Asklepios Campus Hamburg, Germany
∗ Corresponding author. Epidemiology IQVIA, Unterschweinstiege 2–14, 60549, Frankfurt am Main, Germany.
1 These authors contributed equally to this work.
29 11 2022
1 2023
29 11 2022
157 192196
19 10 2022
16 11 2022
28 11 2022
© 2022 Elsevier Ltd. All rights reserved.
2022
Elsevier Ltd
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Background
The aim of this retrospective cohort study was to investigate associations between depression and anxiety disorder and the risk of COVID-19 severity and mortality in patients treated in large hospitals in Germany.
Methods
This retrospective study was based on anonymized electronic medical data from 50 public healthcare service hospitals across Germany. Multivariable logistic regression models were used to study associations between depression, anxiety and mechanical ventilation and mortality due to COVID adjusted for age, sex, time of COVID-19 diagnosis, and pre-defined co-diagnoses.
Results
Of 28,311 patients diagnosed with COVID-19, 1970 (6.9%) had a diagnosis of depression and 369 (1.3%) had a diagnosis of anxiety disorder prior to contracting COVID-19. While multivariable logistic regression models did not indicate any association between depression diagnosis and the risk of mechanical ventilation, depression was associated with a decreased risk of mortality (OR: 0.71; 95% CI: 0.53–0.94). There was no association between anxiety disorders and risk of mortality, but there was a strong positive association between anxiety disorders and the risk of mechanical ventilation (OR: 2.04; 95% CI: 1.35–3.10).
Conclusion
In the present study, depression and anxiety disorder diagnoses were not associated with increased COVID-19 mortality. Anxiety disorder was strongly associated with an increased risk of mechanical ventilation. Further studies are needed to clarify how depression and anxiety disorders may influence COVID-19 severity and mortality.
Keywords
COVID-19
Depression
Anxiety
Elderly
Hospital
Mortality
==== Body
pmc1 Introduction
As of September 7, 2022, more than 600 million people globally have contracted COVID-19, while the number of related deaths has exceeded six million (World Health Organization 2022). In Germany, the respective figures are 32 million positive cases and 148,000 deaths (World Health Organization 2021).
Previous research has shown that people with psychiatric disorders not only have an increased risk of mortality in general (Plana-Ripoll et al., 2019) but also an increased risk of COVID-19 mortality (Vai et al., 2021; Nemani et al., 2021). However, there were relevant discrepancies between these studies with regard to the types of mental disorder examined.
A meta-analysis of 21 studies including >90 million individuals found significantly higher odds of COVID-19 mortality (OR, 1.51) in persons with preexisting mood disorders (Ceban et al., 2021). Another meta-analysis based on 23 studies comprising 1.5 million patients with COVID-19 and mood disorders found a significant association between mood disorders and COVID-19 mortality (OR: 1·99) but not anxiety disorders (OR: 1·07)) (Vai et al., 2021). However, neither of these meta-analyses distinguished between depression, mania, and bipolar disorders.
In a study conducted in California, US, Azar et al. reported significantly higher odds of COVID-19 mortality among those with comorbid depression compared to those without (OR, 2.64) for COVID-19-positive patients with comorbid depression (Azar et al., 2020). In a retrospective case-control study by Douville et al., also conducted in the US, no significant associations were identified between depression and COVID-19-related mechanical ventilation or mortality among hospitalized patients, although the odds of mortality were increased in patients with depression (1.71; 95% CI: 0.87–3.37). However, this association was not significant, likely owing to small patient numbers (n = 398).
In another study by Poblador-Plou et al. including 4412 individuals from Spain, there was no association between anxiety disorders and COVID-19 mortality (Poblador-Plou et al., 2020).
Teixeira et al. analyzed nationwide electronic health record data from more than 2.5 million patients in the US. In this study, COVID-19 patients with comorbid anxiety disorders were found to have a 2.4 times greater risk of mortality than those without anxiety disorders (Teixeira et al., 2020). In a recent study, Catalan et al. analyzed data on 157,246 people from the Basque Country (Spain) and observed that anxiety disorders [OR: 1.54] were associated with an increased risk of hospital admission, but were not associated with a higher mortality risk among admitted patients (Catalan et al., 2022).
Although several original studies and meta-analyses have advanced the field, there are significant differences between findings on the impact of depression and anxiety disorders on COVID-19 severity and mortality. The majority of the aforementioned studies were conducted during the first year of the COVID-19 pandemic and it is possible that the effects of psychiatric disorders on COVID-19 severity and mortality have changed over time. Finally, there is a lack of studies on this topic conducted in Germany.
Therefore, the aim of this retrospective cohort study was to investigate associations between depression, anxiety disorder, and the risk of COVID-19 severity and mortality in patients treated in large hospitals in Germany.
2 Methods
2.1 Study population
This retrospective study based on anonymized electronic medical data from public hospitals across Germany, all belonging to the same hospital group, included 28,311 patients with a confirmed COVID-19 diagnosis (ICD-10 U07.1) hospitalized between March 11, 2020 and July 20, 2022.
Initially, data were collected as part of the “CORONA Germany” study (Clinical Outcome and Risk in hospitalized COVID-19 patients), a multicenter observational, prospective, epidemiological cohort study. All data collected from the data repository were validated using the hospital network's quality management database. The initial results of the prospective study have been published previously (Gessler et al., 2021; Gunawardene et al., 2021). The study was approved by the ethics committee of the General Medical Council (Aerztekammer) for the City of Hamburg and the ethics committee of the General Medical Council (Aerztekammer) for the City of Munich.
Demographic data (age, sex), COVID-19 relevant data (ventilation, mortality), time of COVID-19 diagnosis, and co-diagnosis data were used for the present study.
2.2 Study outcome
The main outcome of the study is associations between unipolar depression (ICD-10: F32, F33) and anxiety disorders (ICD-10: F41) diagnosed prior to COVID-19 and the risk of mechanical ventilation and mortality among COVID-19 patients.
2.3 Statistical analyses
Based on the dominance of the respective variants, the COVID-19 variant was defined as non-omicron (all cases from March 2020 to December 2021) or omicron (all cases from January 2022 to July 2022).
First, baseline characteristics of study patients were calculated as proportions (sex, co-morbidities, probable COVID-19 variant) or mean (SD) (age) separately for patients with and without depression and anxiety disorders. Co-diagnoses included cancer (ICD-10: C00–C97), diabetes mellitus (ICD-10: E10–E14), lipid metabolism disorder (ICD-10: E78), obesity (ICD-10: E66), heart failure (ICD-10: I50), ischemic heart disease (ICD-10: I20–I25), cerebrovascular disease (ICD-10: I60–I69), and cirrhosis of the liver (ICD-10: K70.3, K74) as these diagnoses are known to be common causes of mortality.
Multivariable logistic regression models were used to study associations between depression and anxiety and mechanical ventilation and mortality adjusted for age, sex, time of COVID-19 diagnosis, and co-diagnoses. P-values <0.05 were considered statistically significant. All analyses were carried out using R version 4.2.0 (2022-04-22). The results of the logistic regression analyses are presented as odds ratios (ORs) with 95% confidence intervals (CIs).
3 Results
A total of 28,311 patients diagnosed with COVID-19 were available for analyses. There was a strong relationship between death and the application of ventilation. In total, 3074 patients (10.9%) received ventilation, of whom 40.0% died; of the 25,237 patients without ventilation, 10.0% died.
Of the 28,311 study patients, 1970 (6.9%) had a diagnosis of depression and 369 (1.3%) had a diagnosis of an anxiety disorder. Table 1 shows the baseline characteristics of study patients. Both patients with depression and patients with anxiety disorders were significantly younger (depression: 63 years vs. no depression: 66 years; anxiety disorders: 59 years vs. no anxiety disorders 65 years) and more likely to be female (62.6% vs. 47.5% for depression vs. no depression and 61.2% vs. 48.4% for anxiety disorder vs. no anxiety disorders) than patients without these psychiatric disorders. Patients with anxiety disorders had a higher prevalence of obesity (9.2% vs. 5.3%) than patients without anxiety.Table 1 Baseline characteristics of study patients with and without depression and anxiety disorder diagnosis.
Table 1Variable Patients without depression (N = 26,341)a Patients with depression (N = 1970)a P-valueb Patients without anxiety disorder (N = 27,942)a Patients with anxiety disorder (N = 369)a P-valueb
Female 47.5 62.6 <0.001 48.4 61.2 <0.001
Age (Mean, SD) 66 (21) 63 (22) <0.001 65 (21) 59 (19) <0.001
Cancer 5.5 4.0 0.004 5.4 7.0 0.15
Diabetes mellitus 22.7 20.0 0.005 22.6 19.0 0.10
Lipid metabolism disorder 15.1 17.8 0.001 15.3 13.0 0.22
Obesity 5.2 6.3 0.036 5.3 9.2 <0.001
Heart failure 14.3 15.3 0.22 14.4 11.1 0.07
Ischemic heart disease, 13.9 12.7 0.14 13.8 12.5 0.46
Cerebrovascular disease 7.4 10.4 <0.001 7.6 6.2 0.32
Cirrhosis of the liver 0.9 0.9 0.85 0.9 0.3 0.27
Omicron variant 41.0 44.1 0.006 41.2 40.4 0.75
No omicron variant 59.0 55.9 58.8 59.6
a Data are presented as percentages unless otherwise specified.
b Welch Two Sample t-test; two-sample test for equality of proportions.
Fig. 1 shows the prevalence of mechanical ventilation and mortality in patients with and without depression and anxiety disorders. Some 8.7% of depression patients and 11.0% of non-depression patients received mechanical ventilation, and 8.3% of depression patients and 13.6% of non-depression patients died. 18.7% of anxiety disorder patients and 10.8% of non-anxiety disorder patients received mechanical ventilation, and 7.6% of anxiety disorder patients and 13.3% of non-anxiety disorder patients died (Fig. 1).Fig. 1 Proportion of patients with COVID-19-related ventilation and mortality depending on depression and anxiety diagnosis.
Fig. 1
Although the multivariable logistic regression found no association between depression diagnosis and the risk of mechanical ventilation, depression was associated with a decrease in the risk of mortality(OR: 0.71; 95% CI: 0.53–0.94). There was no association between anxiety disorders and the risk of death, but there was a strong positive association between anxiety disorders and the risk of mechanical ventilation (OR: 2.04; 95% CI: 1.35–3.10) (Table 2 ).Table 2 Association between depression, anxiety disorders, ventilation and death due to COVID-19 in hospitalized patients.
Table 2Variable Odds Ratio (95% CI)a P-value
Depression
Mechanical ventilation 1.13 (0.89–1.43) 0.306
Death 0.71 (0.53–0.94) 0.020
Anxiety disorders
Mechanical ventilation 2.04 (1.35–3.10) <0.001
Death 1.01 (0.58–1.78) 0.964
a Multivariable logistic regression model adjusted for age, sex, cancer, diabetes mellitus, lipid metabolism disorder, obesity, heart failure, ischemic heart disease, cerebrovascular disease, cirrhosis of the liver, and COVID-19 variant.
4 Discussion
4.1 Main findings
This retrospective study including more than 28,000 COVID-19 patients treated in 50 Asklepios hospitals in Germany between March 2020 and July 2022 showed that neither depression nor anxiety disorder were associated with increased COVID-19 mortality; indeed, we actually observed a negative association for depression. Furthermore, anxiety disorders were strongly associated with a risk of mechanical ventilation.
4.2 Interpretation of findings
Recently, a substantial body of research has focused on the impact of depression and anxiety on COVID-19-related mortality. In some studies, depression was associated with an increased mortality risk (Atkins et al., 2020; Azar et al., 2020; Wang et al., 2021). Our study does not confirm a positive association between depression and COVID-19 mortality.
One finding of this study which is difficult to interpret is the negative association we found between depression and COVID-19 mortality. This cannot be explained by the younger age or by the higher proportion of women in the depression group, since the regression analysis corrected for these factors and the negative association was not observed for the demographically similar anxiety group. COVID-19 severity is actually thought to be responsible for the increased mortality risk, and it is possible that depression patients in this study had milder COVID-19 symptoms. Wang et al. showed that chronic depression and loneliness were each associated with subsequent COVID-19 hospitalization at the same severity level (Wang et al., 2022). The study of Oskotsky et al., 2021 supported evidence that selective serotonin reuptake inhibitors (SSRIs) may be associated with reduced severity of COVID-19 as well as COVID-19 mortality. As many depression patients usually receive SSRI, this may be one of the possible reasons for reduced mortality in our study.
However, this finding should be verified in further studies.
The discrepancy between the findings of the published literature and the present study can be explained by the fact that COVID-19 mortality has decreased since the beginning of the pandemic (Fan et al., 2021; Jones et al., 2021).
Regarding anxiety disorders and COVID-19 death, our findings are in line with published sources (Poblador-Plou et al., 2020; Catalan et al., 2022). However, the strong association between anxiety disorders and mechanical ventilation in our study is a novel finding. Several studies have described respiratory manifestations of anxiety disorders in the past. Sardinha et al. discussed causes, consequences, and therapeutic implications of hyperventilation syndromes among panic disorder patients (Sardinha et al., 2009). Nardi et al. explained that panic disorder is associated with respiratory abnormalities such as enhanced CO2 sensitivity (Nardi et al., 2009). The symptom of breathlessness can occur in the case of respiratory disease but also in patients with anxiety disorders. There is a two-way relationship as panic can cause respiratory difficulties and respiratory difficulties can cause panic (Williams and Carel, 2018). COVID-19 patients with comorbid anxiety disorder who experience breathing difficulties may likely find these breathing symptoms particularly frightening and severe. Finally, these symptoms can cause panic attacks, which in turn cause further breathing difficulties. In such cases, anxiety disorder patients may more often need mechanical ventilation, although their objective COVID-19 severity is not associated with an increased risk of death.
In the total population of this study, deaths were higher among patients who received mechanical ventilation than among patients without ventilation. This supports the assumption that mechanical ventilation is applied in patients with greater COVID-19 severity. This trend has already been observed in another large German study (Hobohm et al., 2022), where 5.3 of survivors and 14.2% of non-survivors received mechanical ventilation. Our study shows that the positive relationship between ventilation and mortality observed for the overall COVID-19 population is not necessarily present in patients with anxiety disorders.
4.3 Strengths and limitations
Two major strengths of this study are the large sample size (n = 28,311) and the inclusion of patients diagnosed with COVID-19 over a long period of the pandemic from March 2020 to July 2022. However, our study is also subject to a number of limitations. First, both depression and anxiety disorder diagnosis seem to be underdiagnosed in our cohort as their prevalence in this study is much lower than the prevalence reported for the total population in Germany (Steffen et al., 2020; Niermann et al., 2021; Erhardt et al., 2022). Second, no information was available on psychiatric disease severity. Third, although different chronic conditions were used for adjustment in regression models, other diseases which were not included could have an impact on the study outcome. Furthermore, no detailed information is available on the causes of death in those patients who died. Most but not all mortality cases listed COVID-19 as the main cause of death. In addition, no medications used for COVID-19 therapy and no other medications were analyzed. Information on the vaccination status of the patients included in the study is also missing. Furthermore, given that this study only included patients treated in hospitals, the associations found in this study cannot be generalized to patients treated outside of hospitals. Finally, viral variants were not determined individually for patients. Variants were assigned in accordance with the predominant variant at the time the patient was diagnosed with COVID-19 and a distinction was only made based on whether patients were diagnosed before or since the omicron variant emerged (1/1/2022).
5 Conclusions
This study including approximately 28,000 patients treated in Aklepios hospitals in Germany between 2020 and 2022 found that depression and anxiety disorder diagnoses were not associated with increased COVID-19 mortality. Anxiety disorder was strongly associated with an increased risk of mechanical ventilation. Further studies are needed to investigate how depression and anxiety disorders may influence COVID-19 severity and mortality.
Funding
This research did not receive any specific grants from funding agencies in the public, commercial, or not-for-profit sectors.
Author contributions
KK contributed to the design of the study, managed the literature searches, wrote the first draft of the manuscript, and corrected the manuscript. PW and JF contributed to the study design and performed the statistical analyses. NG and MW revised the manuscript. DA, LJ, MG, TH, AK, RS, LS, and SS corrected the manuscript. All authors contributed to and have approved the final manuscript.
Declarations of competing interest
NG has received grants from Boston Scientific and Medtronic as well as support from Bayer Vital, but not in connection with the submitted work.
MAW has consulted with Allergan/Abbvie and received remuneration for talks from Allergan/Abbvie, Biogen, and Schwabe, but not in connection with the submitted work.
All other authors declare that they have no competing financial interests.
Acknowledgments
The authors would like to thank our study team Claudia Kalkowski, Susanne Scholz, Charlotte Arms, Hanna Nugent, Francis Maren Konermann, Philipp Anders, Tobias Gethmann, Aaron Wilhelm Sievering, and Victor Rechl for their support with data entry and especially Kathrin Heitmann, Ina Koch, Sara Oldfield, and Kai Jaquet for their assistance with management and organization.
We would also like to express our sincere thanks to Christoph Jermann, Monika Grimm, and the team of the Asklepios Campus Hamburg, Semmelweis University for providing us with the infrastructure and support we needed to complete this study. In addition, we warmly thank the IT team, especially Claudio Forte and Thomas Koschmieder, for their assistance.
We would also like to thank all of the other members of the steering committee:
Tino Schnitgerhans, Sebastian Wirtz, Martin Bergmann, Martin Bachmann, Berthold Bein, Axel Stang, Sebastian Wirtz, Claas Wesseler, Klaus Herrlinger, Ulrich-Frank Pape, Christian Gloeckner, Lorenz Nowak, Juergen Behr, and Stephan Willems.
Finally, we thank all the physicians and nurses from the COVID-19 wards of all participating centers for the patient care they have provided during the ongoing pandemic.
==== Refs
References
Atkins J.L. Masoli J.A.H. Delgado J. Pilling L.C. Kuo C.L. Kuchel G.A. Melzer D. Preexisting comorbidities predicting COVID-19 and mortality in the UK biobank community cohort J Gerontol A Biol Sci Med Sci 75 11 2020 2224 2230 10.1093/gerona/glaa183 Oct 15 32687551
Azar K.M.J. Shen Z. Romanelli R.J. Lockhart S.H. Smits K. Robinson S. Brown S. Pressman A.R. Disparities in outcomes among COVID-19 patients in A large health care system in California Health Aff. 39 7 2020 1253 1262 10.1377/hlthaff.2020.00598 Jul
Catalan A. Aymerich C. Bilbao A. Pedruzo B. Pérez J.L. Aranguren N. Salazar de Pablo G. Hedges E. Gil P. Segarra R. González-Pinto A. Fernández-Rivas A. Inchausti L. McGuire P. Fusar-Poli P. González-Torres M.Á. Psychosis and substance abuse increase the COVID-19 mortality risk Psychol. Med. 2022 1 9 10.1017/S0033291722000976 Apr 12
Ceban F. Nogo D. Carvalho I.P. Lee Y. Nasri F. Xiong J. Lui L.M.W. Subramaniapillai M. Gill H. Liu R.N. Joseph P. Teopiz K.M. Cao B. Mansur R.B. Lin K. Rosenblat J.D. Ho R.C. McIntyre R.S. Association between mood disorders and risk of COVID-19 infection, hospitalization, and death: a systematic review and meta-analysis JAMA Psychiatr. 78 10 2021 1079 1091 10.1001/jamapsychiatry.2021.1818 Oct 1
Erhardt A. Gelbrich G. Klinger-König J. Streit F. Kleineidam L. Riedel-Heller S.G. NAKO Investigators Schmidt B. Schmiedek F. Wagner M. Grabe H.J. Rietschel M. Berger K. Deckert J. Generalised anxiety and panic symptoms in the German National Cohort (NAKO) World J. Biol. Psychiatr. 2022 1 16 10.1080/15622975.2021.2011409 Feb 17
Fan G. Yang Z. Lin Q. Zhao S. Yang L. He D. Decreased case fatality rate of COVID-19 in the second wave: a study in 53 countries or regions Transbound. Emerg. Dis. 68 2021 213 215 10.1111/tbed.13819 32892500
Gessler N. Gunawardene M.A. Wohlmuth P. Arnold D. Behr J. Gloeckner C. Herrlinger K. Hoelting T. Pape U.F. Schreiber R. Stang A. Wesseler C. Willems S. Arms C. Herborn C.U. Clinical outcome, risk assessment, and seasonal variation in hospitalized COVID-19 patients-Results from the CORONA Germany study PLoS One 16 6 2021 e0252867 10.1371/journal.pone.0252867 Jun 17
Gunawardene M.A. Gessler N. Wohlmuth P. Heitmann K. Anders P. Jaquet K. Herborn C.U. Arnold D. Bein B. Bergmann M.W. Herrlinger K.R. Stang A. Schreiber R. Wesseler C. Willems S. Prognostic impact of acute cardiovascular events in COVID-19 hospitalized patients-results from the CORONA Germany study J. Clin. Med. 10 17 2021 3982 10.3390/jcm10173982 Sep 2 34501427
Hobohm L. Sagoschen I. Barco S. Schmidtmann I. Espinola-Klein C. Konstantinides S. Münzel T. Keller K. Trends and risk factors of in-hospital mortality of patients with COVID-19 in Germany: results of a large nationwide inpatient sample Viruses 14 2 2022 275 10.3390/v14020275 Jan 28 35215869
Jones S. Mason N. Palser T. Swift S. Petrilli C.M. Horwitz L.I. Trends in risk-adjusted 28-day mortality rates for patients hospitalized with COVID-19 in England J. Hosp. Med. 16 2021 290 293 10.12788/jhm.3599 33617437
Nardi A.E. Freire R.C. Zin W.A. Panic disorder and control of breathing Respir. Physiol. Neurobiol. 167 1 2009 133 143 10.1016/j.resp.2008.07.011 May 30 18707030
Nemani K. Li C. Olfson M. Blessing E.M. Razavian N. Chen J. Petkova E. Goff D.C. Association of psychiatric disorders with mortality among patients with COVID-19 JAMA Psychiatr. 78 4 2021 380 386 10.1001/jamapsychiatry.2020.4442 Apr 1
Niermann H.C.M. Voss C. Pieper L. Venz J. Ollmann T.M. Beesdo-Baum K. Anxiety disorders among adolescents and young adults: prevalence and mental health care service utilization in a regional epidemiological study in Germany J. Anxiety Disord. 83 2021 102453 10.1016/j.janxdis.2021.102453 Oct
Oskotsky T. Maric I. Tang A. Oskotsky B. Wong R.J. Aghaeepour N. Sirota M. Stevenson D.K. Mortality risk among patients with COVID-19 prescribed selective serotonin reuptake inhibitor antidepressants JAMA Netw. Open 4 11 2021 e2133090 10.1001/jamanetworkopen.2021.33090 Nov 1
Poblador-Plou B. Carmona-Pírez J. Ioakeim-Skoufa I. Poncel-Falcó A. Bliek-Bueno K. Cano-Del Pozo M. Gimeno-Feliú L.A. González-Rubio F. Aza-Pascual-Salcedo M. Bandrés-Liso A.C. Díez-Manglano J. Marta-Moreno J. Mucherino S. Gimeno-Miguel A. Prados-Torres A. EpiChron Group Baseline chronic comorbidity and mortality in laboratory-confirmed COVID-19 cases: results from the PRECOVID study in Spain Int. J. Environ. Res. Publ. Health 17 14 2020 5171 10.3390/ijerph17145171 Jul 17
Plana-Ripoll O. Pedersen C.B. Agerbo E. Holtz Y. Erlangsen A. Canudas-Romo V. Andersen P.K. Charlson F.J. Christensen M.K. Erskine H.E. Ferrari A.J. Iburg K.M. Momen N. Mortensen P.B. Nordentoft M. Santomauro D.F. Scott J.G. Whiteford H.A. Weye N. McGrath J.J. Laursen T.M. A comprehensive analysis of mortality-related health metrics associated with mental disorders: a nationwide, register-based cohort study Lancet 394 10211 2019 1827 1835 10.1016/S0140-6736(19)32316-5 Nov 16 31668728
Sardinha A. Freire R.C. Zin W.A. Nardi A.E. Respiratory manifestations of panic disorder: causes, consequences and therapeutic implications J. Bras. Pneumol. 35 7 2009 698 708 10.1590/s1806-37132009000700012 Jul 19669009
Steffen A. Thom J. Jacobi F. Holstiege J. Bätzing J. Trends in prevalence of depression in Germany between 2009 and 2017 based on nationwide ambulatory claims data J. Affect. Disord. 271 2020 239 247 10.1016/j.jad.2020.03.082 Jun 15 32479322
Teixeira A.L. Krause T.M. Ghosh L. Shahani L. Machado-Vieira R. Lane S.D. Boerwinkle E. Soares J.C. Analysis of COVID-19 infection and mortality among patients with psychiatric disorders, 2020 JAMA Netw. Open 4 11 2021 e2134969 10.1001/jamanetworkopen.2021.34969 Nov 1
Vai B. Mazza M.G. Delli Colli C. Foiselle M. Allen B. Benedetti F. Borsini A. Casanova Dias M. Tamouza R. Leboyer M. Benros M.E. Branchi I. Fusar-Poli P. De Picker L.J. Mental disorders and risk of COVID-19-related mortality, hospitalisation, and intensive care unit admission: a systematic review and meta-analysis Lancet Psychiatr. 8 9 2021 797 812 10.1016/S2215-0366(21)00232-7 Sep
Wang Y. Yang Y. Ren L. Shao Y. Tao W. Dai X.J. Preexisting mental disorders increase the risk of COVID-19 infection and associated mortality Front. Public Health 9 2021 684112 10.3389/fpubh.2021.684112 Aug 9
Wang S. Quan L. Ding M. Kang J.H. Koenen K.C. Kubzansky L.D. Branch-Elliman W. Chavarro J.E. Roberts A.L. Depression, worry, and loneliness are associated with subsequent risk of hospitalization for COVID-19: a prospective study Psychol. Med. 2022 1 10 10.1017/S0033291722000691 May 19
Williams T. Carel H. Breathlessness: from bodily symptom to existential experience Aho K. Existential Medicine 2018 Rowman & Littlefield International London (UK) (Chapter 10). Available from: https://www.ncbi.nlm.nih.gov/books/NBK535717/
World Health Organization Coronavirus (COVID-19) dashboard https://covid19.who.int/
| 36481563 | PMC9706218 | NO-CC CODE | 2022-12-05 23:15:30 | no | J Psychiatr Res. 2023 Jan 29; 157:192-196 | utf-8 | J Psychiatr Res | 2,022 | 10.1016/j.jpsychires.2022.11.031 | oa_other |
==== Front
Mol Aspects Med
Mol Aspects Med
Molecular Aspects of Medicine
0098-2997
1872-9452
The Authors. Published by Elsevier Ltd.
S0098-2997(22)00104-2
10.1016/j.mam.2022.101159
101159
Review
Viral proteases as therapeutic targets
Majerová Taťána a
Konvalinka Jan ab∗
a Institute of Organic Chemistry and Biochemistry of the Czech Academy of Sciences, Flemingovo nám. 2, 166 10, Prague 6, Czech Republic
b Department of Biochemistry, Faculty of Science, Charles University in Prague, 128 43, Prague, Czech Republic
∗ Corresponding author. Institute of Organic Chemistry and Biochemistry of the Czech Academy of Sciences, Flemingovo nám. 2, 166 10, Prague 6, Czech Republic.
29 11 2022
12 2022
29 11 2022
88 101159101159
13 8 2022
21 11 2022
23 11 2022
© 2022 The Authors
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Some medically important viruses―including retroviruses, flaviviruses, coronaviruses, and herpesviruses―code for a protease, which is indispensable for viral maturation and pathogenesis. Viral protease inhibitors have become an important class of antiviral drugs. Development of the first-in-class viral protease inhibitor saquinavir, which targets HIV protease, started a new era in the treatment of chronic viral diseases. Combining several drugs that target different steps of the viral life cycle enables use of lower doses of individual drugs (and thereby reduction of potential side effects, which frequently occur during long term therapy) and reduces drug-resistance development. Currently, several HIV and HCV protease inhibitors are routinely used in clinical practice. In addition, a drug including an inhibitor of SARS-CoV-2 main protease, nirmatrelvir (co-administered with a pharmacokinetic booster ritonavir as Paxlovid®), was recently authorized for emergency use. This review summarizes the basic features of the proteases of human immunodeficiency virus (HIV), hepatitis C virus (HCV), and SARS-CoV-2 and discusses the properties of their inhibitors in clinical use, as well as development of compounds in the pipeline.
==== Body
pmc1 Antivirals
Established and emerging viruses cause a wide range of illnesses (McArthur, 2019). Unlike bacteria, viruses replicate solely intracellularly and rely on the synthetic machinery of the host cell. Thus, development of drugs targeting viruses without affecting the host cell is challenging (Kausar et al., 2021).
Idoxuridine was the first antiviral drug approved by the Food and Drug Administration (FDA) in 1963. This compound is a nucleoside analogue, which targets synthesis of herpesviral DNA (Maxwell, 1963; Prusoff, 1959). Other drugs with a similar mechanism of action followed, including the widely used acyclovir (De Clercq and Li, 2016; Elion et al., 1977; Furman et al., 1981; Hermans and Cockerill, 1983). These nucleoside analogues enter the cell, where they are phosphorylated. Subsequently, such phosphorylated drugs can block viral DNA or RNA polymerases from intracellular synthesis of the functional viral genome. Viral polymerases are less specific than mammalian polymerases, which means that in many cases, a nucleoside derivative designed against one viral polymerase also acts against several others (Kataev and Garifullin, 2021). Due to viral polymerase promiscuity, nucleoside drugs can be considered broad-spectrum, or “less-narrow-spectrum” antivirals. This attribute is beneficial for drug repurposing, an emerging drug discovery strategy in which approved or investigational drugs are used for a condition different to their original medical indications. Recent examples of drug repurposing include remdesivir and molnupiravir. These compounds were originally developed as anti-HCV, anti-Ebola or anti-influenza drugs but are now used to treat the SARS-CoV-2 infection (de Wit et al., 2020; Kabinger et al., 2021; Sheahan et al., 2017; Toots et al., 2019, 2020; Warren et al., 2016; Williamson et al., 2020). The major disadvantage of nucleoside analogues is their potential toxicity and mutagenicity for the host cells (Wutzler and Thust, 2001; Zhou et al., 2021). Off-target actions are particularly problematic in the treatment of chronic infections, when the drugs must be administered for a prolonged period of time. Development of structurally diverse antivirals with a better safety profile, which also retain activity against drug-resistant viral variants, has become of primary interest (Bean, 1992; Kuroki et al., 2021; Meganck and Baric, 2021). These needs have turned the attention of researchers to compounds with new mechanisms of action.
The first antiviral compound with a completely new mechanism of action was saquinavir, a peptidomimetic inhibitor of HIV protease (Craig et al., 1991; James, 1995; Roberts et al., 1990). Other inhibitors of viral proteases and compounds targeting various steps of viral infection followed (De Clercq, 2004). The availability of compounds with different mechanisms of action (Matthew et al., 2021) enables several steps of the viral replication cycle to be targeted simultaneously. Such combination therapies can be more efficient and safer, as concomitantly administrated low doses can induce antiviral action while reducing the side effects (Hézode, 2018; Paredes and Clotet, 2010).
Nevertheless, all compounds in current clinical use are virostatic drugs that stop further viral replication, but cannot eliminate viral particles that have already been formed (Pankey and Sabath, 2004; Sutton et al., 2021). Due to the inherent nature of virostatic drugs, antivirals have to be administered in the early phases of acute viral infection to minimize the viral load in the infected individual (Dunning et al., 2020; Mehta et al., 2021).
2 Viral polyprotein processing
Some viruses express their proteins in the form of a polyprotein containing one or more proteases. Such proteases autocatalytically release themselves from the precursor and cleave the remaining parts of the polyprotein into functional proteins. Inhibition of this proteolytic activity can block production of infectious viral progeny and reduce pathogenic processes (Han et al., 1995; Schneider and Kent, 1988; Seelmeier et al., 1988; Tsu et al., 2021). All positive single-stranded RNA viruses and some DNA viruses belong to this group (see Table 1 for summary). Examples of clinically important inhibitors of viral proteases include inhibitors of human immunodeficiency virus (HIV) and hepatitis C virus (HCV) proteases (Skwarecki et al., 2021). In addition, the FDA recently authorized an inhibitor of the main protease of SARS-CoV-2 (Hammond et al., 2022; Owen et al., 2021).Table 1 Medically important families of viruses (Siegel, 2018) coding for one or more proteases.
Table 1 Family Genus Examples of viruses
Positive single-stranded RNA viruses, enveloped Retroviridae Lentivirus HIV-1, HIV-2, HTLV-1, HTLV-2
Flaviviridae Flavivirus Dengue virus, Zika virus, yellow fever virus, Japanese encephalitis virus, West Nile virus, tick-borne encephalitis virus
Hepacivirus Hepatitis C virus
Coronaviridae Coronavirus SARS-CoV, SARS-CoV-2, MERS-CoV, OC43,32 229E, NL63,33, HKU134
Togaviridae Rubivirus Rubeola virus
Alphavirus Chikungunya virus
Positive single-stranded RNA viruses, non-enveloped Picornaviridae Enterovirus Polio virus, rhinoviruses, coxsackieviruses, echoviruses
Hepatovirus Hepatitis A virus
Hepeviridae Hepevirus Hepatitis E virus
Caliciviridae Norovirus Noroviruses causing gastroenteritis
Sapovirus Sapoviruses causing gastroenteritis
Negative single-stranded RNA viruses, enveloped Nairoviridae Orthonairovirus Crimean-Congo hemorrhagic fever virus
Double-stranded DNA viruses, enveloped Herpesviridae Simplexvirus Herpes simplex virus 1 and 2
Varicellovirus Varicella-zoster virus
Cytomegalovirus Cytomegalovirus
Rhadinovirus Kaposi sarcoma-associated herpesvirus
Poxviridae Orthopoxvirus Small pox virus
Double-stranded DNA viruses, non-enveloped Adenoviridae Mastadenovirus Human adenoviruses 1-57
Some RNA and DNA viruses cannot be targeted by protease inhibitors, as they exploit different strategies, such as alternative mRNA splicing, to express several proteins from one RNA molecule (Ho et al., 2021). These include negative single-stranded RNA viruses (Payne, 2017), such as influenza viruses (Majerová et al., 2010), Lassa virus, Marburg virus, morbillivirus, mumps virus, lyssavirus, and vesicular stomatitis virus (Trovato et al., 2020); double-stranded RNA viruses including rotaviruses; and certain DNA viruses, such as papillomaviruses (Graham, 2017) and polyomaviruses (Saribas et al., 2019).
3 HIV and anti-HIV therapy
HIV is the retrovirus that causes acquired immunodeficiency syndrome (AIDS) (Barré-Sinoussi et al., 1983; Gallo et al., 1983; Sharp and Hahn, 2011). There are two types of HIV – HIV-1 and HIV-2 – with HIV-2 being less transmissible and less virulent than HIV-1. While HIV-1 is spread worldwide, HIV-2 is confined mainly to West Africa (Clavel et al., 1986). Other medically significant retroviruses include endemic human T-lymphotropic viruses (Gallo et al., 1981). Human DNA also harbors fragments of ancient retroviruses, some of which are partially retained as functional genes (Grandi and Tramontano, 2018; Lander et al., 2001).
Retroviruses utilize two important enzymes supporting their unique replication cycle: reverse transcriptase, which copies viral genomic RNA into DNA, and integrase, which integrates this DNA (called proviral DNA) into the host cell genome. As cells divide, the integrated proviral DNA remains a part of the genome of newly arising cells. Proliferating host cells, such as CD4+ T-cells, with integrated proviral DNA, form the latent HIV reservoir (Chun et al., 1995; Morcilla et al., 2021). Because the integrated proviral DNA becomes part of the host cell genome, it cannot be removed by commonly used antivirals. Recently, in animal models, integrated provirus was removed from the host cell genome using the genome-editing technology CRISPR/Cas9 (Dash et al., 2019; Mancuso et al., 2020). These promising results led to initiation of an early-stage clinical trial, announced by Excision BioTherapetics in September 2021 (Excision; https://www.excision.bio/technology). Another potentially promising approach involves activation of latent virus reservoirs combined with active elimination of virus-producing cells (recently reviewed in (Ward et al., 2021)).
Although antiretroviral therapy currently cannot eliminate proviral DNA from the genome, it averts development of AIDS. The first anti-HIV drug – 3′-azido-2′,3′-dideoxythymidine (AZT or zidovudine) – is a nucleoside chain terminator of viral reverse transcription (De Clercq, 1987; Mitsuya et al., 1985; Robins and Robins, 1964). After FDA approval of AZT in 1987, other compounds with the same mechanism of action followed (Broder, 2010). A novel mechanism of action – HIV protease inhibition – was introduced by the development of saquinavir (Invirase®), which was approved by the FDA in 1995 (Craig et al., 1991; James, 1995; Roberts et al., 1990). This enabled simultaneous use of nucleosides targeting reverse transcription and protease inhibitors blocking viral maturation. Simultaneously targeting these two steps of the viral life cycle led to a decrease of viremia below a detectable level, followed by an increase in CD4+ T-cell count to a normal level. This approach started a new era in anti-HIV therapy: HAART (highly active antiretroviral therapy; today referred to as cART or ART) (Smart, 1995).
Subsequently, inhibitors with other mechanisms of action have been approved: non-nucleoside reverse transcriptase inhibitors, integrase inhibitors, and compounds blocking the initial steps of the viral life cycle (enfuvirtide, maraviroc, fostemsavir, and the anti-CD4 monoclonal antibody ibalizumab) (Tseng et al., 2015; Weichseldorfer et al., 2021). At present, a wide range of drug combinations is available to achieve maximal antiviral effect while minimizing the side effects. However, the risk of development of drug-resistant viral variants still persists (Fig. 1 ). Although the drug-resistant variants are usually less viable, other compensatory mutations can restore viral fitness. In addition, the number of compensatory mutations likely decreases the probability of reappearance of revertant mutants when therapy is discontinued (Zhang et al., 2020b). HIV subtypes have different virulence and drug susceptibility (Spira et al., 2003). Potential differences in virulence among drug-resistant variants have yet to be described.Fig. 1 A simplified schematic representation of drug resistance development: Due to errors randomly incorporated into the viral genome during viral replication, mutated viral variants continuously evolve. Most mutations are lethal for the virus or neutral. When the virus is under selection pressure (e.g., a person is treated by an antiviral drug), only viral variants able to circumvent the action of the drug can produce a new viral progeny.
Fig. 1
3.1 HIV protease
HIV expresses its proteins as the Gag-Pol polyprotein (Fig. 2 ). While Gag region bears viral structural proteins, Pol contains viral enzymes (protease, reverse transcriptase, RNase H, and integrase). During translation, either only the Gag region or the entire Gag-Pol polyprotein is synthesized. Synthesis of the whole Gag-Pol polyprotein is enabled by suppression of the stop codon between the Gag and Pol regions by a −1 frameshift at a specific mRNA site (Jacks et al., 1988). This process is precisely regulated and occurs in 5% of cases, as this ratio of Gag to Gag-Pol is crucial for efficient virus assembly, genome packaging, and maturation (Shehu-Xhilaga et al., 2001). Gag and Gag-Pol are myristoylated at the N-terminus by the host cell machinery. These myristoyl tags enable anchoring of the Gag and Gag-Pol polyproteins into the cell membrane (Hermida-Matsumoto and Resh, 1999), which gives rise to the sites of viral assembly. The viral genomic RNA is subsequently recruited to these sites, followed by formation of immature viral particles and their budding from the cell. To be able to infect new host cells, viral particles released from the cell of origin must undergo maturation, during which Gag and Gag-Pol polyproteins are cleaved by HIV protease into functional proteins (Katsumoto et al., 1987; Tabler et al., 2022).Fig. 2 A schematic representation of initial cleavage sites in HIV Gag-Pol polyprotein cleaved by HIV protease in cis (intramolecularly). The Gag region harbors viral structural proteins (matrix, capsid and nucleocapsid), whereas viral enzymes are expressed in the Pol region. Each Gag-Pol polyprotein bears one monomer of HIV protease, which must form a homodimer to be catalytically active. The numerals 1, 2 and 3 denote the order of cis-cleavage events.
Fig. 2
HIV protease is an aspartic protease, active only as a homodimer. Each monomeric subunit provides one catalytic Asp-Ser-Gly triad to form the active site. Gag and Gag-Pol cleavage by HIV protease must be preceded by autoactivation of the protease, resulting in its release from the Gag-Pol precursor. The initial cuts occur exclusively in cis, i.e. intramolecularly (Pettit et al., 2004) (Fig. 2). The cleavage sites are upstream of the N-terminus of HIV protease – between the p2 peptide and the nucleocapsid, and between the trans-frame octapeptide (TFP) and the p6*peptide. Then, the free N-terminus of HIV protease is released from the precursor by cleaving out the p6*peptide (Pettit et al., 2003, 2004). The subsequent cleavages of the Gag and Gal-Pol polyprotein by HIV protease occur in trans (intermolecularly) (Pettit et al., 2005).
Once the viral polyproteins are cleaved by HIV protease in the process of maturation, the virus becomes infectious and can attack a new host cell (Kohl et al., 1988; Tabler et al., 2022). For successful maturation, the proteolysis of viral polyproteins must be perfectly timed. This was demonstrated experimentally by adding a protease inhibitor during viral assembly and subsequently removing it from purified immature virions. While cleavage of Gag and Gag-Pol was restored after the inhibitor was removed, the viral particles remained noninfectious (Mattei et al., 2014). Indeed, inhibitors of HIV protease are capable of locking viral particles in a noninfectious immature state and thus play a crucial role in current anti-HIV treatment (Konvalinka et al., 2015).
Interestingly, release of HIV protease from the Gag-Pol polyprotein prior to virion assembly results in decreased production of viral particles and might lead to elimination of infected cells due to the cytotoxicity of HIV protease (Jochmans et al., 2010; Kaplan and Swanstrom, 1991; Kräusslich, 1991; Majerová and Novotný, 2021; Pan et al., 2012; Sudo et al., 2013; Trinité et al., 2019; Wang et al., 2021). A drug that would be able to provoke this premature release of HIV protease from Gag-Pol could potentially eliminate host cells with integrated proviral DNA in the genome. This would represent a complete cure of HIV infection.
3.1.1 Inhibitors of HIV protease
Inhibitors of HIV protease are important components of ART. A typical initial ART regimen consists of two nucleoside reverse transcriptase inhibitors and one integrase inhibitor or protease inhibitor. The regimen is usually modified over the course of therapy due to development of adverse drug reactions or drug resistance (Vitoria et al., 2019). Adverse events include gastrointestinal problems and hyperlipidemia (except for atazanavir), hepatic problems (mainly ritonanir, tipranavir, darunavir), rash (amprenavir, tipranavir, darunavir), hyperbilirubinemia (indinavir, atazanavir), paresthesia (ritonavir, amprenavir), nephrolithiasis (indinavir, atazanavir), retinoid-like effects (indinavir) (Boesecke and Cooper, 2008), cardiovascular problems (Alvi et al., 2018), and reversible inhibition of insulin secretion (mainly indinavir; also amprenavir, nelfinavir, and ritonavir) (Koster et al., 2003; Pokorná et al., 2009).
3.1.1.1 First-generation HIV protease inhibitors
The first-generation HIV protease inhibitors primarily targets the active site. They bind reversibly into the substrate pocket of the enzyme, hindering binding of substrates. Most of these inhibitors are peptidomimetics derived from natural cleavage sites. They include saquinavir, ritonavir, indinavir, nelfinavir, amprenavir, and fosmaprenavir (Ghosh et al., 2016) (Fig. 3 ).Fig. 3 Structures of the first-generation inhibitors of HIV protease in clinical use.
Fig. 3
The peptidomimetic structure of the first-in-class drug saquinavir (Invirase®, Fortovase®) was inspired by the Tyr-Pro cleavage sites naturally occurring in the Gag-Pol polyprotein; the scissile bond is replaced with an uncleavable hydroxyethylamine transition state mimetic. Saquinavir interacts with all subsites of the substrate binding cavity (Craig et al., 1991; Thompson et al., 1993). As Tyr-Pro substrates are atypical for mammalian proteases, the inhibitors derived from unique structures are anticipated to be specific for viral proteases, reducing off-target activities against host proteolytic enzymes (Jacobsen et al., 1995; Roberts et al., 1990).
Ritonavir (Norvir®) was approved by the FDA as a second HIV protease inhibitor in 1996 (Kempf et al., 1995). However, high doses of ritonavir cause strong gastrointestinal side effects, as ritonavir also potently inhibits the 3A4 isoenzyme of hepatic cytochrome P-450. Inhibition of cytochrome P-450 3A4 blocks metabolism of many drugs, including HIV protease inhibitors, and thus leads to increased plasma levels of these drugs (Kumar et al., 1996). This surprising observation lead to the development of new class of compounds, inhibitors of cytochromes P450 that can be used as general “boosters” improving of therapeutics with compromised serum half-life, including HIV protease inhibitors. Ritonavir became “first in class” of these boosters (Hakkola et al., 2020; Kempf et al., 1997).
Another example of a pharmacokinetic booster used in anti-HIV therapy is cobicistat, which also inhibits cytochrome P-450 3A4 but does not inhibit HIV protease (Elion et al., 2011; Mathias et al., 2010). Moreover, cobicistat is more specific to the P-450 3A4 isoenzyme and does not interact with other cytochrome isoenzymes to much extent. It is thus a preferred choice of pharmacokinetic booster in newer drug formulations. However, in contrast to ritonavir, cobicistat may influence serum creatinine and is less suitable for use in pregnancy (Nguyen et al., 2016; Sherman et al., 2015). The effectiveness of some compounds is not improved equally by addition of cobicistat or ritonavir. For instance, tipranavir is boosted by ritonavir more effectively than by cobicistat (Ramanathan et al., 2016).
Indinavir (Crixivan®), the next HIV protease inhibitor to be approved by the FDA, in 1996, was developed based on the structure of experimental renin inhibitors (Churchill, 1996; Vacca et al., 1994). Researchers screened a library of renin inhibitors to provide a lead peptide structure for HIV protease inhibition. A small modification of the lead molecule (removal of N-terminal phenylalanine) resulted in a compound lacking anti-renin activity, but retaining anti-HIV activity (Vacca et al., 1991). The compound was further modified to improve its pharmacokinetic properties (Vacca et al., 1994). However, indinavir has a very short half-life (1.8 h) under physiological conditions and thus requires frequent administration (Churchill, 1996).
Nelfinavir (Viracept®), approved by the FDA in 1997, is the first nonpeptidic HIV protease inhibitor obtained by rational, iterative, structure-based design combined with pharmacokinetic optimizations (Gehlhaar et al., 1995; Kaldor et al., 1997; Shetty et al., 1996). Nelfinavir shares some structural features with saquinavir, including a 2-quinoline-carboxamide moiety (Fig. 3) (Ghosh et al., 2016).
Amprenavir (Agenerase®) was approved by the FDA in 1999 to treat HIV infection (Kim et al., 1995; Miller, 1999; Murphy et al., 1999). This compound was again developed based on modified experimental renin inhibitors. It contains a sulfonamide moiety and a hydroxyethylamine isostere, which serves to mimic the transition state (Kim et al., 1995). To improve the pharmacokinetics a hydrophilic phosphate ester of amprenavir―fosamprenavir (Lexiva®) ―was developed as a prodrug and introduced into clinical use in 2003. During absorption in the gut, fosamprenavir is converted to amprenavir by host phosphatases (Furfine et al., 2004; Wire et al., 2006).
3.1.1.2 Drug resistance development
The first generation of HIV protease inhibitors suffered from suboptimal pharmacokinetics, requiring administration of high doses of the drugs several times per day (van Heeswijk et al., 2001). This dosing reinforced the side effects of these drugs, leading to health problems that have sometimes required changes in the treatment regimen and/or caused poor adherence. Consequently, suboptimal dosing accelerated the development of drug-resistant viral variants, which became an important problem in anti-HIV therapy (Pokorná et al., 2009).
Due to the reverse transcription step, replication of HIV continuously gives rise to mutated variants. Retroviral reverse transcriptase lacks proofreading activity and introduces mutations into nascent proviral DNA at a frequency of 1 error per 1,700–4,000 bases, i.e. 2 to 6 errors per each HIV-1 genome molecule. The errors are not distributed equally and are more frequent in mutation hotspots (Preston et al., 1988; Roberts et al., 1988), which are likely delineated by the 3D structure of the viral template genetic information. Most mutations in coding regions are deleterious or neutral (they do not influence the function of a mutated protein). Occasionally, mutations that confer an advantage to the virus appear. Viruses with such mutations can produce viral progeny more efficiently, and the viral population becomes gradually enriched with such variants. In some cases, a mutation can only be beneficial under certain conditions, such as in the presence of a drug (Nijhuis et al., 2009) (Fig. 1).
Complete blockade of production of viral progeny during ART is crucial, as suboptimal drug concentrations lead to selection and replication of drug-resistant viral variants. Interestingly, mutations that originate under selection pressure do not necessarily have to result in increased HIV protease activity, as even minor HIV protease activity is sufficient for production of infectious viral progeny. This was demonstrated with an active site mutant of HIV protease, in which a threonine adjacent to the catalytic aspartate was replaced with a serine. Even though the purified mutant protease had one-order-of-magnitude lower activity than the wild type, the virus bearing this mutant protease showed no significant differences in polyprotein cleavage, infection kinetics, and titer compared with the wild-type virus (Konvalinka et al., 1995). Similarly, some drug-resistant HIV protease mutants show lower activity in vitro, while the infectivity of viruses bearing these protease variants remains unchanged (Weber et al., 2002).
The first-generation HIV protease inhibitors share some structural features and target similar binding sites (Fig. 3). It is therefore not surprising that the development of drug resistance under selection pressure of one inhibitor usually causes cross-resistance to one or more other inhibitors. Fortunately, the genetic drug resistance barrier is quite high, i.e., a virus usually needs to accumulate several mutations to become resistant to a drug (Chang and Torbett, 2011; Shah et al., 2020; Weikl and Hemmateenejad, 2020). Interestingly, the mutations that confer resistance to HIV protease inhibitors do not occur only within the active site of the protease but also in other parts of the Gag-Pol polyprotein, mainly in the polyprotein cleavage sites. This brings even greater complexity to the problem of anti-HIV drug resistance (Aoki et al., 2009; Kozísek et al., 2012; Nijhuis et al., 2007; Zhang et al., 1997).
Characteristic primary drug-resistance-conferring mutations developed under the selection pressure of saquinavir include G48V and L90M (Jacobsen et al., 1995). The D30N mutation arises with nelfinavir treatment (Kožíšek et al., 2007; Patick et al., 1998). Amprenavir/fosamprenavir treatment can lead to M46I/L, I47V, I50L, I54 L/V, and I84V variants (Arabi et al., 2021; Marcelin et al., 2004). The mutations M46L/I and V82A arise upon treatment with indinavir, later followed by a mutation at either I54V or A71 V/T (Zhang et al., 1997). Typical mutations that confer cross-resistance include M46I/L, I47V, G48V, V82 A/T, I84V, and L90M (Svicher et al., 2005). Cross-resistance to saquinavir and ritonavir or to nelfinavir and indinavir is common (Majerová-Uhlíková et al., 2006; Pawar et al., 2018; Shah et al., 2020).
3.1.1.3 Second-generation HIV protease inhibitors
Rapid development of drug resistance to first-generation HIV protease inhibitors led to efforts to design drugs with higher barriers to resistance development. Second-generation inhibitors of HIV protease were designed to form fewer binding contacts with non-conserved residues prone to mutations and more binding contacts with the backbone of the enzyme and conserved residues (Ghosh et al., 2008). The latter includes interactions beyond the active site cavity of the enzyme, namely with the flaps of HIV protease (Weber et al., 2015). Additionally, the enthalpic and entropic contributions of the protease-inhibitor interaction were optimized, and attempts were made to improve the pharmacokinetic profiles of the new compounds (Ghosh et al., 2016; Majerová and Konvalinka, 2021; Wensing et al., 2010). Second-generation inhibitors include lopinavir, tipranavir, atazanavir, and darunavir (Fig. 4 ).Fig. 4 Structures of the second-generation inhibitors of HIV protease.
Fig. 4
Lopinavir (Kaletra®) was approved by the FDA in 2004. Its structure was derived from ritonavir and modified to minimize contacts with Val82 of HIV protease, a residue inside the substrate-binding pocket that is critical in drug resistance development. Lopinavir was the first HIV protease inhibitor co-administered with the pharmacokinetic booster ritonavir (Sham et al., 1998; Vogel and Rockstroh, 2005). The main mutations associated with failure of lopinavir treatment are K20 M/R and F53L. These mutations, in combination with other compensatory mutations, dramatically decrease the protease's sensitivity to lopinavir (Kempf et al., 2001). Other lopinavir-resistance-conferring mutations include L76V, Q58E, L90M, and I54V (Champenois et al., 2011). Some mutations that evolve under the selection pressure of lopinavir even result in sub-optimal binding of some first-generation HIV protease inhibitors. For example, the I47A mutation causes a 100-fold decrease in sensitivity to lopinavir as well as resistance to amprenavir/fosamprenavir. Interestingly, it also leads to hyper-susceptibility to saquinavir (de Mendoza et al., 2006; Sasková et al., 2008, 2009).
Tipranavir (Aptivus®), which the FDA approved in 2005, is based on the nonpeptidic compound phenprocoumon, which was identified as a lead structure during compound library screening for inhibition of HIV protease. Tipranavir has no amide group but contains one sulfonamide moiety. Unlike other inhibitors of HIV protease, tipranavir not only binds to the substrate pocket of the enzyme (Thaisrivongs et al., 1996; Thaisrivongs and Strohbach, 1999), but also blocks HIV protease dimerization (Aoki et al., 2012). Tipranavir is a strong inducer of cytochrome P450 and is thus co-administered with higher doses of the pharmacokinetic booster ritonavir (Morello et al., 2007).
Tipranavir retains a high activity against drug-resistant variants of HIV protease. Loss of sensitivity to tipranavir is accompanied by loss of sensitivity to other specific HIV protease inhibitors, except saquinavir. Decreased susceptibility to tipranavir has been reported for viral variants accumulating different combinations of the HIV protease mutations L10F, I13V, V32I, L33F, M36I, K45I, I54V, A71V, V82L, and I84V, as well as one cleavage site within the Gag polyprotein. These tipranavir-resistant variants have lower replicative capacity (Doyon et al., 2005).
Atazanavir (Reyataz®), which the FDA approved in 2003, is an azapeptide inhibitor with a long half-life (Bold et al., 1998; Piliero, 2002). It can be administered once per day with cobicistat (Evotaz®) or without a pharmacokinetic booster (von Hentig, 2008). As atazanavir is both a substrate and inhibitor of cytochrome P450, its administration increases plasma levels of some other drugs, including other HIV protease inhibitors, statins, warfarin, and drugs used to treat gastric ulcers (Busti et al., 2004). The I50L mutation is a hallmark of drug resistance to atazanavir (Goldsmith and Perry, 2003).
Darunavir is the most recently developed FDA-approved protease inhibitor, which was approved for single-agent administration in 2006 (Prezista®) (Ghosh et al., 2007; Koh et al., 2003). Additionally, darunavir is available in combinations with cobicistat (Prezcobix®) or cobicistat and two nucleoside reverse transcriptase inhibitors (Symtuza®) (Rittweger and Arastéh, 2007). The structure of darunavir resembles that of amprenavir, as it contains a sulfonamide moiety and hydroxyethylamine, which mimics the transition state of the substrate cleavage (Fig. 4). In contrast to amprenavir, darunavir also contains a bis-tetrahydrofuranyl group at its N-terminus, which mediates hydrogen bonding with the backbone of the enzyme in the vicinity of residues Asp29 and Asp30, mimicking conserved enzyme-substrate bonds (Koh et al., 2003). While darunavir primarily binds to the HIV protease active site, a secondary binding site blocking HIV protease dimerization has been reported (Hayashi et al., 2014). Darunavir also seems to have a high affinity to the precursor form of HIV protease. In fact, inhibition of HIV protease in its noncleaved precursor form has been observed for other inhibitors, but it required much higher (several orders of magnitude) doses of the drugs. Darunavir has the highest affinity to the precursor form of all inhibitors in clinical use, possibly because of the contribution of the secondary binding site (Davis et al., 2012; Huang et al., 2019; Humpolíčková et al., 2018; Louis et al., 2011; Park et al., 2016). Although darunavir has a high barrier to drug-resistance development, darunavir-resistant variants accumulating over 20 mutations have been reported (Kožíšek et al., 2014; Sasková et al., 2009) (Fig. 5 ). The hallmark mutations are V11I, V32I, L33F, I47V, I50V, I54 L/M, G73S, L76V, I84V, and L89V (Tremblay, 2008).Fig. 5 A. Superimposition of the wild type (blue, pdb structure 4LL3) and a darunavir-resistant mutant of HIV-1 protease homodimer, pdb code 3TTP (Kožíšek et al., 2014). The 3TTP structure is grey, mutated positions are marked in green, RMSD between the structures was calculated as 0.49 Å. The inset shows the inhibitor in the substrate binding site using protein surface representation. The surface color denotes hydrophobic (grey) and hydrophilic (red and blue) atoms of amino acid residues. B. 2D interaction diagram of darunavir bound to the dimeric HIV-1 protease (pdb structure 3TTP). Mutated residues are marked with the green stars. The figures were generated using PyMol (Schrödinger and DeLano, 2020) and the PDBe application (https://www.ebi.ac.uk/pdbe/entry/pdb/3TTP). (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)
Fig. 5
3.1.1.4 Drugs with non-canonical mechanisms of action targeting HIV protease
Some non-nucleoside HIV reverse transcriptase inhibitors are also able to trigger premature release of HIV protease out of the Gag-Pol polyprotein. Rilpivirine, efavirenz, and etravirine can bind to the reverse transcriptase domain of the Gag-Pol precursor and enhance dimerization of two Gag-Pol molecules, bringing together their protease domains. This dimerization then results in autocatalytic release of HIV protease from the Gag-Pol precursor prior to assembly of new virions. This leads to decreased production of viral particles. At the same time, HIV protease is cytotoxic and its presence within the host cell could lead to elimination of infected cells harboring integrated proviral DNA (Jochmans et al., 2010; Kaplan and Swanstrom, 1991; Kräusslich, 1991; Majerová and Novotný, 2021; Pan et al., 2012; Sudo et al., 2013; Trinité et al., 2019; Wang et al., 2021). Although this phenomenon has been observed in cell culture experiments with such high doses of non-nucleoside HIV reverse transcriptase inhibitors that cannot be reached in vivo, it nevertheless represents an interesting proof-of-principle of a potential approach to causative cure of HIV infection.
3.1.1.5 Repurposing of HIV protease inhibitors
Some evidence suggests that HIV protease inhibitors could be beneficial for treatment of Kaposi sarcoma; however, the reports are contradictory and further research is needed (Lebbé et al., 1998; Palich et al., 2021; Subeha and Telleria, 2020). Nelfinavir has been studied for its potential anticancer effect (Allegra et al., 2020; Subeha and Telleria, 2020), as it interferes with the cell cycle (Brüning et al., 2010; Chow et al., 2006; Sun et al., 2012) and proteasomal degradation (Fassmannová et al., 2020; Gu et al., 2020); affects signal transduction pathways (Subeha and Telleria, 2020) and mitochondrial processes (Deng et al., 2010; Utkina-Sosunova et al., 2013); and induces cell death (Bissinger et al., 2015; Kawabata et al., 2012), endoplasmic reticulum stress (Okubo et al., 2018), and autophagy (Gills et al., 2008; Xia et al., 2019). Nelfinavir also interferes with late steps of adenovirus and herpesvirus replication in vitro, but it does not affect adenoviral cysteine protease or herpes serine protease (Georgi et al., 2020; Kalu et al., 2014). Similarly, lopinavir could act against papillomaviruses through its interactions with host targets (Barillari et al., 2018; Batman et al., 2011; Loharamtaweethong et al., 2019; Zehbe et al., 2011).
Recently, HIV protease inhibitors have been studied as potential anti-SARS-CoV-2 drugs. Even though SARS-CoV-2 harbors two proteases, neither has a specificity or mechanism of action resembling that of HIV protease. Thus, it is not probable that HIV protease inhibitors would inhibit SARS-CoV-2 proteases. Indeed, clinical trials assessing potential use of lopinavir-ritonavir among critically ill COVID-19 patients showed that this combination was not beneficial (Cao et al., 2020; Lecronier et al., 2020); in fact, they seemed to have adverse effect (Arabi et al., 2021).
However, there have been speculations that HIV PR inhibitors could affect a different step in the SARS-CoV-2 life cycle. Nelfinavir might act by blocking cell-cell fusion mediated by the spike protein (Yousefi et al., 2021). Amprenavir might destabilize the complex between the host ACE2 receptor and the SARS-CoV-2 spike protein and thus inhibit entry of the virus into the host cells (Buitrón-González et al., 2021). Viral entry might be also blocked by lopinavir and darunavir (Singh et al., 2020). No clinical data confirming these speculations are available at present.
4 Hepatitis C virus (HCV) and antiviral therapy
Flaviviruses are enveloped positive-sense single-stranded RNA viruses that possess a serine protease with a chymotrypsin-like fold embedded in the viral polyprotein. Once the host cell is infected by a flavivirus, the viral polyprotein is synthesized and intertwined across the membrane of the endoplasmic reticulum with the protease located on the cytoplasmic side (Pierson and Diamond, 2020). The protease is part of the multidomain protein NS3, which possesses proteolytic, NTPase, and helicase activity. To be fully active, NS3 needs an activating peptide (Erbel et al., 2006; Hilgenfeld et al., 2018; Kang et al., 2017; Lei et al., 2016; Luo et al., 2015; Majerová et al., 2019; Nie et al., 2021; Tomei et al., 1996). Although inhibitors of flaviviral proteases have been designed, no classical inhibitor of the NS2B/NS3 protease from the Flavivirus genus (such as Dengue, Zika, tick-born encephalitis virus or West-Nile virus) is currently undergoing the approval process. However, a closed derivative of JNJ-A07, a compound blocking protein-protein interactions between Dengue NS3 and NS4, has entered into clinical trials (Behnam and Klein, 2021, 2022; Kaptein et al., 2021).
The situation is different in the case of HCV (a flavivirus from the Hepacivirus genus), against which inhibitors have been developed and are successfully used clinically. HCV is a very important human pathogen: as many as 58 million people worldwide have chronic HCV infection and approximately 290 thousand of them die every year (WHO; https://www.who.int/news-room/fact-sheets/detail/hepatitis-c). Acute infection with HCV can be asymptomatic. While 25% of patients recover spontaneously, 75% develop chronic hepatitis C, which can lead to cirrhosis, hepatocellular carcinoma, or chronic system inflammatory problems (Pol and Lagaye, 2019). Seven HCV genotypes have been reported, differing in their susceptibility to antiviral drugs (Han et al., 2019; Smith et al., 2014).
HCV enters host cells by receptor-mediated endocytosis. After endosomal fusion, the viral particle is uncoated and viral RNA is released into the host cytoplasm. The 9.6-kb viral genomic RNA is used directly as a template for translation into a single viral polyprotein (Rosenberg, 2001). The nascent viral polyprotein is entwined through the membrane of the endoplasmic reticulum and co- and post-translationally processed by proteases on both sides of the membrane. Host proteases ensure processing of viral structural proteins on the luminal side, whereas viral proteases process the nonstructural (NS) proteins on the cytoplasmic side (Lohmann et al., 1996). The viral cysteine protease NS2 autocatalytically cleaves between NS2 and NS3 (Isken et al., 2022; Santolini et al., 1995); the viral serine protease NS3 ensures maturation of viral proteins expressed downstream of NS3, including NS5 replication complex. NS3, which harbors not only a protease but also a helicase, requires an activating cofactor, NS4A, for proteolytic activity. NS4A also anchors NS3 to the membrane of the endoplasmic reticulum (Hahm et al., 1995).
During viral RNA replication, the endoplasmic reticulum membranes undergo extensive rearrangement. Assembly of new viral particles is initiated at specific sites of rearranged membrane structures. Such membrane vesicles pass through the Golgi apparatus and are released by exocytosis (Alazard-Dany et al., 2019; Dustin et al., 2016; Li et al., 2021). Many host biomolecules are exploited during the HCV replication cycle, including cyclophilin A, liver-specific miRNAs, phosphatidylinositol 4-kinase IIIα, diacylglycerol O-acyltransferase 1, and endosomal sorting complexes required for transport (ESCRT) proteins (Manns et al., 2017).
As HCV is an RNA virus that does not require a DNA intermediate and thus does not integrate into the host cell genome, antivirals can cure HCV infection (in contrast with HIV infection). A combination of two or three direct-acting antiviral agents (DAAs) administered for 8–24 weeks cures HCV infection in more than 90% of patients (Manns et al., 2017).
Three classes of antivirals are currently in clinical use: NS5 replication complex inhibitors (daclatasvir, elbasvir, ledipasvir, ombitasvir, velpatasvir), NS5B RNA polymerase inhibitors (the nucleoside inhibitor sofosbuvir and the nonnucleoside inhibitor dasabuvir), and NS3/NS4A protease inhibitors (Cervino and Hynicka, 2018). Previous treatment regimens involved therapy using PEGylated interferons and the nucleoside analogue ribavirin (Palumbo, 2011). Several order-of-magnitude increase in the HCV mutation rate and shifts in the mutation spectrum were documented in viral genomes obtained from patients after six months of such a treatment indicating that lethal mutagenesis of the virus is one of the mechanisms of the antiviral effect (Cuevas et al., 2009).
The propensity of HCV to generate mutations is comparable to that of HIV. Thus, minimizing the risk of drug-resistance development became an important factor in therapeutic decision-making and drug design. HCV's RNA genome is replicated by NS5 RNA polymerase, which lacks a proofreading mechanism. The mutation rate is 2.5 × 10⁻⁵ mutations per nucleotide per genome replication (range 1.6–6.2 × 10⁻⁵) (Ribeiro et al., 2012), which translates to two to six mutations per each newly synthesized molecule of viral RNA. The mutations are not spread equally throughout the genome. Studies have indicated differences in mutability across genome sites of more than three orders of magnitude (Geller et al., 2016).
4.1 HCV NS3/NS4A protease inhibitors
Heterodimeric HCV NS3/NS4A protease adopts a chymotrypsin-like fold with the characteristic catalytic triad His-Asp-Ser. All first-generation HCV NS3/NS4A protease inhibitors bind to the active site of the enzyme. In contrast with the aspartic HIV protease, to which inhibitors bind only via non-covalent interactions (e.g. hydrogen bonds, hydrophobic interactions), inhibitors of the serine HCV NS3/NS4A protease bind both via non-covalent interactions and formation of a covalent bond with the catalytic serine. The substrate binding pocket of HCV NS3/NS4A protease is flat, not charged, and exposed to solvent. Moreover, the whole enzyme is structurally flexible and can adapt its conformation depending on the inhibitor structure (Kim et al., 1996; Kwong et al., 1998; Love et al., 1996). Interestingly, early work demonstrated that N-terminal cleavage products of substrate peptides act as potent inhibitors of HCV NS3/NS4A protease. This encouraged the design of inhibitors bound to P positions, with a warhead in the P1 position that covalently modifies the catalytic serine (Steinkühler et al., 1998). First-generation inhibitors of HCV NS3/NS4A protease include the linear compounds boceprevir, telaprevir, and narlaprevir and “a second wave first generation inhibitors” comprising the P1–P3 macrocycles simeprevir and danoprevir (Fig. 6 ) (Sarrazin et al., 2012). Boceprevir, telapreveir, and simeprevir are approved by the FDA, although newer second-generation drugs (see chapter 3.4) are currently preferred in clinical use (Zephyr et al., 2022). Narlaprevir and danoprevir have not been approved by the FDA, but they are used in eastern markets (Baker et al., 2021; Miao et al., 2020).Fig. 6 Structures of the first-generation inhibitors of HCV NS3/4A protease.
Fig. 6
The first-in-class inhibitor boceprevir (Victrelis®) was approved by the FDA in 2011 (Foote et al., 2011). It is a ketoamide that forms a reversible covalent bond with the catalytic serine of HCV NS3/NS4A protease. Boceprevir was obtained after extensive modifications of the lead ketoamide undecapeptide to maximize inhibition, minimize off-target effects, and optimize pharmacokinetic properties (Madison et al., 2008; Prongay et al., 2007; Venkatraman et al., 2006). The broad-spectrum serine protease neutrophil elastase was used as a model off-target enzyme; boceprevir inhibited HCV NS3/NS4A protease three orders of magnitude more potently than the elastase (Njoroge et al., 2008). Side effects of boceprevir include changes in blood counts, vomiting, changes in taste sensing, and chills (Marks and Jacobson, 2012).
Telaprevir (Incivek® and Incivo®) was also approved by the FDA in 2011 (Traynor, 2011). Like boceprevir, telaprevir is a peptidomimetic with a ketoamide warhead. Its structure was obtained after extensive optimization of all amino acid positions (Kwong et al., 2011; Perni et al., 2003, 2004a, 2004b, 2007) of the natural substrate. Side effects of telaprevir include skin and anorectal problems, gastrointestinal problems, and dysgeusia (Marcellin et al., 2011; Marks and Jacobson, 2012).
Despite differences in the structures of both inhibitors (see Fig. 6), the same patterns of drug-resistance mutations were identified in the S2 subsite of the substrate binding cavity of HCV NS3/NS4A protease: A156 T/S, T54S, V36M, and R155K (Howe and Venkatraman, 2013; Lin et al., 2004; Marks and Jacobson, 2012; Tong et al., 2006; Venkatraman, 2012). Nevertheless, the replicative capacity of telaprevir-resistant variants is lower than that of the wild type virus (Jiang et al., 2013).
The clinical use of boceprevir and telaprevir is limited by their adverse events, suboptimal pharmacokinetics, and low genetic barrier to drug-resistance development (even a single mutation can decrease the sensitivity of the protease to the inhibitor by one order of magnitude (Jiang et al., 2013). Moreover, only HCV genotype 1 is susceptible to these inhibitors (Dienstag, 2015). Because compounds with better features are now available, boceprevir and telaprevir have been withdrawn from the market.
Narlaprevir (Arlansa®) is active against HCV genotypes 1–3. Narlaprevir has a similar profile of drug-resistance development as boceprevir and telaprevir, but HCV harboring drug-resistant variants of NS3/NS4A protease retains some sensitivity to narlaprevir due to its tighter binding (Arasappan et al., 2010; Tong et al., 2010).
4.2 “Second wave” first-generation HCV NS3/NS4A protease inhibitors
The second wave drugs were obtained by cyclization of peptidomimetic inhibitors between the P1 and P3 residues. Cyclization increases the metabolic stability of the drug and decreases the variability of its conformational states which might result in tighter binding to the substrate cavity of the enzyme (Rosenquist et al., 2014).
Simeprevir (Olysiowas) was approved by the FDA in 2013. Simeprevir acts only against HCV genotype 1, but in comparison with other first-generation inhibitors, it has an improved pharmacokinetic profile, reduced side effects, and dramatically reduced drug-drug interactions (Lin et al., 2009; Raboisson et al., 2008; Reesink et al., 2010; You and Pockros, 2013). A polymorphic variant (Q80K) and the mutations A156 T/V, R155K, and D168 T/Y/H/A/V/I confer resistance to this drug (Shepherd et al., 2015).
Another compound developed by a similar strategy is danoprevir (Ganovo®). Like simeprevir, danoprevir occupies the S3–S1′subsites of the substrate binding pocket, but the interactions with the S4 subsite differ between the two compounds (Markham and Keam, 2018; Seiwert et al., 2008; Zephyr et al., 2021).
4.3 Second-generation HCV NS3/NS4A protease inhibitors
The second-generation HCV NS3/NS4A protease inhibitors were designed to act against a broader spectrum of natural HCV phenotypes and drug-resistant variants, as well as possess a higher barrier to drug-resistance development. The design was inspired by the substrate-envelope hypothesis, which proposes that interactions of an inhibitor with the enzyme binding site should not protrude beyond the substrate envelope, otherwise, the risk increases that a drug-resistant variant will evade inhibitor binding via a mutation outside the substrate envelope, while the interaction with natural substrates will be retained (Nalam et al., 2010). Therefore, the second-generation inhibitors contain a macrocyclic structure formed by connecting the P2 and P4 positions with a linker (Fig. 7 ), which ensures restricted conformational motion. The P2 position is locked in a suitable conformation and modified to block inhibitor contact with R155, as mutation at this position is known to give rise to drug-resistant variants. Instead, the interactions of the P2 position of inhibitors with the strictly conserved catalytic residues D75 and H57 are reinforced (Matthew et al., 2020; Taylor et al., 2019). The second-generation HCV NS3/NS4A protease inhibitors include grazoprevir, paritaprevir, glecaprevir, and voxilaprevir (Fig. 7, Fig. 8 ). Only grazoprevir, glecaprevir, and voxilaprevir are currently in clinical use (Zephyr et al., 2022).Fig. 7 Structures of the second-generation inhibitors of HCV NS3/4A protease.
Fig. 7
Fig. 8 A. Structure of HCV NS3/4A protease in the complex with voxilaprevir, pdb code 6NZT (Taylor et al., 2019). The activating peptide is highlighted in green, catalytic residues His 57, Asp 81 and Ser 139 are highlighted in cyan. The inset shows in detail the substrate binding cavity using surface representation with the inhibitor bound. The surface denotes hydrophobic (grey) and hydrophilic (red and blue) atoms of amino acid residues. B. 2D interaction diagram of voxilaprevir bound to HCV NS3/4A protease. The figures were generated using PyMol (Schrödinger and DeLano, 2020) and the PDBe application (https://www.ebi.ac.uk/pdbe/entry/pdb/6NZT). (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)
Fig. 8
Grazoprevir is typically used together with the NS5 replication complex inhibitor elbasvir in a combined formulation (Zepatier®), which was approved by the FDA in 2016 (Harper et al., 2012; Keating, 2016; Summa et al., 2012). It acts against genotypes 1 and 4 (Jacobson et al., 2017b; Zeuzem et al., 2015). The side effects of this therapy include fatigue, headache, and nausea. Even though grazoprevir has a high barrier to drug resistance development, mutations connected with treatment failure can evolve, including V36 L/M, Y56 F/H, Q80 K/L, R155I/K/L/S, A156 G/M/T/V, V158A, and D168 A/C/E/G/K/N/V (Sorbo et al., 2018).
Unlike other second-generation HCV NS3/NS4A inhibitors, paritaprevir is cyclized at positions P1–P3. This macrocyclic acylsulfonamide is used mainly against HCV genotype 1. Drug combinations involving paritaprevir were approved by the FDA in 2016 (Saab et al., 2016); however, they are not currently available on the market. Paritaprevir was a component of Technivie®, which contained ombitasvir (an inhibitor of viral RNA polymerase), paritaprevir, and ritonavir as a pharmacokinetic booster, and Viekira Pak®, which contained ombitasvir, paritaprevir, ritonavir, and dasabuvir. These treatments were effective and well-tolerated (Bacinschi et al., 2022; Chen et al., 2019; Lawitz et al., 2013). Drug resistance was associated with the mutations R155K and D168 N/Y/V (Boonma et al., 2019; Krishnan et al., 2015).
Glecaprevir, approved by the FDA in 2017, is a component of Mavyret®/Maviret®, which also contains the replicase complex inhibitor pibrentasvir (Mensa et al., 2019). It acts against all HCV genotypes, including genotype 3, which is not otherwise sensitive to inhibition (Gane et al., 2016a; Lawitz et al., 2013, 2015). Common drug-resistant HCV mutants are somewhat more sensitive to glecaprevir compared to other HCV protease inhibitors. Glecaprevir failure may occur in the presence of V36M, Y56 H/N, Q80 K/R, R155T, A156 G/T/V, and Q168 A/K/L/R mutations (Sorbo et al., 2018). Glecaprevir is well-tolerated (Lin et al., 2017; Park et al., 2021) and associated with a low frequency of adverse events, with serious adverse events reported in 0.8–7.5% of cases. One study indicated that therapy was stopped due to adverse events in 0.3–3.8% of cases (Cotter and Jensen, 2019), another study reported 1% of serious adverse events, leading in 0.6% to discontinuation of the therapy. For comparison, discontinuation of the therapy due to adverse events was one order of magnitude higher, e.g. 18% for boceprevir and telaprevir (Gordon et al., 2015).
Voxilaprevir (Fig. 7, Fig. 8), approved by the FDA in 2017, is a component of Vosevi®, which also includes the anti NS5 RNA polymerase drugs sofosbuvir and velpatasvir. It is active against multiple genotypes and is also useful in re-treatment regimens (Gane et al., 2016b, 2016c; Jacobson et al., 2017a; Lawitz et al., 2016, 2017; Rodriguez-Torres et al., 2016). Mutations associated with drug therapy failure include Y56F for the G1b phenotype and A166T for the G3a phenotype (Garcia-Cehic et al., 2021). The side effects include fatigue, headache, nausea, and diarrhea (Heo and Deeks, 2018).
4.4 Repurposing of HCV NS3/NS4A protease inhibitors
In vitro assays revealed that some HCV NS3/NS4A protease inhibitors also act against SARS-CoV-2 main protease. However, the inhibition constants (IC50 and EC50) of the HCV NS3/NS4A protease inhibitors against SARS-CoV-2 main protease were in the micromolar range in experiments with the purified enzyme and antiviral tissue culture assays (Baker et al., 2021; Ma et al., 2020; Oerlemans et al., 2021). This range is probably above the concentrations needed for effective and safe use of these compounds in clinical practice.
The HCV NS3/NS4A protease inhibitors simeprevir, vaniprevir, paritaprevir, and grazoprevir showed a similar potency against SARS-CoV-2 papain-like protease as against the main SARS-CoV-2 protease. Importantly, these four compounds exhibited a synergistic effect with remdesivir (an inhibitor of SARS-CoV-2 RNA polymerase) (Boras et al., 2021) in antiviral cell culture experiments. This effect was not observed for compounds that inhibit solely SARS-CoV-2 main protease (e.g., boceprevir) (Bafna et al., 2021). None of the inhibitors of HCV NS3/NS4A protease has been approved for treatment of COVID-19.
5 SARS-CoV-2 and antiviral therapy
Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is the third emerging coronavirus within the last twenty years. It was preceded by SARS-CoV (Qin et al., 2003) and Middle East Respiratory Syndrome Coronavirus (MERS-CoV) (van Boheemen et al., 2012). SARS-CoV was first identified in 2003 in China and spread to 29 countries, leading to 8,096 cases and 774 confirmed deaths. This epidemic was successfully eradicated by strict public health measures. MERS-CoV, originating from natural reservoirs in bats and transmitted via dromedary camels, was first identified in humans in 2012 in Saudi Arabia. All reported cases in humans have been related to residence in or travel to the Middle East, with 2,519 cases and 866 deaths from MERS-CoV infection reported by January 2020 (da Costa et al., 2020). The other human coronaviruses―HCoV-229E, HCoV-NL63, HCoV-OC43, and HCoV-HKU1―have been circulating in the human population for a long time and cause common cold (Corman et al., 2018).
The first SARS-CoV-2 infections were reported in Wuhan, China at the end of 2019 (Hui et al., 2020) The virus has spread worldwide, resulting in a pandemic that has led to the loss of several million lives and caused severe economic and social consequences. Several antigenic variants, differing in virulence and infectivity, have evolved over time. The Omicron variants that are currently dominant seem to be more contagious than previous variants but cause less serious illness that is more frequently restricted to the upper respiratory tract (Pia and Rowland-Jones, 2022; Shuai et al., 2022).
Like other coronaviruses, SARS-CoV-2 has a 30-kb single-stranded positive RNA genome (Bar-On et al., 2020). Viral structural proteins, among them the surface spike protein, are expressed from alternatively spliced mRNAs. Viral enzymes are translated as a shorter polyprotein called pp1a (harboring proteins nsp1–nsp11) and as the longer pp1ab polyprotein (harboring proteins nsp1–nsp16) after a frameshift suppressing the stop codon at the terminus of pp1a (Bhatt et al., 2021). This mechanism likely ensures an optimal ratio of viral non-structural proteins. Two proteases are embedded in pp1a: SARS-CoV-2 main protease (also known as chymotrypsin-like main protease 3CLpro or nsp5) and papain-like protease (PLpro), which is a part of a multifunctional multidomain protein called nsp3 (Klemm et al., 2020; Osipiuk et al., 2021). Both enzymes are cysteine proteases (Arya et al., 2021).
PLpro (nsp3) cleaves the first three nonstructural proteins of the viral polyprotein into the functional proteins nsp1, nsp2, and nsp3. This protease also cleaves ubiquitin-like interferon-stimulated gene 15 protein (ISG15) and has a deubiquitylating activity, which can disrupt the host interferon response and interfere with NF-κB pathways. In vitro, inhibition of SARS-CoV-2 PLpro by the compound GRL-0617, originally designed to inhibit SARS-CoV PLpro (Ratia et al., 2008), reduced virus-induced cytopathogenic effects, preserved the antiviral interferon response, and reduced production of viral particles in infected cells (Shin et al., 2020). As noted, some inhibitors of HCV NS3/NS4A protease are able to inhibit PLpro in vitro and act synergistically with other compounds (Bafna et al., 2021). However, no inhibitors of PLpro are currently in clinical use.
The second protease, SARS-CoV-2 main protease, cleaves the 13 remaining sites of the viral polyprotein (Jin et al., 2020; Zhang et al., 2020a). Although the polyprotein harbors nsp3, nsp4, and nsp6 proteins, which are intertwined across the membrane of endoplasmic reticulum, all of the cleavage sites are accessible from the cytosolic side (Majerová et al., 2019; V'Kovski et al., 2021).
Over the last two years, efforts to develop antivirals against SARS-CoV-2 have led to discoveries and FDA emergency use authorization or approvals of several drugs, including anti-spike specific monoclonal antibodies (Hwang et al., 2022) and the viral RNA polymerase inhibitors remdesivir (Veklury®) and molnupiravir (Lagevrio®). Remdesivir is administered intravenously, while molnupiravir is available orally (de Wit et al., 2020; Kabinger et al., 2021; Sheahan et al., 2017; Toots et al., 2019, 2020; Warren et al., 2016; Williamson et al., 2020). At the end of 2021, the FDA authorized nirmatrelvir (PF-07321332) (Owen et al., 2021) co-administered with the pharmacokinetic booster ritonavir (Paxlovid®). Nirmatrelvir is a specific inhibitor of SARS-CoV-2 main protease. In addition, the following immunomodulating compounds are currently available to treat COVID-19: baricitinib (Olumiant®, Janus kinase inhibitor), recently approved by both the FDA and the European Medicines Agency (EMA) (Rubin, 2022a), and anakinra (Kineret®, antagonist of interleukin-1), only approved by EMA (Atluri et al., 2022).
Interestingly, most of these drugs were obtained from drug repurposing programs. Remdesivir was originally studied as a potential anti-HCV and later anti-Ebola drug (Sheahan et al., 2017; Warren et al., 2016), and molnupiravir as a potential anti-influenza drug (Toots et al., 2019). Baricitinib and anakinra were originally approved for the treatment of chronic inflammatory diseases. Another compound with a similar mechanism of action as baricitinib―tofacitinib―has recently entered the authorization process (Ely et al., 2022; Zhang et al., 2021). Different supportive drugs such as corticoids and anticoagulants are also sometimes used to improve COVID-19 symptoms (Napoli et al., 2021).
Like other viruses, SARS-CoV-2 acquires mutations over time. The mutation rate of coronaviruses is lower than that of other RNA viruses, as their exonucleases are capable of proofreading (Denison et al., 2011; Liu et al., 2021). SARS-CoV-2 typically accumulates two point mutations per month, while the mutation rates of influenza and HIV are two- and four-times higher, respectively (Callaway, 2020), Importantly, viruses harboring mutations in the spike protein, which are responsible for overcoming acquired host immunity, do not harbor mutations in the viral enzymes exploited as drug targets. Indeed, even the Omicron variant, which has an increased ability to escape antibodies after previous immunizations, retains sensitivity to low-molecular-weight antivirals (Vangeel et al., 2022). As clinical use of novel antiviral drugs expands, resistant variants of SARS-CoV-2 may evolve under selection pressure. However, a potential trade-off between virulence and transmissibility (Blanquart et al., 2016; Simmonds et al., 2019) (together with a necessity to circumvent acquired immunity and/or to reduce susceptibility to antivirals) could favor the selection of milder viral variants.
5.1 SARS-CoV-2 main protease as an established therapeutic target
SARS-CoV-2 main protease (SARS-CoV-2 Mpro) is a key player in viral maturation and an important factor for SARS-CoV-2 pathogenicity (Pablos et al., 2021). SARS-CoV-2 Mpro was also exploited for a potential detection of the virus in clinical samples using activity-based probes (Rut et al., 2021). SARS-CoV-2 main protease is a cysteine protease with a catalytic dyad of Cys 145 and His 41 (Lee et al., 2020). The protease is embedded in viral polyproteins pp1a and pp1ab, from which it is autocatalytically released. Each SARS-CoV-2 Mpro monomer consists of three domains: rigid N-terminal domains I and II with a chymotrypsin-like fold and the more dynamic domain III. The protease is almost inactive in its monomeric form, and dimerization between the N-terminus of one monomer and domain II of a second monomer is required to establish a fully active enzyme. Binding of a ligand then stabilizes the active conformation (Jin et al., 2020; Zhang et al., 2020a). Interestingly, although the dimer harbors two active sites, only one monomeric subunit binds the substrate, and the monomeric subunits alternate in their activity through a flip-flop mechanism (Sheik Amamuddy et al., 2021). In contrast, SARS-CoV-2 main protease inhibitors bind to both monomeric subunits simultaneously, as observed in X-ray structures (Jin et al., 2020; Zhang et al., 2020a).
5.2 Inhibitors of SARS-CoV-2 main protease
Although allosteric binding sites in the SARS-CoV-2 main protease structure have been reported (El-Baba et al., 2020), most inhibitors being developed, including nirmatrelvir (PF-07321332) (Fig. 9, Fig. 10 ), target the active site of the enzyme (Chia et al., 2022; Owen et al., 2021). The development of nirmatrelvir was facilitated by prior efforts to target SARS-CoV that emerged in years 2002–2003. SARS-CoV main protease is 96% identical with the SARS-CoV-2 main protease (Zhang et al., 2020a). An inhibitor of SARS-CoV main protease PF-00835231 was developed by Pfizer. The α-hydroxymethylketone warhead of this compound forms a covalent reversible bond with the catalytic cysteine of the protease (Hoffman et al., 2020). Masking this reactive warhead by phosphorylation resulted in a prodrug lufotrelvir (PF-07304814) (Fig. 9), which is suitable for intravenous application. This prodrug is more soluble than the parental compound. The phosphate group is removed by host alkaline phosphatase, which is abundant in liver, lung and kidney. Lufotrelvir (PF-07304814) is now in clinical trials for treatment of COVID-19 in hospitalized patients (Boras et al., 2021).Fig. 9 Rational design of nirmatrelvir - the first FDA-authorized inhibitor of SARS-CoV-2 main protease (emergency use authorization). Nirmatrelvir shares similar features with lufotrelvir (a phosphate prodrug in the pipeline) and with boceprevir (an approved inhibitor of HCV NS3/NS4A protease).
Fig. 9
Fig. 10 A. A homodimeric structure of SARS-CoV-2 main protease with nirmatrelvir bound to one of the monomeric subunits (grey) with the inset showing in detail the nirmatrelvir using the substrate binding site of the protein surface representation. The surface denotes hydrophobic (grey) and hydrophilic (red and blue) atoms of amino acid residues. B. 2D interaction diagram of nirmatrelvir bound to the monomer of SARS-CoV-2 main protease (pdb 7RFW, (Owen et al., 2021). The figures were generated using PyMol (Schrödinger, 2020) and the PDBe application (https://www.ebi.ac.uk/pdbe/entry/pdb/7RFW). (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)
Fig. 10
The structure of nirmatrelvir was also inspired by PF-00835231 (Fig. 9) (Boras et al., 2021; Hoffman et al., 2020; Owen et al., 2021). Nirmatrelvir was optimized for the antiviral effect and for stability enabling oral administration. Similar to PF-00835231, the glutamine, which is essential at the P1 position of natural substrates, was replaced with the γ-lactam moiety (Dragovich et al., 1999). To reduce the chance of off-target binding, the α-hydroxymethylketone warhead of PF-00835231 was replaced with a nitrile group. The nitrile group retains its reactivity with the thiol group of the catalytic cysteine in the active-site of SARS-CoV-2 Mpro, but the binding is more reversible (Boike et al., 2022; Boras et al., 2021; Chuck et al., 2013; Owen et al., 2021). The P2 and P3 positions are formed by a dipeptide fragment of boceprevir. Boceprevir is an inhibitor of HCV protease approved for clinical use, and it showed moderate activity against SARS-CoV-2 in vitro (Fu et al., 2020; Ma et al., 2020; Madison et al., 2008; Prongay et al., 2007; Shekhar et al., 2022). Leucine in the P2 position is preferred in natural SARS-CoV-2 substrates (Rut et al., 2021). To preserve this structural feature in the inhibitor, a leucine mimetic harboring 6,6-dimethyl-3-azabicyclo[3.1.0]hexane structure was incorporated. This moiety retains hydrophobic interactions and is more conformationally rigid (Fu et al., 2020). The trifluoromethyl group in the P4 position increases hydrophobic interaction with the substrate binding pocket, membrane permeability and metabolic stability (Gillis et al., 2015; Owen et al., 2021).
A combination of nirmatrelvir with the pharmacokinetic booster ritonavir (Paxlovid®) was authorized by the FDA for treatment of COVID-19 in late 2021. Clinical studies showed that when nirmatrelvir/ritonavir therapy was initiated 3 or 5 days after diagnosis, the probability of hospitalization was reduced 8.8- and 6.7-fold, respectively (Couzin-Frankel, 2021). Co-administration of ritonavir with nirmatrelvir is necessary to achieve a plasma concentration of nirmatrelvir sufficient for an efficient antiviral effect (Lamb, 2022). Even though ritonavir is a well-established pharmacokinetic booster, caution is required when using it in combination with nirmatrelvir (Heskin et al., 2022). Several adverse events have been reported among people using this therapy, with dysgeusia, diarrhea, hypertension, and myalgia being the most frequent. Additionally, ritonavir is not only a potent inhibitor of CYP3A4, but also shows inhibitory effects on other isoenzymes of cytochrome P450 (namely CYP3A4, CYP2D6, CYP2C19, CYP2C8, and CYP2C9) as well as on the multidrug resistance transporter ABCB5 P- glycoprotein, breast cancer resistance protein (ABCG2), the liver organic anion transporter (hOCT1), and the drug efflux mediating protein MATE1. Moreover, ritonavir induces isoenzymes CYP1A2, CYP2B6, CYP2C9, and CYP2C19. It is thus not surprising that this compound has significant drug-drug interactions, including with anticoagulants, antiarrhythmics, statins, steroids, and sedative hypnotics. Many of these agents are prescribed separately to elderly patients who are at the greatest risk of developing complications from SARS-CoV-2 infection (Heskin et al., 2022). Patients under long-term medication containing a pharmacokinetic booster (ritonavir or cobicistat) can use nirmatrelvir/ritonavir, if necessary (Marzolini et al., 2022).
Nirmatrelvir is a reversible covalent inhibitor with a nitrile warhead targeting the catalytic Cys 145 of SARS-CoV-2 Mpro (Fig. 9, Fig. 10). Upon binding, the nitrile group of the inhibitor forms a thioimidate adduct with Cys 145; the binding is stabilized by a hydrogen bond with the backbone of Gly 143. The (S)-γ-lactam ring of the glutamine surrogate in the P1 position of the inhibitor forms hydrogen bonds with the side chains of His 163 and Glu 166 and the main chain of His 164. A conformationally constrained hydrophobic dimethylcyclopropylproline group in the P2 position interacts with the side chains of His 41, Met 49, Tyr 54, Met 165, and Asp 189, and with the main chains of Asp 187 and Arg 188. The tert-butyl group in the P3 position of the inhibitor is exposed to solvent and has limited interactions with Mpro. The amide nitrogen of the trifluoroacetyl group in the P4 position of the inhibitor forms a hydrogen bond with the backbone carbonyl oxygen of Glu 166. Moreover, the trifluoromethyl group forms hydrogen bonds with Gln 192 and binds two ordered water molecules, which stabilizes the interaction with the S4 subsite of Mpro (Zhao et al., 2021) (Fig. 9, Fig. 10). The contacts between the inhibitor and the protein backbone are important, as they reduce the risk of drug-resistance development. Nevertheless, (see above) nirmatrelvir forms several contacts outside the “substrate envelope” with Asn 142 at P1 position, with Asp 187, Gln 189, Thr 190, Gln 192 at P2 position and with Met 165 in P3 position (Shaqra et al., 2022). For potential development of drug resistance were predicted residues in regions 45–51 and 186–192. Such mutations were found in sequences of clinical viral isolates, but they were not phenotypically evaluated (Yang et al., 2022). In vitro selection revealed three mutations that conferred resistance to nirmatrelvir: L50F, E166 A/V and L167F. (Jochmans et al., 2022; Zhou et al., 2022). Furthermore, although the E166V mutation resulted in a reduction of sensitivity to nirmatrelvir by two orders of magnitude, this mutation also resulted in viruses bearing it losing replicative fitness. The fitness was restored by compensatory mutations L50F and T21I (Iketani et al., 2022). Mpro mutants of all common viral variants found in patients were susceptible to inhibition by nirmatrelvir (Greasley et al., 2022).
Although Paxlovid® reduces hospitalizations and deaths (Wen et al., 2022), a rebound phenomenon has been reported. In some patients, the virus is eradicated during the course of Paxlovid® therapy, but several days after the therapy is discontinued, the patient tests positive for SARS-CoV-2 infection again (Burki, 2022; Coulson et al., 2022). No drug-resistant variants have appeared, and no antibody-evading spike mutations have been identified in samples from such patients (Carlin et al., 2022). There are hypotheses that this phenomenon results from a suboptimal dose of the drug or that the early onset of the therapy blocks an efficient immune response (Rubin, 2022b). Inter and extracellular membrane structures containing the virus could play a role in virus and viral fragments shedding (e.g. preserving intact genomic viral RNA) or even in spread via intercellular transport pathways acting as a “Trojan horse”. Such viruses could be reactivated after finishing of the antiviral therapy (Badierah et al., 2021; Borowiec et al., 2021; Merolli et al., 2022). Further investigation of the rare rebound phenomenon is required.
Combinations of drugs targeting different steps of the viral life cycle have proven to be efficient in the therapy of viral infections. For SARS-CoV-2, possible combinations include nirmatrelvir with either molnupiravir or remdesivir. A study has shown improvement in survival of SARS-CoV-2 infected mice under such experimental combination therapies (Jeong et al., 2022), but clinical data are needed. Co-administration of these drugs could synergize some adverse drug reactions.
For individual treatment optimization, in terms of circumventing development of SARS-CoV-2 drug resistance and obtaining drugs with acceptable side effects and drug-drug interactions as well as with improved plasma stability without the necessity to use pharmacokinetics boosters, the availability of additional Mpro inhibitors would be beneficial. Currently, several Mpro inhibitors are in the pipeline (Fig. 9, Fig. 11 ), including lufotrelvir (PF-07304814), a peptidomimetic prodrug intended for intravenous application; ensitrelvir (S-217622, Xocova®) (Sasaki et al., 2022; Unoh et al., 2022); PBI-0451 (Pardesbio; https://www.pardesbio.com/pipeline/) as well as 13b (Cully, 2022; Zhang et al., 2020a); and tollovir (Todosmedical; https://todosmedical.com/tollovir). As main protease is well-conserved (Yang et al., 2022), inhibitors of SARS-CoV-2 Mpro may be at least partially active against new SARS-CoV-2 variants, and could be useful for repurposing against potential novel emerging coronaviruses.Fig. 11 Structures of compounds in the pipeline.
Fig. 11
Computer aided design speeds up drug development, reduces experimental work and makes processes more cost-effective. Approaches, such as structure-based and ligand-based virtual screening, help to identify novel drugs, to optimize leading compounds (da Silva Rocha et al., 2019; Evenseth et al., 2020; Proia et al., 2022) or can help in drug-repurposing strategies (Gan et al., 2022; Mody et al., 2021). Positive hits identified, must be verified experimentally. Even inhibition of a target in an experiment does not necessarily mean clinically relevant inhibition.
For design of novel structures an innovative approach “Virtual SYNThhon Hierarchical Enumeration Screening” (V-SYNTHES) has been employed: final molecules are obtained after several iterative steps of screening of small fragment-like molecules followed by building a focused library of larger compounds combining fragment hits as building blocks (Sadybekov et al., 2022). Machine learning further boosts drug design. Here, an algorithm searches for general structural patterns of desired pharmacodynamical and/or pharmacokinetic properties in known molecules and tries to transfer these properties into structures of newly designed compounds (Graff et al., 2021; Priya et al., 2022; Tanramluk et al., 2022; Zhang and Lee, 2019; Zhang et al., 2022). All the artificial intelligence (AI) approaches need experimental data sets for a starting input as well as for validation (e.g., X-ray structures, inhibition constants). High quality and quantity of experimental data boost the future success of AI in drug design.
AI has been involved in the COVID Moonshot initiative for non-profit patent-free development of antivirals, when non-covalent SARS-CoV-2 inhibitors were designed (Achdout et al., 2022; Morris et al., 2021). Compounds obtained by the Moonshot initiative and non-covalent inhibitors designed earlier against SARS-CoV were helpful in construction of pharmacophores in the binding site of SARS-CoV-2 main protease and subsequent virtual screening followed by biological screening. This effort led to the identification of compound S-217622, which is currently being approved in Japan (Fig. 11) (Han et al., 2022; Jacobs et al., 2013; Unoh et al., 2022).
6 Important viral proteases for future targeting
Even though the field of antiviral research has been rapidly advancing, there are viral infections with limited or no therapy options. Indeed, no inhibitors of SARS-CoV-2 papain like protease (PLpro), herpes serine proteases, or poxviral protease are currently in clinical use, although such drugs would increase the possibility of individual treatment optimization. Moreover, no treatments for infections caused by the flaviviruses Dengue virus, Zika virus, West Nile virus, yellow fever virus, and tick-born encephalitis virus are available. Efforts are underway to develop such drug candidates.
SARS-CoV-2 papain like protease (SARS-CoV-2 PLpro) plays an important role in SARS-CoV-2 maturation and contributes to evasion of the host antiviral immune response. It is embedded in the large multifunctional multidomain membrane protein nsp3, which is released from the pp1a and pp1ab precursor by the PLpro domain (Klemm et al., 2020; Lei et al., 2018). SARS-CoV-2 PLpro contains a catalytic cysteine and four conserved non-active-site cysteines, which coordinate a zinc cation. Clinical availability of SARS-CoV-2 PLpro inhibitors would be beneficial, as they could reduce viral replication and support an innate immune response to virus-induced cytopathogenic effects (Osipiuk et al., 2021). Design of specific inhibitors targeting the active site of the enzyme might be problematic due to substrate specificity similarities with host enzymes, particularly deubiquitinylases. However, highly specific active-site targeted inhibitors have been designed (Rut et al., 2020). Compounds with alternative binding modes could represent another interesting alternative. For instance, some inhibitors reported to bind SARS-CoV-2 PLpro and showing no serious cytotoxicity (including GRL-0617) bind atypically to the substrate cavity, as they interact with the P3 and P4 positions and omit the catalytic site (Lv et al., 2022). In addition, allosteric compounds might have a reduced risk of off-target binding (Armstrong et al., 2021; Majerová and Novotný, 2021). Flaviviral proteases (viruses like Dengue, Zika, West-Nile, tick-born encephalitis) have some similar structural and enzymological features that could be exploited for design of panflaviviral inhibitors (Akaberi et al., 2021). On the other hand, less specific compounds acting against a wide range of viruses might inadvertently damage the host virome (De Vlaminck et al., 2013; Liang and Bushman, 2021).
Studies of autoprocessing of Dengue protease revealed a trans-dominant inhibitory effect of unprocessed precursors on production of infectious viral particles. This trans-dominant effect (i.e. blocking active enzyme molecules via interactions with inactive enzyme species) could be achieved either by partial inhibition of polyprotein processing by a PR inhibitor, or by co-infection of the cells by an autoprocessing defective mutant. This effect could slowdown potential drug resistance development during treatment of flavivirus infections using specific protease inhibitors (Constant et al., 2018; Majerová and Novotný, 2021).
Herpes viruses are enveloped double-stranded DNA viruses that cause a wide range of diseases including cold sores, varicella, mononucleosis, and meningoencephalitis. Nucleoside derivatives targeting synthesis of herpesviral DNA are in clinical use (De Clercq, 2004). Herpes serine proteases have been extensively studied, but no inhibitors of these interesting homodimeric enzymes have entered clinical use (Gable et al., 2016; Khayat et al., 2003; Lee et al., 2011; Shimba et al., 2004; Tong et al., 1996, 1998).
Poxviruses such as monkeypox and smallpox viruses, also possess a protease that could be pharmaceutically targeted (Aleshin et al., 2012; Bunge et al., 2022). Recent outburst of the monkeypox epidemy (declared by WHO in July 2022 as global emergency) represents very strong incentive for the research of novel antivirals based on PR inhibition.
7 Conclusion
Viral proteases are suitable targets for therapeutic intervention due to their essential roles in viral life cycles and clearly represent one of the first and most convincing examples of the power of the structure-assisted drug design. Among individual viral families, proteases differ in mechanism of action and specificity. Viral protease inhibitors are established components of HIV and HCV therapies, and the first protease inhibitor for treatment of SARS-CoV-2 infection was recently authorized, promising major breakthrough in the treatment of this pandemic.
All viral protease inhibitors in clinical use bind to the active site of the enzyme. To efficiently block viral maturation and reduce off-target binding, which could lead to adverse drug reactions, interaction between the inhibitor and enzyme must be tight and highly specific. For these reasons, successful repurposing of a viral protease inhibitor to inhibit a protease from an unrelated viral species might not be possible. Nevertheless, given the wide range of clinical experience and drug approvals, these compounds potentially could be repurposed for treatment of other viral diseases by targeting other steps in the viral life cycle or for treatment of noninfectious diseases, such as cancer.
In chronic infections, inhibitors of viral proteases are usually co-administered with antivirals with different mechanisms of action to increase the efficiency of the therapy and reduce the probability of development of drug-resistant mutants.
Due to continual development of drug-resistant viral mutants under the selection pressure of drugs and sometimes-intolerable drug side effects, novel antivirals are still in demand. For targeting of viral proteases, not only active-site inhibitors, but also allosteric modulators may open the door to new therapeutic options.
Acknowledgment
We want to thank Barbora Marešová, Blanka Collis, Tereza Ormsby, and Hillary Hoffman for useful comments, manuscript editing and language corrections. This work was supported by the National Institute of virology and bacteriology (Programme EXCELES, ID Project No. LX22NPO5103) - Funded by the European Union - Next Generation EU.
==== Refs
References
Achdout H. Aimon A. Bar-David E. Barr H. Ben-Shmuel A. Bennett J. Bilenko V.A. Bilenko V.A. Boby M.L. Borden B. Bowman G.R. Brun J. Bvnbs S. Calmiano M. Carbery A. Carney D. Cattermole E. Chang E. Chernyshenko E. Chodera J.D. Clyde A. Coffland J.E. Cohen G. Cole J. Contini A. Cox L. Cvitkovic M. Dias A. Donckers K. Dotson D.L. Douangamath A. Duberstein S. Dudgeon T. Dunnett L. Eastman P.K. Erez N. Eyermann C.J. Fairhead M. Fate G. Fearon D. Fedorov O. Ferla M. Fernandes R.S. Ferrins L. Foster R. Foster H. Gabizon R. Garcia-Sastre A. Gawriljuk V.O. Gehrtz P. Gileadi C. Giroud C. Glass W.G. Glen R. Glinert I. Godoy A.S. Gorichko M. Gorrie-Stone T. Griffen E.J. Hart S.H. Heer J. Henry M. Hill M. Horrell S. Huliak V.D. Hurley M.F.D. Israely T. Jajack A. Jansen J. Jnoff E. Jochmans D. John T. Jonghe S.D. Kantsadi A.L. Kenny P.W. Kiappes J.L. Kinakh S.O. Koekemoer L. Kovar B. Krojer T. Lee A. Lefker B.A. Levy H. Logvinenko I.G. London N. Lukacik P. Macdonald H.B. MacLean B. Malla T.R. Matviiuk T. McCorkindale W. McGovern B.L. Melamed S. Melnykov K.P. Michurin O. Mikolajek H. Milne B.F. Morris A. Morris G.M. Morwitzer M.J. Moustakas D. Nakamura A.M. Neto J.B. Neyts J. Nguyen L. Noske G.D. Oleinikovas V. Oliva G. Overheul G.J. Owen D. Pai R. Pan J. Paran N. Perry B. Pingle M. Pinjari J. Politi B. Powell A. Psenak V. Puni R. Rangel V.L. Reddi R.N. Reid S.P. Resnick E. Ripka E.G. Robinson M.C. Robinson R.P. Rodriguez-Guerra J. Rosales R. Rufa D. Saar K. Saikatendu K.S. Schofield C. Shafeev M. Shaikh A. Shi J. Shurrush K. Singh S. Sittner A. Skyner R. Smalley A. Smeets B. Smilova M.D. Solmesky L.J. Spencer J. Strain-Damerell C. Swamy V. Tamir H. Tennant R. Thompson W. Thompson A. Tomasio S. Tsurupa I.S. Tumber A. Vakonakis I. van Rij R.P. Vangeel L. Varghese F.S. Vaschetto M. Vitner E.B. Voelz V. Volkamer A. von Delft F. von Delft A. Walsh M. Ward W. Weatherall C. Weiss S. White K.M. Wild C.F. Wittmann M. Wright N. Yahalom-Ronen Y. Zaidmann D. Zidane H. Zitzmann N. Open Science Discovery of Oral Non-covalent SARS-CoV-2 Main Protease Inhibitor Therapeutics 2022 bioRxiv 10.1101/2020.10.29.339317 2020.2010.2029.339317
Akaberi D. Båhlström A. Chinthakindi P.K. Nyman T. Sandström A. Järhult J.D. Palanisamy N. Lundkvist Å. Lennerstrand J. Targeting the NS2B-NS3 protease of tick-borne encephalitis virus with pan-flaviviral protease inhibitors Antivir. Res. 190 2021 105074 10.1016/j.antiviral.2021.105074
Alazard-Dany N. Denolly S. Boson B. Cosset F.-L. Overview of HCV life cycle with a special focus on current and possible future antiviral targets Viruses 11 1 2019 30 10.3390/v11010030 30621318
Aleshin A.E. Drag M. Gombosuren N. Wei G. Mikolajczyk J. Satterthwait A.C. Strongin A.Y. Liddington R.C. Salvesen G.S. Activity, specificity, and probe design for the smallpox virus protease K7L J. Biol. Chem. 287 47 2012 39470 39479 10.1074/jbc.M112.388678 23012361
Allegra A. Innao V. Allegra A.G. Pulvirenti N. Pugliese M. Musolino C. Antitumorigenic action of nelfinavir: effects on multiple myeloma and hematologic malignancies (Review) Oncol. Rep. 43 6 2020 1729 1736 10.3892/or.2020.7562 32236596
Alvi R.M. Neilan A.M. Tariq N. Awadalla M. Afshar M. Banerji D. Rokicki A. Mulligan C. Triant V.A. Zanni M.V. Neilan T.G. Protease inhibitors and cardiovascular outcomes in patients with HIV and heart failure J. Am. Coll. Cardiol. 72 5 2018 518 530 10.1016/j.jacc.2018.04.083 30049313
Aoki M. Venzon D.J. Koh Y. Aoki-Ogata H. Miyakawa T. Yoshimura K. Maeda K. Mitsuya H. Non-cleavage site gag mutations in amprenavir-resistant human immunodeficiency virus type 1 (HIV-1) predispose HIV-1 to rapid acquisition of amprenavir resistance but delay development of resistance to other protease inhibitors J. Virol. 83 7 2009 3059 3068 10.1128/jvi.02539-08 19176623
Aoki M. Danish M.L. Aoki-Ogata H. Amano M. Ide K. Das D. Koh Y. Mitsuya H. Loss of the protease dimerization inhibition activity of tipranavir (TPV) and its association with the acquisition of resistance to TPV by HIV-1 J. Virol. 86 24 2012 13384 13396 10.1128/jvi.07234-11 23015723
Arabi Y.M. Gordon A.C. Derde L.P.G. Nichol A.D. Murthy S. Beidh F.A. Annane D. Swaidan L.A. Beane A. Beasley R. Berry L.R. Bhimani Z. Bonten M.J.M. Bradbury C.A. Brunkhorst F.M. Buxton M. Buzgau A. Cheng A. De Jong M. Detry M.A. Duffy E.J. Estcourt L.J. Fitzgerald M. Fowler R. Girard T.D. Goligher E.C. Goossens H. Haniffa R. Higgins A.M. Hills T.E. Horvat C.M. Huang D.T. King A.J. Lamontagne F. Lawler P.R. Lewis R. Linstrum K. Litton E. Lorenzi E. Malakouti S. McAuley D.F. McGlothlin A. McGuinness S. McVerry B.J. Montgomery S.K. Morpeth S.C. Mouncey P.R. Orr K. Parke R. Parker J.C. Patanwala A.E. Rowan K.M. Santos M.S. Saunders C.T. Seymour C.W. Shankar-Hari M. Tong S.Y.C. Turgeon A.F. Turner A.M. Van de Veerdonk F.L. Zarychanski R. Green C. Berry S. Marshall J.C. McArthur C. Angus D.C. Webb S.A. Lopinavir-ritonavir and hydroxychloroquine for critically ill patients with COVID-19: REMAP-CAP randomized controlled trial Intensive Care Med. 47 8 2021 867 886 10.1007/s00134-021-06448-5 34251506
Arasappan A. Bennett F. Bogen S.L. Venkatraman S. Blackman M. Chen K.X. Hendrata S. Huang Y. Huelgas R.M. Nair L. Padilla A.I. Pan W. Pike R. Pinto P. Ruan S. Sannigrahi M. Velazquez F. Vibulbhan B. Wu W. Yang W. Saksena A.K. Girijavallabhan V. Shih N.Y. Kong J. Meng T. Jin Y. Wong J. McNamara P. Prongay A. Madison V. Piwinski J.J. Cheng K.C. Morrison R. Malcolm B. Tong X. Ralston R. Njoroge F.G. Discovery of narlaprevir (SCH 900518): a potent, second generation HCV NS3 serine protease inhibitor ACS Med. Chem. Lett. 1 2 2010 64 69 10.1021/ml9000276 24900178
Armstrong L.A. Lange S.M. Dee Cesare V. Matthews S.P. Nirujogi R.S. Cole I. Hope A. Cunningham F. Toth R. Mukherjee R. Bojkova D. Gruber F. Gray D. Wyatt P.G. Cinatl J. Dikic I. Davies P. Kulathu Y. Biochemical characterization of protease activity of Nsp3 from SARS-CoV-2 and its inhibition by nanobodies PLoS One 16 7 2021 e0253364 10.1371/journal.pone.0253364
Arya R. Kumari S. Pandey B. Mistry H. Bihani S.C. Das A. Prashar V. Gupta G.D. Panicker L. Kumar M. Structural insights into SARS-CoV-2 proteins J. Mol. Biol. 433 2 2021 166725 10.1016/j.jmb.2020.11.024
Atluri K. Aimlin I. Arora S. Current effective therapeutics in management of COVID-19 J. Clin. Med. 11 13 2022 10.3390/jcm11133838
Bacinschi X. Popescu G.C. Zgura A. Gales L. Rodica A. Mercan A. Serban D. Haineala B. Toma L. Iliescu L. A real-world study to compare the safety and efficacy of paritaprevir/ombitasvir/ritonavir and dasabuvir, with or without ribavirin, in 587 patients with chronic hepatitis C at the fundeni clinical Institute, bucharest, Romania Med. Sci. Mon. Int. Med. J. Exp. Clin. Res. 28 2022 e936706 10.12659/msm.936706
Badierah R.A. Uversky V.N. Redwan E.M. Dancing with Trojan horses: an interplay between the extracellular vesicles and viruses J. Biomol. Struct. Dyn. 39 8 2021 3034 3060 10.1080/07391102.2020.1756409 32351170
Bafna K. White K. Harish B. Rosales R. Ramelot T.A. Acton T.B. Moreno E. Kehrer T. Miorin L. Royer C.A. García-Sastre A. Krug R.M. Montelione G.T. Hepatitis C virus drugs that inhibit SARS-CoV-2 papain-like protease synergize with remdesivir to suppress viral replication in cell culture Cell Rep. 35 7 2021 109133 10.1016/j.celrep.2021.109133
Baker J.D. Uhrich R.L. Kraemer G.C. Love J.E. Kraemer B.C. A drug repurposing screen identifies hepatitis C antivirals as inhibitors of the SARS-CoV2 main protease PLoS One 16 2 2021 e0245962 10.1371/journal.pone.0245962
Bar-On Y.M. Flamholz A. Phillips R. Milo R. SARS-CoV-2 (COVID-19) by the numbers Elife 9 2020 e57309 10.7554/eLife.57309
Barillari G. Monini P. Sgadari C. Ensoli B. The impact of human papilloma viruses, matrix metallo-proteinases and HIV protease inhibitors on the onset and progression of uterine cervix epithelial tumors: a review of preclinical and clinical studies Int. J. Mol. Sci. 19 5 2018 10.3390/ijms19051418
Barré-Sinoussi F. Chermann J.C. Rey F. Nugeyre M.T. Chamaret S. Gruest J. Dauguet C. Axler-Blin C. Vézinet-Brun F. Rouzioux C. Rozenbaum W. Montagnier L. Isolation of a T-lymphotropic retrovirus from a patient at risk for acquired immune deficiency syndrome (AIDS) Science 220 4599 1983 868 871 10.1126/science.6189183 6189183
Batman G. Oliver A.W. Zehbe I. Richard C. Hampson L. Hampson I.N. Lopinavir up-regulates expression of the antiviral protein ribonuclease L in human papillomavirus-positive cervical carcinoma cells Antivir. Ther. 16 4 2011 515 525 10.3851/imp1786 21685539
Bean B. Antiviral therapy: current concepts and practices Clin. Microbiol. Rev. 5 2 1992 146 182 10.1128/cmr.5.2.146 1576586
Behnam M.A.M. Klein C.D. On track to tackle dengue: history and future of NS4B ligands Cell Host Microbe 29 12 2021 1735 1737 10.1016/j.chom.2021.11.010 34883063
Behnam M.A.M. Klein C.D. Corona versus dengue: distinct mechanisms for inhibition of polyprotein processing by antiviral drugs ACS Pharmacol Transl Sci 5 7 2022 508 511 10.1021/acsptsci.2c00105 35837137
Bhatt P.R. Scaiola A. Loughran G. Leibundgut M. Kratzel A. Meurs R. Dreos R. O'Connor K.M. McMillan A. Bode J.W. Thiel V. Gatfield D. Atkins J.F. Ban N. Structural basis of ribosomal frameshifting during translation of the SARS-CoV-2 RNA genome Science 372 6548 2021 1306 1313 10.1126/science.abf3546 34029205
Bissinger R. Waibel S. Lang F. Induction of suicidal erythrocyte death by nelfinavir Toxins 7 5 2015 1616 1628 10.3390/toxins7051616 26008229
Blanquart F. Grabowski M.K. Herbeck J. Nalugoda F. Serwadda D. Eller M.A. Robb M.L. Gray R. Kigozi G. Laeyendecker O. Lythgoe K.A. Nakigozi G. Quinn T.C. Reynolds S.J. Wawer M.J. Fraser C. A transmission-virulence evolutionary trade-off explains attenuation of HIV-1 in Uganda Elife 5 2016 e20492 10.7554/eLife.20492
Boesecke C. Cooper D.A. Toxicity of HIV protease inhibitors: clinical considerations Curr. Opin. HIV AIDS 3 6 2008 653 659 10.1097/COH.0b013e328312c392 19373038
Boike L. Henning N.J. Nomura D.K. Advances in covalent drug discovery Nat. Rev. Drug Discov. 2022 1 18 10.1038/s41573-022-00542-z
Bold G. Fässler A. Capraro H.G. Cozens R. Klimkait T. Lazdins J. Mestan J. Poncioni B. Rösel J. Stover D. Tintelnot-Blomley M. Acemoglu F. Beck W. Boss E. Eschbach M. Hürlimann T. Masso E. Roussel S. Ucci-Stoll K. Wyss D. Lang M. New aza-dipeptide analogues as potent and orally absorbed HIV-1 protease inhibitors: candidates for clinical development J. Med. Chem. 41 18 1998 3387 3401 10.1021/jm970873c 9719591
Boonma T. Nutho B. Rungrotmongkol T. Nunthaboot N. Understanding of the drug resistance mechanism of hepatitis C virus NS3/4A to paritaprevir due to D168N/Y mutations: a molecular dynamics simulation perspective Comput. Biol. Chem. 83 2019 107154 10.1016/j.compbiolchem.2019.107154
Boras B. Jones R.M. Anson B.J. Arenson D. Aschenbrenner L. Bakowski M.A. Beutler N. Binder J. Chen E. Eng H. Hammond H. Hammond J. Haupt R.E. Hoffman R. Kadar E.P. Kania R. Kimoto E. Kirkpatrick M.G. Lanyon L. Lendy E.K. Lillis J.R. Logue J. Luthra S.A. Ma C. Mason S.W. McGrath M.E. Noell S. Obach R.S. Mn O.B. O'Connor R. Ogilvie K. Owen D. Pettersson M. Reese M.R. Rogers T.F. Rosales R. Rossulek M.I. Sathish J.G. Shirai N. Steppan C. Ticehurst M. Updyke L.W. Weston S. Zhu Y. White K.M. García-Sastre A. Wang J. Chatterjee A.K. Mesecar A.D. Frieman M.B. Anderson A.S. Allerton C. Preclinical characterization of an intravenous coronavirus 3CL protease inhibitor for the potential treatment of COVID19 Nat. Commun. 12 1 2021 6055 10.1038/s41467-021-26239-2 34663813
Borowiec B.M. Angelova Volponi A. Mozdziak P. Kempisty B. Dyszkiewicz-Konwińska M. Small extracellular vesicles and COVID19-using the "trojan horse" to tackle the giant Cells 10 12 2021 10.3390/cells10123383
Broder S. The development of antiretroviral therapy and its impact on the HIV-1/AIDS pandemic Antivir. Res. 85 1 2010 1 18 10.1016/j.antiviral.2009.10.002 20018391
Brüning A. Friese K. Burges A. Mylonas I. Tamoxifen enhances the cytotoxic effects of nelfinavir in breast cancer cells Breast Cancer Res. 12 4 2010 R45 10.1186/bcr2602 20594311
Buitrón-González I. Aguilera-Durán G. Romo-Mancillas A. In-silico drug repurposing study: amprenavir, enalaprilat, and plerixafor, potential drugs for destabilizing the SARS-CoV-2 S-protein-angiotensin-converting enzyme 2 complex Results Chem 3 2021 100094 10.1016/j.rechem.2020.100094
Bunge E.M. Hoet B. Chen L. Lienert F. Weidenthaler H. Baer L.R. Steffen R. The changing epidemiology of human monkeypox-A potential threat? A systematic review PLoS Neglected Trop. Dis. 16 2 2022 e0010141 10.1371/journal.pntd.0010141
Burki T. The future of Paxlovid for COVID-19 Lancet Respir. Med. 10 7 2022 e68 10.1016/s2213-2600(22)00192-8 35623373
Busti A.J. Hall R.G. Margolis D.M. Atazanavir for the treatment of human immunodeficiency virus infection Pharmacotherapy 24 12 2004 1732 1747 10.1592/phco.24.17.1732.52347 15585441
Callaway E. The coronavirus is mutating - does it matter? Nature 585 7824 2020 174 177 10.1038/d41586-020-02544-6 32901123
Cao B. Wang Y. Wen D. Liu W. Wang J. Fan G. Ruan L. Song B. Cai Y. Wei M. Li X. Xia J. Chen N. Xiang J. Yu T. Bai T. Xie X. Zhang L. Li C. Yuan Y. Chen H. Li H. Huang H. Tu S. Gong F. Liu Y. Wei Y. Dong C. Zhou F. Gu X. Xu J. Liu Z. Zhang Y. Li H. Shang L. Wang K. Li K. Zhou X. Dong X. Qu Z. Lu S. Hu X. Ruan S. Luo S. Wu J. Peng L. Cheng F. Pan L. Zou J. Jia C. Wang J. Liu X. Wang S. Wu X. Ge Q. He J. Zhan H. Qiu F. Guo L. Huang C. Jaki T. Hayden F.G. Horby P.W. Zhang D. Wang C. A trial of lopinavir–ritonavir in adults hospitalized with severe covid-19 N. Engl. J. Med. 382 19 2020 1787 1799 10.1056/NEJMoa2001282 32187464
Carlin A.F. Clark A.E. Chaillon A. Garretson A.F. Bray W. Porrachia M. Santos A.T. Rana T.M. Smith D.M. Virologic and immunologic characterization of COVID-19 recrudescence after nirmatrelvir/ritonavir treatment Clin. Infect. Dis. 2022 10.1093/cid/ciac496
Cervino L. Hynicka L.M. Direct-acting antivirals to prevent vertical transmission of viral hepatitis C: when is the optimal time to treat? Ann. Pharmacother. 52 11 2018 1152 1157 10.1177/1060028018772181 29681166
Champenois K. Baras A. Choisy P. Ajana F. Melliez H. Bocket L. Yazdanpanah Y. Lopinavir/ritonavir resistance in patients infected with HIV-1: two divergent resistance pathways? J. Med. Virol. 83 10 2011 1677 1681 10.1002/jmv.22161 21755502
Chang M.W. Torbett B.E. Accessory mutations maintain stability in drug-resistant HIV-1 protease J. Mol. Biol. 410 4 2011 756 760 10.1016/j.jmb.2011.03.038 21762813
Chen C.-H. Chen C.-H. Lin C.-L. Lin C.-Y. Hu T.-H. Tung S.-Y. Hsieh S.-Y. Lu S.-N. Chien R.-N. Hung C.-H. Sheen I.S. Real-world safety and efficacy of paritaprevir/ritonavir/ombitasvir plus dasabuvir ± ribavirin in patients with hepatitis C virus genotype 1 and advanced hepatic fibrosis or compensated cirrhosis: a multicenter pooled analysis Sci. Rep. 9 1 2019 7086 10.1038/s41598-019-43554-3 31068655
Chia C.S.B. Xu W. Shuyi Ng P. A patent review on SARS coronavirus main protease (3CL(pro)) inhibitors ChemMedChem 17 1 2022 e202100576 10.1002/cmdc.202100576 e202100576
Chow W.A. Guo S. Valdes-Albini F. Nelfinavir induces liposarcoma apoptosis and cell cycle arrest by upregulating sterol regulatory element binding protein-1 Anti Cancer Drugs 17 8 2006 891 903 10.1097/01.cad.0000224448.08706.76 16940799
Chuck C.P. Chen C. Ke Z. Wan D.C. Chow H.F. Wong K.B. Design, synthesis and crystallographic analysis of nitrile-based broad-spectrum peptidomimetic inhibitors for coronavirus 3C-like proteases Eur. J. Med. Chem. 59 2013 1 6 10.1016/j.ejmech.2012.10.053 23202846
Chun T.W. Finzi D. Margolick J. Chadwick K. Schwartz D. Siliciano R.F. In vivo fate of HIV-1-infected T cells: quantitative analysis of the transition to stable latency Nat. Med. 1 12 1995 1284 1290 10.1038/nm1295-1284 7489410
Churchill S.A. Protease inhibitors: implications for HIV research and treatment J. Int. Assoc. Phys. AIDS Care 2 1 1996 13 18
Clavel F. Guétard D. Brun-Vézinet F. Chamaret S. Rey M.A. Santos-Ferreira M.O. Laurent A.G. Dauguet C. Katlama C. Rouzioux C. Isolation of a new human retrovirus from West African patients with AIDS Science 233 4761 1986 343 346 10.1126/science.2425430 2425430
Constant D.A. Mateo R. Nagamine C.M. Kirkegaard K. Targeting intramolecular proteinase NS2B/3 cleavages for trans-dominant inhibition of dengue virus Proc. Natl. Acad. Sci. U. S. A. 115 40 2018 10136 10141 10.1073/pnas.1805195115 30228122
Corman V.M. Muth D. Niemeyer D. Drosten C. Hosts and sources of endemic human coronaviruses Adv. Virus Res. 100 2018 163 188 10.1016/bs.aivir.2018.01.001 29551135
Cotter T.G. Jensen D.M. Glecaprevir/pibrentasvir for the treatment of chronic hepatitis C: design, development, and place in therapy Drug Des. Dev. Ther. 13 2019 2565 2577 10.2147/dddt.S172512
Coulson J.M. Adams A. Gray L.A. Evans A. COVID-19 "Rebound" associated with nirmatrelvir/ritonavir pre-hospital therapy J. Infect. 2022 10.1016/j.jinf.2022.06.011
Couzin-Frankel J. Antiviral pills could change pandemic's course Science 374 6569 2021 799 800 10.1126/science.acx9605 34762459
Craig J.C. Duncan I.B. Hockley D. Grief C. Roberts N.A. Mills J.S. Antiviral properties of Ro 31-8959, an inhibitor of human immunodeficiency virus (HIV) proteinase Antivir. Res. 16 4 1991 295 305 10.1016/0166-3542(91)90045-s 1810306
Cuevas J.M. González-Candelas F. Moya A. Sanjuán R. Effect of ribavirin on the mutation rate and spectrum of hepatitis C virus in vivo J. Virol. 83 11 2009 5760 5764 10.1128/jvi.00201-09 19321623
Cully M. A tale of two antiviral targets - and the COVID-19 drugs that bind them Nat. Rev. Drug Discov. 21 1 2022 3 5 10.1038/d41573-021-00202-8 34857884
da Costa V.G. Moreli M.L. Saivish M.V. The emergence of SARS, MERS and novel SARS-2 coronaviruses in the 21st century Arch. Virol. 165 7 2020 1517 1526 10.1007/s00705-020-04628-0 32322993
da Silva Rocha S.F.L. Olanda C.G. Fokoue H.H. Sant'Anna C.M.R. Virtual screening techniques in drug discovery: review and recent applications Curr. Top. Med. Chem. 19 19 2019 1751 1767 10.2174/1568026619666190816101948 31418662
Dash P.K. Kaminski R. Bella R. Su H. Mathews S. Ahooyi T.M. Chen C. Mancuso P. Sariyer R. Ferrante P. Donadoni M. Robinson J.A. Sillman B. Lin Z. Hilaire J.R. Banoub M. Elango M. Gautam N. Mosley R.L. Poluektova L.Y. McMillan J. Bade A.N. Gorantla S. Sariyer I.K. Burdo T.H. Young W.B. Amini S. Gordon J. Jacobson J.M. Edagwa B. Khalili K. Gendelman H.E. Sequential LASER ART and CRISPR treatments eliminate HIV-1 in a subset of infected humanized mice Nat. Commun. 10 1 2019 2753 10.1038/s41467-019-10366-y 31266936
Davis D.A. Soule E.E. Davidoff K.S. Daniels S.I. Naiman N.E. Yarchoan R. Activity of human immunodeficiency virus type 1 protease inhibitors against the initial autocleavage in Gag-Pol polyprotein processing Antimicrob. Agents Chemother. 56 7 2012 3620 3628 10.1128/AAC.00055-12 22508308
De Clercq E. Perspectives for the chemotherapy of AIDS Anticancer Res. 7 5b 1987 1023 1038 3324934
De Clercq E. Antiviral drugs in current clinical use J. Clin. Virol. 30 2 2004 115 133 10.1016/j.jcv.2004.02.009 15125867
De Clercq E. Li G. Approved antiviral drugs over the past 50 years Clin. Microbiol. Rev. 29 3 2016 695 747 10.1128/CMR.00102-15 27281742
de Mendoza C. Valer L. Bacheler L. Pattery T. Corral A. Soriano V. Prevalence of the HIV-1 protease mutation I47A in clinical practice and association with lopinavir resistance AIDS 20 7 2006 1071 1074 10.1097/01.aids.0000222084.44411.cc 16603864
De Vlaminck I. Khush K.K. Strehl C. Kohli B. Luikart H. Neff N.F. Okamoto J. Snyder T.M. Cornfield D.N. Nicolls M.R. Weill D. Bernstein D. Valantine H.A. Quake S.R. Temporal response of the human virome to immunosuppression and antiviral therapy Cell 155 5 2013 1178 1187 10.1016/j.cell.2013.10.034 24267896
de Wit E. Feldmann F. Cronin J. Jordan R. Okumura A. Thomas T. Scott D. Cihlar T. Feldmann H. Prophylactic and therapeutic remdesivir (GS-5734) treatment in the rhesus macaque model of MERS-CoV infection Proc. Natl. Acad. Sci. U. S. A. 117 12 2020 6771 6776 10.1073/pnas.1922083117 32054787
Deng W. Baki L. Yin J. Zhou H. Baumgarten C.M. HIV protease inhibitors elicit volume-sensitive Cl- current in cardiac myocytes via mitochondrial ROS J. Mol. Cell. Cardiol. 49 5 2010 746 752 10.1016/j.yjmcc.2010.08.013 20736017
Denison M.R. Graham R.L. Donaldson E.F. Eckerle L.D. Baric R.S. Coronaviruses: an RNA proofreading machine regulates replication fidelity and diversity RNA Biol. 8 2 2011 270 279 10.4161/rna.8.2.15013 21593585
Dienstag J.L. 46 - antiviral drugs against hepatitis viruses Bennett J.E. Dolin R. Blaser M.J. Mandell, Douglas, and Bennett's Principles and Practice of Infectious Diseases eighth ed. 2015 W.B. Saunders Philadelphia 563 575 e563
Doyon L. Tremblay S. Bourgon L. Wardrop E. Cordingley M.G. Selection and characterization of HIV-1 showing reduced susceptibility to the non-peptidic protease inhibitor tipranavir Antivir. Res. 68 1 2005 27 35 10.1016/j.antiviral.2005.07.003 16122817
Dragovich P.S. Prins T.J. Zhou R. Webber S.E. Marakovits J.T. Fuhrman S.A. Patick A.K. Matthews D.A. Lee C.A. Ford C.E. Burke B.J. Rejto P.A. Hendrickson T.F. Tuntland T. Brown E.L. Meador J.W. 3rd Ferre R.A. Harr J.E. Kosa M.B. Worland S.T. Structure-based design, synthesis, and biological evaluation of irreversible human rhinovirus 3C protease inhibitors. 4. Incorporation of P1 lactam moieties as L-glutamine replacements J. Med. Chem. 42 7 1999 1213 1224 10.1021/jm9805384 10197965
Dunning J. Thwaites R.S. Openshaw P.J.M. Seasonal and pandemic influenza: 100 years of progress, still much to learn Mucosal Immunol. 13 4 2020 566 573 10.1038/s41385-020-0287-5 32317736
Dustin L.B. Bartolini B. Capobianchi M.R. Pistello M. Hepatitis C virus: life cycle in cells, infection and host response, and analysis of molecular markers influencing the outcome of infection and response to therapy Clin. Microbiol. Infect. 22 10 2016 826 832 10.1016/j.cmi.2016.08.025 27592089
El-Baba T.J. Lutomski C.A. Kantsadi A.L. Malla T.R. John T. Mikhailov V. Bolla J.R. Schofield C.J. Zitzmann N. Vakonakis I. Robinson C.V. Allosteric inhibition of the SARS-CoV-2 main protease: insights from mass spectrometry based assays Angew Chem. Int. Ed. Engl. 59 52 2020 23544 23548 10.1002/anie.202010316 32841477
Elion G.B. Furman P.A. Fyfe J.A. de Miranda P. Beauchamp L. Schaeffer H.J. Selectivity of action of an antiherpetic agent, 9-(2-hydroxyethoxymethyl) guanine Proc. Natl. Acad. Sci. U. S. A. 74 12 1977 5716 5720 10.1073/pnas.74.12.5716 202961
Elion R. Cohen C. Gathe J. Shalit P. Hawkins T. Liu H.C. Mathias A.A. Chuck S.L. Kearney B.P. Warren D.R. Phase 2 study of cobicistat versus ritonavir each with once-daily atazanavir and fixed-dose emtricitabine/tenofovir df in the initial treatment of HIV infection Aids 25 15 2011 1881 1886 10.1097/QAD.0b013e32834b4d48 21811136
Ely E.W. Ramanan A.V. Kartman C.E. de Bono S. Liao R. Piruzeli M.L.B. Goldman J.D. Saraiva J.F.K. Chakladar S. Marconi V.C. Efficacy and safety of baricitinib plus standard of care for the treatment of critically ill hospitalised adults with COVID-19 on invasive mechanical ventilation or extracorporeal membrane oxygenation: an exploratory, randomised, placebo-controlled trial Lancet Respir. Med. 2022 10.1016/s2213-2600(22)00006-6
Erbel P. Schiering N. D'Arcy A. Renatus M. Kroemer M. Lim S.P. Yin Z. Keller T.H. Vasudevan S.G. Hommel U. Structural basis for the activation of flaviviral NS3 proteases from dengue and West Nile virus Nat. Struct. Mol. Biol. 13 4 2006 372 373 10.1038/nsmb1073 16532006
Evenseth L.S.M. Gabrielsen M. Sylte I. The GABA(B) receptor-structure, ligand binding and drug development Molecules 25 13 2020 10.3390/molecules25133093
Excision https://www.excision.bio/technology
Fassmannová D. Sedlák F. Sedláček J. Špička I. Grantz Šašková K. Nelfinavir inhibits the TCF11/nrf1-mediated proteasome recovery pathway in multiple myeloma Cancers 12 5 2020 10.3390/cancers12051065
Foote B.S. Spooner L.M. Belliveau P.P. Boceprevir: a protease inhibitor for the treatment of chronic hepatitis C Ann. Pharmacother. 45 9 2011 1085 1093 10.1345/aph.1P744 21828346
Fu L. Ye F. Feng Y. Yu F. Wang Q. Wu Y. Zhao C. Sun H. Huang B. Niu P. Song H. Shi Y. Li X. Tan W. Qi J. Gao G.F. Both Boceprevir and GC376 efficaciously inhibit SARS-CoV-2 by targeting its main protease Nat. Commun. 11 1 2020 4417 10.1038/s41467-020-18233-x 32887884
Furfine E.S. Baker C.T. Hale M.R. Reynolds D.J. Salisbury J.A. Searle A.D. Studenberg S.D. Todd D. Tung R.D. Spaltenstein A. Preclinical pharmacology and pharmacokinetics of GW433908, a water-soluble prodrug of the human immunodeficiency virus protease inhibitor amprenavir Antimicrob. Agents Chemother. 48 3 2004 791 798 10.1128/aac.48.3.791-798.2004 14982766
Furman P.A. de Miranda P. St Clair M.H. Elion G.B. Metabolism of acyclovir in virus-infected and uninfected cells Antimicrob. Agents Chemother. 20 4 1981 518 524 10.1128/aac.20.4.518 6282196
Gable J.E. Lee G.M. Acker T.M. Hulce K.R. Gonzalez E.R. Schweigler P. Melkko S. Farady C.J. Craik C.S. Fragment-based protein-protein interaction antagonists of a viral dimeric protease ChemMedChem 11 8 2016 862 869 10.1002/cmdc.201500526 26822284
Gallo R.C. Poiesz B.J. Ruscetti F.W. Regulation of human T-cell proliferation: T-cell growth factor and isolation of a new class of type-C retroviruses from human T-cells Haematol. Blood Transfus. 26 1981 502 514 10.1007/978-3-642-67984-1_93 6274766
Gallo R.C. Sarin P.S. Gelmann E.P. Robert-Guroff M. Richardson E. Kalyanaraman V.S. Mann D. Sidhu G.D. Stahl R.E. Zolla-Pazner S. Leibowitch J. Popovic M. Isolation of human T-cell leukemia virus in acquired immune deficiency syndrome (AIDS) Science 220 4599 1983 865 867 10.1126/science.6601823 6601823
Gan J.H. Liu J.X. Liu Y. Chen S.W. Dai W.T. Xiao Z.X. Cao Y. DrugRep: an automatic virtual screening server for drug repurposing Acta Pharmacol. Sin. 1–9 2022 10.1038/s41401-022-00996-2
Gane E. Poordad F. Wang S. Asatryan A. Kwo P.Y. Lalezari J. Wyles D.L. Hassanein T. Aguilar H. Maliakkal B. Liu R. Lin C.W. Ng T.I. Kort J. Mensa F.J. High efficacy of ABT-493 and ABT-530 treatment in patients with HCV genotype 1 or 3 infection and compensated cirrhosis Gastroenterology 151 4 2016 651 659 10.1053/j.gastro.2016.07.020 e651 27456384
Gane E.J. Kowdley K.V. Pound D. Stedman C.A. Davis M. Etzkorn K. Gordon S.C. Bernstein D. Everson G. Rodriguez-Torres M. Tsai N. Khalid O. Yang J.C. Lu S. Dvory-Sobol H. Stamm L.M. Brainard D.M. McHutchison J.G. Tong M. Chung R.T. Beavers K. Poulos J.E. Kwo P.Y. Nguyen M.H. Efficacy of sofosbuvir, velpatasvir, and GS-9857 in patients with hepatitis C virus genotype 2, 3, 4, or 6 infections in an open-label, phase 2 trial Gastroenterology 151 5 2016 902 909 10.1053/j.gastro.2016.07.038 27486033
Gane E.J. Schwabe C. Hyland R.H. Yang Y. Svarovskaia E. Stamm L.M. Brainard D.M. McHutchison J.G. Stedman C.A. Efficacy of the combination of sofosbuvir, velpatasvir, and the NS3/4A protease inhibitor GS-9857 in treatment-naïve or previously treated patients with hepatitis C virus genotype 1 or 3 infections Gastroenterology 151 3 2016 448 456 10.1053/j.gastro.2016.05.021 e441 27240903
Garcia-Cehic D. Rando A. Rodriguez-Frias F. Gregori J. Costa J.G. Carrión J.A. Macenlle R. Pamplona J. Castro-Iglesias A. Cañizares A. Tabernero D. Campos C. Buti M. Esteban J.I. Quer J. Resistance-associated substitutions after sofosbuvir/velpatasvir/voxilaprevir triple therapy failure J. Viral Hepat. 28 9 2021 1319 1324 10.1111/jvh.13497 33720484
Gehlhaar D.K. Verkhivker G.M. Rejto P.A. Sherman C.J. Fogel D.B. Fogel L.J. Freer S.T. Molecular recognition of the inhibitor AG-1343 by HIV-1 protease: conformationally flexible docking by evolutionary programming Chem. Biol. 2 5 1995 317 324 10.1016/1074-5521(95)90050-0 9383433
Geller R. Estada Ú. Peris J.B. Andreu I. Bou J.V. Garijo R. Cuevas J.M. Sabariegos R. Mas A. Sanjuán R. Highly heterogeneous mutation rates in the hepatitis C virus genome Nat. Microbiol. 1 7 2016 16045 10.1038/nmicrobiol.2016.45
Georgi F. Andriasyan V. Witte R. Murer L. Hemmi S. Yu L. Grove M. Meili N. Kuttler F. Yakimovich A. Turcatti G. Greber U.F. The FDA-approved drug nelfinavir inhibits lytic cell-free but not cell-associated nonlytic transmission of human adenovirus Antimicrob. Agents Chemother. 64 9 2020 10.1128/AAC.01002-20 e01002-01020
Ghosh A.K. Dawson Z.L. Mitsuya H. Darunavir, a conceptually new HIV-1 protease inhibitor for the treatment of drug-resistant HIV Bioorg. Med. Chem. 15 24 2007 7576 7580 10.1016/j.bmc.2007.09.010 17900913
Ghosh A.K. Chapsal B.D. Weber I.T. Mitsuya H. Design of HIV protease inhibitors targeting protein backbone: an effective strategy for combating drug resistance Acc. Chem. Res. 41 1 2008 78 86 10.1021/ar7001232 17722874
Ghosh A.K. Osswald H.L. Prato G. Recent progress in the development of HIV-1 protease inhibitors for the treatment of HIV/AIDS J. Med. Chem. 59 11 2016 5172 5208 10.1021/acs.jmedchem.5b01697 26799988
Gillis E.P. Eastman K.J. Hill M.D. Donnelly D.J. Meanwell N.A. Applications of fluorine in medicinal chemistry J. Med. Chem. 58 21 2015 8315 8359 10.1021/acs.jmedchem.5b00258 26200936
Gills J.J. Lopiccolo J. Dennis P.A. Nelfinavir, a new anti-cancer drug with pleiotropic effects and many paths to autophagy Autophagy 4 1 2008 107 109 10.4161/auto.5224 18000394
Goldsmith D.R. Perry C.M. Atazanavir. Drugs 63 16 2003 1679 1693 10.2165/00003495-200363160-00003 discussion 1694-1675 12904086
Gordon S.C. Muir A.J. Lim J.K. Pearlman B. Argo C.K. Ramani A. Maliakkal B. Alam I. Stewart T.G. Vainorius M. Peter J. Nelson D.R. Fried M.W. Reddy K.R. Safety profile of boceprevir and telaprevir in chronic hepatitis C: real world experience from HCV-TARGET J. Hepatol. 62 2 2015 286 293 10.1016/j.jhep.2014.08.052 25218788
Graff D.E. Shakhnovich E.I. Coley C.W. Accelerating high-throughput virtual screening through molecular pool-based active learning Chem. Sci. 12 22 2021 7866 7881 10.1039/d0sc06805e 34168840
Graham S.V. The human papillomavirus replication cycle, and its links to cancer progression: a comprehensive review Clin. Sci. (Lond.) 131 17 2017 2201 2221 10.1042/cs20160786 28798073
Grandi N. Tramontano E. Human endogenous retroviruses are ancient acquired elements still shaping innate immune responses Front. Immunol. 9 2039 2018 10.3389/fimmu.2018.02039
Greasley S.E. Noell S. Plotnikova O. Ferre R. Liu W. Bolanos B. Fennell K. Nicki J. Craig T. Zhu Y. Stewart A.E. Steppan C.M. Structural basis for the in vitro efficacy of nirmatrelvir against SARS-CoV-2 variants J. Biol. Chem. 298 6 2022 101972 10.1016/j.jbc.2022.101972
Gu Y. Wang X. Wang Y. Wang Y. Li J. Yu F.X. Nelfinavir inhibits human DDI2 and potentiates cytotoxicity of proteasome inhibitors Cell. Signal. 75 2020 109775 10.1016/j.cellsig.2020.109775
Hahm B. Han D.S. Back S.H. Song O.K. Cho M.J. Kim C.J. Shimotohno K. Jang S.K. NS3-4A of hepatitis C virus is a chymotrypsin-like protease J. Virol. 69 4 1995 2534 2539 10.1128/jvi.69.4.2534-2539.1995 7884903
Hakkola J. Hukkanen J. Turpeinen M. Pelkonen O. Inhibition and induction of CYP enzymes in humans: an update Arch. Toxicol. 94 11 2020 3671 3722 10.1007/s00204-020-02936-7 33111191
Hammond J. Leister-Tebbe H. Gardner A. Abreu P. Bao W. Wisemandle W. Baniecki M. Hendrick V.M. Damle B. Simón-Campos A. Pypstra R. Rusnak J.M. Oral nirmatrelvir for high-risk, nonhospitalized adults with covid-19 N. Engl. J. Med. 386 15 2022 1397 1408 10.1056/NEJMoa2118542 35172054
Han D.S. Hahm B. Rho H.M. Jang S.K. Identification of the protease domain in NS3 of hepatitis C virus J. Gen. Virol. 76 Pt 4 1995 985 993 10.1099/0022-1317-76-4-985 9049347
Han B. Martin R. Xu S. Parvangada A. Svarovskaia E.S. Mo H. Dvory-Sobol H. Sofosbuvir susceptibility of genotype 1 to 6 HCV from DAA-naïve subjects Antivir. Res. 170 2019 104574 10.1016/j.antiviral.2019.104574
Han S.H. Goins C.M. Arya T. Shin W.J. Maw J. Hooper A. Sonawane D.P. Porter M.R. Bannister B.E. Crouch R.D. Lindsey A.A. Lakatos G. Martinez S.R. Alvarado J. Akers W.S. Wang N.S. Jung J.U. Macdonald J.D. Stauffer S.R. Structure-based optimization of ml300-derived, noncovalent inhibitors targeting the severe acute respiratory syndrome coronavirus 3CL protease (SARS-CoV-2 3CL(pro)) J. Med. Chem. 65 4 2022 2880 2904 10.1021/acs.jmedchem.1c00598 34347470
Harper S. McCauley J.A. Rudd M.T. Ferrara M. DiFilippo M. Crescenzi B. Koch U. Petrocchi A. Holloway M.K. Butcher J.W. Romano J.J. Bush K.J. Gilbert K.F. McIntyre C.J. Nguyen K.T. Nizi E. Carroll S.S. Ludmerer S.W. Burlein C. DiMuzio J.M. Graham D.J. McHale C.M. Stahlhut M.W. Olsen D.B. Monteagudo E. Cianetti S. Giuliano C. Pucci V. Trainor N. Fandozzi C.M. Rowley M. Coleman P.J. Vacca J.P. Summa V. Liverton N.J. Discovery of MK-5172, a macrocyclic hepatitis C virus NS3/4a protease inhibitor ACS Med. Chem. Lett. 3 4 2012 332 336 10.1021/ml300017p 24900473
Hayashi H. Takamune N. Nirasawa T. Aoki M. Morishita Y. Das D. Koh Y. Ghosh A.K. Misumi S. Mitsuya H. Dimerization of HIV-1 protease occurs through two steps relating to the mechanism of protease dimerization inhibition by darunavir Proc. Natl. Acad. Sci. USA 111 33 2014 12234 12239 10.1073/pnas.1400027111 25092296
Heo Y.A. Deeks E.D. Sofosbuvir/velpatasvir/voxilaprevir: a review in chronic hepatitis C Drugs 78 5 2018 577 587 10.1007/s40265-018-0895-5 29546556
Hermans P.E. Cockerill F.R. 3rd Antiviral agents Mayo Clin. Proc. 58 4 1983 217 222 6339831
Hermida-Matsumoto L. Resh M.D. Human immunodeficiency virus type 1 protease triggers a myristoyl switch that modulates membrane binding of Pr55(gag) and p17MA J. Virol. 73 3 1999 1902 1908 10.1128/jvi.73.3.1902-1908.1999 9971769
Heskin J. Pallett S.J.C. Mughal N. Davies G.W. Moore L.S.P. Rayment M. Jones R. Caution required with use of ritonavir-boosted PF-07321332 in COVID-19 management Lancet 399 10319 2022 21 22 10.1016/S0140-6736(21)02657-X 34973713
Hézode C. Treatment of hepatitis C: results in real life Liver Int. 38 Suppl. 1 2018 21 27 10.1111/liv.13638 29427481
Hilgenfeld R. Lei J. Zhang L. The structure of the Zika virus protease, NS2B/NS3(pro) Adv. Exp. Med. Biol. 1062 2018 131 145 10.1007/978-981-10-8727-1_10 29845530
Ho J.S.Y. Zhu Z. Marazzi I. Unconventional viral gene expression mechanisms as therapeutic targets Nature 593 7859 2021 362 371 10.1038/s41586-021-03511-5 34012080
Hoffman R.L. Kania R.S. Brothers M.A. Davies J.F. Ferre R.A. Gajiwala K.S. He M. Hogan R.J. Kozminski K. Li L.Y. Lockner J.W. Lou J. Marra M.T. Mitchell L.J. Jr. Murray B.W. Nieman J.A. Noell S. Planken S.P. Rowe T. Ryan K. Smith G.J. 3rd Solowiej J.E. Steppan C.M. Taggart B. Discovery of ketone-based covalent inhibitors of coronavirus 3CL proteases for the potential therapeutic treatment of COVID-19 J. Med. Chem. 63 21 2020 12725 12747 10.1021/acs.jmedchem.0c01063 33054210
Howe A.Y.M. Venkatraman S. The discovery and development of boceprevir: a novel, first-generation inhibitor of the hepatitis C virus NS3/4A serine protease J. Clin. Transl. Hepatol. 1 1 2013 22 32 10.14218/JCTH.2013.002XX 26357603
Huang L. Li L. Tien C. LaBarbera D.V. Chen C. Targeting HIV-1 protease autoprocessing for high-throughput drug discovery and drug resistance assessment Sci. Rep. 9 1 2019 301 10.1038/s41598-018-36730-4 30670786
Hui D.S. E I.A. Madani T.A. Ntoumi F. Kock R. Dar O. Ippolito G. McHugh T.D. Memish Z.A. Drosten C. Zumla A. Petersen E. The continuing 2019-nCoV epidemic threat of novel coronaviruses to global health - the latest 2019 novel coronavirus outbreak in Wuhan, China Int. J. Infect. Dis. 91 2020 264 266 10.1016/j.ijid.2020.01.009 31953166
Humpolíčková J. Weber J. Starková J. Mašínová E. Günterová J. Flaisigová I. Konvalinka J. Majerová T. Inhibition of the precursor and mature forms of HIV-1 protease as a tool for drug evaluation Sci. Rep. 8 1 2018 10438 10.1038/s41598-018-28638-w
Hwang Y.-C. Lu R.-M. Su S.-C. Chiang P.-Y. Ko S.-H. Ke F.-Y. Liang K.-H. Hsieh T.-Y. Wu H.-C. Monoclonal antibodies for COVID-19 therapy and SARS-CoV-2 detection J. Biomed. Sci. 29 1 2022 1 10.1186/s12929-021-00784-w 34983527
Iketani S. Mohri H. Culbertson B. Hong S.J. Duan Y. Luck M.I. Annavajhala M.K. Guo Y. Sheng Z. Uhlemann A.C. Goff S.P. Sabo Y. Yang H. Chavez A. Ho D.D. Multiple pathways for SARS-CoV-2 resistance to nirmatrelvir Nature 2022 10.1038/s41586-022-05514-2
Isken O. Walther T. Wong-Dilworth L. Rehders D. Redecke L. Tautz N. Identification of NS2 determinants stimulating intrinsic HCV NS2 protease activity PLoS Pathog. 18 6 2022 e1010644 10.1371/journal.ppat.1010644
Jacks T. Power M.D. Masiarz F.R. Luciw P.A. Barr P.J. Varmus H.E. Characterization of ribosomal frameshifting in HIV-1 gag-pol expression Nature 331 6153 1988 280 283 10.1038/331280a0 2447506
Jacobs J. Grum-Tokars V. Zhou Y. Turlington M. Saldanha S.A. Chase P. Eggler A. Dawson E.S. Baez-Santos Y.M. Tomar S. Mielech A.M. Baker S.C. Lindsley C.W. Hodder P. Mesecar A. Stauffer S.R. Discovery, synthesis, and structure-based optimization of a series of N-(tert-butyl)-2-(N-arylamido)-2-(pyridin-3-yl) acetamides (ML188) as potent noncovalent small molecule inhibitors of the severe acute respiratory syndrome coronavirus (SARS-CoV) 3CL protease J. Med. Chem. 56 2 2013 534 546 10.1021/jm301580n 23231439
Jacobsen H. Yasargil K. Winslow D.L. Craig J.C. Kröhn A. Duncan I.B. Mous J. Characterization of human immunodeficiency virus type 1 mutantswith decreased sensitivity to proteinase inhibitor Ro 31-8959 Virology 206 1 1995 527 534 10.1016/S0042-6822(95)80069-7 7831807
Jacobson I.M. Lawitz E. Gane E.J. Willems B.E. Ruane P.J. Nahass R.G. Borgia S.M. Shafran S.D. Workowski K.A. Pearlman B. Hyland R.H. Stamm L.M. Svarovskaia E. Dvory-Sobol H. Zhu Y. Subramanian G.M. Brainard D.M. McHutchison J.G. Bräu N. Berg T. Agarwal K. Bhandari B.R. Davis M. Feld J.J. Dore G.J. Stedman C.A.M. Thompson A.J. Asselah T. Roberts S.K. Foster G.R. Efficacy of 8 Weeks of sofosbuvir, velpatasvir, and voxilaprevir in patients with chronic HCV infection: 2 phase 3 randomized trials Gastroenterology 153 1 2017 113 122 10.1053/j.gastro.2017.03.047 28390869
Jacobson I.M. Lawitz E. Kwo P.Y. Hézode C. Peng C.Y. Howe A.Y.M. Hwang P. Wahl J. Robertson M. Barr E. Haber B.A. Safety and efficacy of elbasvir/grazoprevir in patients with hepatitis C virus infection and compensated cirrhosis: an integrated analysis Gastroenterology 152 6 2017 1372 1382 10.1053/j.gastro.2017.01.050 e1372 28193518
James J.S. Saquinavir (Invirase): first protease inhibitor approved--reimbursement, information hotline numbers AIDS Treat. News 237 1995 1 2
Jeong J.H. Chokkakula S. Min S.C. Kim B.K. Choi W.S. Oh S. Yun Y.S. Kang D.H. Lee O.J. Kim E.G. Choi J.H. Lee J.Y. Choi Y.K. Baek Y.H. Song M.S. Combination therapy with nirmatrelvir and molnupiravir improves the survival of SARS-CoV-2 infected mice Antivir. Res. 208 2022 105430 10.1016/j.antiviral.2022.105430
Jiang M. Mani N. Lin C. Ardzinski A. Nelson M. Reagan D. Bartels D. Zhou Y. Nicolas O. Rao B.G. Müh U. Hanzelka B. Tigges A. Rijnbrand R. Kieffer T.L. In vitro phenotypic characterization of hepatitis C virus NS3 protease variants observed in clinical studies of telaprevir Antimicrob. Agents Chemother. 57 12 2013 6236 6245 10.1128/AAC.01578-13 24100495
Jin Z. Du X. Xu Y. Deng Y. Liu M. Zhao Y. Zhang B. Li X. Zhang L. Peng C. Duan Y. Yu J. Wang L. Yang K. Liu F. Jiang R. Yang X. You T. Liu X. Yang X. Bai F. Liu H. Liu X. Guddat L.W. Xu W. Xiao G. Qin C. Shi Z. Jiang H. Rao Z. Yang H. Structure of M(pro) from SARS-CoV-2 and discovery of its inhibitors Nature 582 7811 2020 289 293 10.1038/s41586-020-2223-y 32272481
Jochmans D. Anders M. Keuleers I. Smeulders L. Kräusslich H.-G. Kraus G. Müller B. Selective killing of human immunodeficiency virus infected cells by non-nucleoside reverse transcriptase inhibitor-induced activation of HIV protease Retrovirology 7 1 2010 89 10.1186/1742-4690-7-89 20950436
Jochmans D. Liu C. Donckers K. Stoycheva A. Boland S. Stevens S.K. De Vita C. Vanmechelen B. Maes P. Trüeb B. Ebert N. Thiel V. De Jonghe S. Vangeel L. Bardiot D. Jekle A. Blatt L.M. Beigelman L. Symons J.A. Raboisson P. Chaltin P. Marchand A. Neyts J. Deval J. Vandyck K. The substitutions L50F, E166A and L167F in SARS-CoV-2 3CLpro are selected by a protease inhibitor in vitro and confer resistance to nirmatrelvir bioRxiv 2022 10.1101/2022.06.07.495116 2022.2006.2007.495116
Kabinger F. Stiller C. Schmitzová J. Dienemann C. Kokic G. Hillen H.S. Höbartner C. Cramer P. Mechanism of molnupiravir-induced SARS-CoV-2 mutagenesis Nat. Struct. Mol. Biol. 28 9 2021 740 746 10.1038/s41594-021-00651-0 34381216
Kaldor S.W. Kalish V.J. Davies J.F. Shetty B.V. Fritz J.E. Appelt K. Burgess J.A. Campanale K.M. Chirgadze N.Y. Clawson D.K. Dressman B.A. Hatch S.D. Khalil D.A. Kosa M.B. Lubbehusen P.P. Muesing M.A. Patick A.K. Reich S.H. Su K.S. Tatlock J.H. Viracept (nelfinavir mesylate, AG1343): a potent, orally bioavailable inhibitor of HIV-1 protease J. Med. Chem. 40 24 1997 3979 3985 10.1021/jm9704098 9397180
Kalu N.N. Desai P.J. Shirley C.M. Gibson W. Dennis P.A. Ambinder R.F. Nelfinavir inhibits maturation and export of herpes simplex virus 1 J. Virol. 88 10 2014 5455 5461 10.1128/JVI.03790-13 24574416
Kang C. Keller T.H. Luo D. Zika virus protease: an antiviral drug target Trends Microbiol. 25 10 2017 797 808 10.1016/j.tim.2017.07.001 28789826
Kaplan A.H. Swanstrom R. Human immunodeficiency virus type 1 Gag proteins are processed in two cellular compartments Proc. Natl. Acad. Sci. USA 88 10 1991 4528 4532 10.1073/pnas.88.10.4528 2034693
Kaptein S.J.F. Goethals O. Kiemel D. Marchand A. Kesteleyn B. Bonfanti J.F. Bardiot D. Stoops B. Jonckers T.H.M. Dallmeier K. Geluykens P. Thys K. Crabbe M. Chatel-Chaix L. Münster M. Querat G. Touret F. de Lamballerie X. Raboisson P. Simmen K. Chaltin P. Bartenschlager R. Van Loock M. Neyts J. A pan-serotype dengue virus inhibitor targeting the NS3-NS4B interaction Nature 598 7881 2021 504 509 10.1038/s41586-021-03990-6 34616043
Kataev V.E. Garifullin B.F. Antiviral nucleoside analogs Chem. Heterocycl. Compd. 57 4 2021 326 341 10.1007/s10593-021-02912-8
Katsumoto T. Hattori N. Kurimura T. Maturation of human immunodeficiency virus, strain LAV, in vitro Intervirology 27 3 1987 148 153 10.1159/000149733 3667187
Kausar S. Said Khan F. Ishaq Mujeeb Ur Rehman M. Akram M. Riaz M. Rasool G. Hamid Khan A. Saleem I. Shamim S. Malik A. A review: mechanism of action of antiviral drugs Int. J. Immunopathol. Pharmacol. 35 2021 10.1177/20587384211002621
Kawabata S. Gills J.J. Mercado-Matos J.R. Lopiccolo J. Wilson W. 3rd Hollander M.C. Dennis P.A. Synergistic effects of nelfinavir and bortezomib on proteotoxic death of NSCLC and multiple myeloma cells Cell Death Dis. 3 7 2012 e353 10.1038/cddis.2012.87 22825471
Keating G.M. Elbasvir/grazoprevir: first global approval Drugs 76 5 2016 617 624 10.1007/s40265-016-0558-3 26943930
Kempf D.J. Marsh K.C. Denissen J.F. McDonald E. Vasavanonda S. Flentge C.A. Green B.E. Fino L. Park C.H. Kong X.P. ABT-538 is a potent inhibitor of human immunodeficiency virus protease and has high oral bioavailability in humans Proc. Natl. Acad. Sci. U. S. A. 92 7 1995 2484 2488 10.1073/pnas.92.7.2484 7708670
Kempf D.J. Marsh K.C. Kumar G. Rodrigues A.D. Denissen J.F. McDonald E. Kukulka M.J. Hsu A. Granneman G.R. Baroldi P.A. Sun E. Pizzuti D. Plattner J.J. Norbeck D.W. Leonard J.M. Pharmacokinetic enhancement of inhibitors of the human immunodeficiency virus protease by coadministration with ritonavir Antimicrob. Agents Chemother. 41 3 1997 654 660 10.1128/aac.41.3.654 9056009
Kempf D.J. Isaacson J.D. King M.S. Brun S.C. Xu Y. Real K. Bernstein B.M. Japour A.J. Sun E. Rode R.A. Identification of genotypic changes in human immunodeficiency virus protease that correlate with reduced susceptibility to the protease inhibitor lopinavir among viral isolates from protease inhibitor-experienced patients J. Virol. 75 16 2001 7462 7469 10.1128/jvi.75.16.7462-7469.2001 11462018
Khayat R. Batra R. Qian C. Halmos T. Bailey M. Tong L. Structural and biochemical studies of inhibitor binding to human cytomegalovirus protease Biochemistry 42 4 2003 885 891 10.1021/bi027045s 12549906
Kim E. Baker C. Dwyer M. Murcko M. Rao B. Tung R. Navia M. Crystal structure of HIV-1 protease in complex with VX-478, a potent and orally bioavailable inhibitor of the enzyme J. Am. Chem. Soc. 117 3 1995 1181 1182
Kim J.L. Morgenstern K.A. Lin C. Fox T. Dwyer M.D. Landro J.A. Chambers S.P. Markland W. Lepre C.A. O'Malley E.T. Harbeson S.L. Rice C.M. Murcko M.A. Caron P.R. Thomson J.A. Crystal structure of the hepatitis C virus NS3 protease domain complexed with a synthetic NS4A cofactor peptide Cell 87 2 1996 343 355 10.1016/s0092-8674(00)81351-3 8861917
Klemm T. Ebert G. Calleja D.J. Allison C.C. Richardson L.W. Bernardini J.P. Lu B.G. Kuchel N.W. Grohmann C. Shibata Y. Gan Z.Y. Cooney J.P. Doerflinger M. Au A.E. Blackmore T.R. van der Heden van Noort G.J. Geurink P.P. Ovaa H. Newman J. Riboldi-Tunnicliffe A. Czabotar P.E. Mitchell J.P. Feltham R. Lechtenberg B.C. Lowes K.N. Dewson G. Pellegrini M. Lessene G. Komander D. Mechanism and inhibition of the papain-like protease, PLpro, of SARS-CoV-2 EMBO J. 39 18 2020 e106275 10.15252/embj.2020106275
Koh Y. Nakata H. Maeda K. Ogata H. Bilcer G. Devasamudram T. Kincaid J.F. Boross P. Wang Y.F. Tie Y. Volarath P. Gaddis L. Harrison R.W. Weber I.T. Ghosh A.K. Mitsuya H. Novel bis-tetrahydrofuranylurethane-containing nonpeptidic protease inhibitor (PI) UIC-94017 (TMC114) with potent activity against multi-PI-resistant human immunodeficiency virus in vitro Antimicrob. Agents Chemother. 47 10 2003 3123 3129 10.1128/aac.47.10.3123-3129.2003 14506019
Kohl N.E. Emini E.A. Schleif W.A. Davis L.J. Heimbach J.C. Dixon R.A. Scolnick E.M. Sigal I.S. Active human immunodeficiency virus protease is required for viral infectivity Proc. Natl. Acad. Sci. U. S. A. 85 13 1988 4686 4690 10.1073/pnas.85.13.4686 3290901
Konvalinka J. Litterst M.A. Welker R. Kottler H. Rippmann F. Heuser A.M. Kräusslich H.G. An active-site mutation in the human immunodeficiency virus type 1 proteinase (PR) causes reduced PR activity and loss of PR-mediated cytotoxicity without apparent effect on virus maturation and infectivity J. Virol. 69 11 1995 7180 7186 10.1128/jvi.69.11.7180-7186.1995 7474139
Konvalinka J. Kräusslich H.G. Müller B. Retroviral proteases and their roles in virion maturation Virology 479–480 2015 403 417 10.1016/j.virol.2015.03.021
Koster J.C. Remedi M.S. Qiu H. Nichols C.G. Hruz P.W. HIV protease inhibitors acutely impair glucose-stimulated insulin release Diabetes 52 7 2003 1695 1700 10.2337/diabetes.52.7.1695 12829635
Kožíšek M. Bray J. Řezáčová P. Šašková K. Brynda J. Pokorná J. Mammano F. Rulíšek L. Konvalinka J. Molecular analysis of the HIV-1 resistance development: enzymatic activities, crystal structures, and thermodynamics of nelfinavir-resistant HIV protease mutants J. Mol. Biol. 374 4 2007 1005 1016 10.1016/j.jmb.2007.09.083 17977555
Kozísek M. Henke S. Sasková K.G. Jacobs G.B. Schuch A. Buchholz B. Müller V. Kräusslich H.G. Rezácová P. Konvalinka J. Bodem J. Mutations in HIV-1 gag and pol compensate for the loss of viral fitness caused by a highly mutated protease Antimicrob. Agents Chemother. 56 8 2012 4320 4330 10.1128/aac.00465-12 22644035
Kožíšek M. Lepšík M. Grantz Šašková K. Brynda J. Konvalinka J. Rezáčová P. Thermodynamic and structural analysis of HIV protease resistance to darunavir - analysis of heavily mutated patient-derived HIV-1 proteases FEBS J. 281 7 2014 1834 1847 10.1111/febs.12743 24785545
Kräusslich H.G. Human immunodeficiency virus proteinase dimer as component of the viral polyprotein prevents particle assembly and viral infectivity Proc. Natl. Acad. Sci. USA 88 8 1991 3213 3217 10.1073/pnas.88.8.3213 2014242
Krishnan P. Tripathi R. Schnell G. Reisch T. Beyer J. Irvin M. Xie W. Larsen L. Cohen D. Podsadecki T. Pilot-Matias T. Collins C. Resistance analysis of baseline and treatment-emergent variants in hepatitis C virus genotype 1 in the AVIATOR study with paritaprevir-ritonavir, ombitasvir, and dasabuvir Antimicrob. Agents Chemother. 59 9 2015 5445 5454 10.1128/aac.00998-15 26100711
Kumar G.N. Rodrigues A.D. Buko A.M. Denissen J.F. Cytochrome P450-mediated metabolism of the HIV-1 protease inhibitor ritonavir (ABT-538) in human liver microsomes J. Pharmacol. Exp. Therapeut. 277 1 1996 423 431
Kuroki A. Tay J. Lee G.H. Yang Y.Y. Broad-spectrum antiviral peptides and polymers Adv. Healthc. Mater. 10 23 2021 e2101113 10.1002/adhm.202101113
Kwong A.D. Kim J.L. Rao G. Lipovsek D. Raybuck S.A. Hepatitis C virus NS3/4A protease Antivir. Res. 40 1–2 1998 1 18 10.1016/s0166-3542(98)00043-6 9864043
Kwong A.D. Kauffman R.S. Hurter P. Mueller P. Discovery and development of telaprevir: an NS3-4A protease inhibitor for treating genotype 1 chronic hepatitis C virus Nat. Biotechnol. 29 11 2011 993 1003 10.1038/nbt.2020 22068541
Lamb Y.N. Nirmatrelvir plus ritonavir: first approval Drugs 82 5 2022 585 591 10.1007/s40265-022-01692-5 35305258
Lander E.S. Linton L.M. Birren B. Nusbaum C. Zody M.C. Baldwin J. Devon K. Dewar K. Doyle M. FitzHugh W. Funke R. Gage D. Harris K. Heaford A. Howland J. Kann L. Lehoczky J. LeVine R. McEwan P. McKernan K. Meldrim J. Mesirov J.P. Miranda C. Morris W. Naylor J. Raymond C. Rosetti M. Santos R. Sheridan A. Sougnez C. Stange-Thomann Y. Stojanovic N. Subramanian A. Wyman D. Rogers J. Sulston J. Ainscough R. Beck S. Bentley D. Burton J. Clee C. Carter N. Coulson A. Deadman R. Deloukas P. Dunham A. Dunham I. Durbin R. French L. Grafham D. Gregory S. Hubbard T. Humphray S. Hunt A. Jones M. Lloyd C. McMurray A. Matthews L. Mercer S. Milne S. Mullikin J.C. Mungall A. Plumb R. Ross M. Shownkeen R. Sims S. Waterston R.H. Wilson R.K. Hillier L.W. McPherson J.D. Marra M.A. Mardis E.R. Fulton L.A. Chinwalla A.T. Pepin K.H. Gish W.R. Chissoe S.L. Wendl M.C. Delehaunty K.D. Miner T.L. Delehaunty A. Kramer J.B. Cook L.L. Fulton R.S. Johnson D.L. Minx P.J. Clifton S.W. Hawkins T. Branscomb E. Predki P. Richardson P. Wenning S. Slezak T. Doggett N. Cheng J.F. Olsen A. Lucas S. Elkin C. Uberbacher E. Frazier M. Gibbs R.A. Muzny D.M. Scherer S.E. Bouck J.B. Sodergren E.J. Worley K.C. Rives C.M. Gorrell J.H. Metzker M.L. Naylor S.L. Kucherlapati R.S. Nelson D.L. Weinstock G.M. Sakaki Y. Fujiyama A. Hattori M. Yada T. Toyoda A. Itoh T. Kawagoe C. Watanabe H. Totoki Y. Taylor T. Weissenbach J. Heilig R. Saurin W. Artiguenave F. Brottier P. Bruls T. Pelletier E. Robert C. Wincker P. Smith D.R. Doucette-Stamm L. Rubenfield M. Weinstock K. Lee H.M. Dubois J. Rosenthal A. Platzer M. Nyakatura G. Taudien S. Rump A. Yang H. Yu J. Wang J. Huang G. Gu J. Hood L. Rowen L. Madan A. Qin S. Davis R.W. Federspiel N.A. Abola A.P. Proctor M.J. Myers R.M. Schmutz J. Dickson M. Grimwood J. Cox D.R. Olson M.V. Kaul R. Raymond C. Shimizu N. Kawasaki K. Minoshima S. Evans G.A. Athanasiou M. Schultz R. Roe B.A. Chen F. Pan H. Ramser J. Lehrach H. Reinhardt R. McCombie W.R. de la Bastide M. Dedhia N. Blöcker H. Hornischer K. Nordsiek G. Agarwala R. Aravind L. Bailey J.A. Bateman A. Batzoglou S. Birney E. Bork P. Brown D.G. Burge C.B. Cerutti L. Chen H.C. Church D. Clamp M. Copley R.R. Doerks T. Eddy S.R. Eichler E.E. Furey T.S. Galagan J. Gilbert J.G. Harmon C. Hayashizaki Y. Haussler D. Hermjakob H. Hokamp K. Jang W. Johnson L.S. Jones T.A. Kasif S. Kaspryzk A. Kennedy S. Kent W.J. Kitts P. Koonin E.V. Korf I. Kulp D. Lancet D. Lowe T.M. McLysaght A. Mikkelsen T. Moran J.V. Mulder N. Pollara V.J. Ponting C.P. Schuler G. Schultz J. Slater G. Smit A.F. Stupka E. Szustakowki J. Thierry-Mieg D. Thierry-Mieg J. Wagner L. Wallis J. Wheeler R. Williams A. Wolf Y.I. Wolfe K.H. Yang S.P. Yeh R.F. Collins F. Guyer M.S. Peterson J. Felsenfeld A. Wetterstrand K.A. Patrinos A. Morgan M.J. de Jong P. Catanese J.J. Osoegawa K. Shizuya H. Choi S. Chen Y.J. Szustakowki J. Initial sequencing and analysis of the human genome Nature 409 6822 2001 860 921 10.1038/35057062 11237011
Lawitz E. Poordad F. Kowdley K.V. Cohen D.E. Podsadecki T. Siggelkow S. Larsen L. Menon R. Koev G. Tripathi R. Pilot-Matias T. Bernstein B. A phase 2a trial of 12-week interferon-free therapy with two direct-acting antivirals (ABT-450/r, ABT-072) and ribavirin in IL28B C/C patients with chronic hepatitis C genotype 1 J. Hepatol. 59 1 2013 18 23 10.1016/j.jhep.2013.02.009 23439262
Lawitz E.J. O'Riordan W.D. Asatryan A. Freilich B.L. Box T.D. Overcash J.S. Lovell S. Ng T.I. Liu W. Campbell A. Lin C.W. Yao B. Kort J. Potent antiviral activities of the direct-acting antivirals ABT-493 and ABT-530 with three-day monotherapy for hepatitis C virus genotype 1 infection Antimicrob. Agents Chemother. 60 3 2015 1546 1555 10.1128/aac.02264-15 26711747
Lawitz E. Reau N. Hinestrosa F. Rabinovitz M. Schiff E. Sheikh A. Younes Z. Herring R. Jr. Reddy K.R. Tran T. Bennett M. Nahass R. Yang J.C. Lu S. Dvory-Sobol H. Stamm L.M. Brainard D.M. McHutchison J.G. Pearlman B. Shiffman M. Hawkins T. Curry M. Jacobson I. Efficacy of sofosbuvir, velpatasvir, and GS-9857 in patients with genotype 1 hepatitis C virus infection in an open-label, phase 2 trial Gastroenterology 151 5 2016 893 901 10.1053/j.gastro.2016.07.039 e891 27486034
Lawitz E. Poordad F. Wells J. Hyland R.H. Yang Y. Dvory-Sobol H. Stamm L.M. Brainard D.M. McHutchison J.G. Landaverde C. Gutierrez J. Sofosbuvir-velpatasvir-voxilaprevir with or without ribavirin in direct-acting antiviral-experienced patients with genotype 1 hepatitis C virus Hepatology 65 6 2017 1803 1809 10.1002/hep.29130 28220512
Lebbé C. Blum L. Pellet C. Blanchard G. Vérola O. Morel P. Danne O. Calvo F. Clinical and biological impact of antiretroviral therapy with protease inhibitors on HIV-related Kaposi's sarcoma AIDS 12 7 1998 F45 F49 10.1097/00002030-199807000-00002 9619797
Lecronier M. Beurton A. Burrel S. Haudebourg L. Deleris R. Le Marec J. Virolle S. Nemlaghi S. Bureau C. Mora P. De Sarcus M. Clovet O. Duceau B. Grisot P.H. Pari M.H. Arzoine J. Clarac U. Boutolleau D. Raux M. Delemazure J. Faure M. Decavele M. Morawiec E. Mayaux J. Demoule A. Dres M. Comparison of hydroxychloroquine, lopinavir/ritonavir, and standard of care in critically ill patients with SARS-CoV-2 pneumonia: an opportunistic retrospective analysis Crit. Care 24 1 2020 418 10.1186/s13054-020-03117-9 32653015
Lee G.M. Shahian T. Baharuddin A. Gable J.E. Craik C.S. Enzyme inhibition by allosteric capture of an inactive conformation J. Mol. Biol. 411 5 2011 999 1016 10.1016/j.jmb.2011.06.032 21723875
Lee J. Worrall L.J. Vuckovic M. Rosell F.I. Gentile F. Ton A.T. Caveney N.A. Ban F. Cherkasov A. Paetzel M. Strynadka N.C.J. Crystallographic structure of wild-type SARS-CoV-2 main protease acyl-enzyme intermediate with physiological C-terminal autoprocessing site Nat. Commun. 11 1 2020 5877 10.1038/s41467-020-19662-4 33208735
Lei J. Hansen G. Nitsche C. Klein C.D. Zhang L. Hilgenfeld R. Crystal structure of Zika virus NS2B-NS3 protease in complex with a boronate inhibitor Science 353 6298 2016 503 505 10.1126/science.aag2419 27386922
Lei J. Kusov Y. Hilgenfeld R. Nsp3 of coronaviruses: structures and functions of a large multi-domain protein Antivir. Res. 149 2018 58 74 10.1016/j.antiviral.2017.11.001 29128390
Li H.-C. Yang C.-H. Lo S.-Y. Hepatitis C viral replication complex Viruses 13 3 2021 520 10.3390/v13030520 33809897
Liang G. Bushman F.D. The human virome: assembly, composition and host interactions Nat. Rev. Microbiol. 19 8 2021 514 527 10.1038/s41579-021-00536-5 33785903
Lin C. Lin K. Luong Y.-P. Rao B.G. Wei Y.-Y. Brennan D.L. Fulghum J.R. Hsiao H.-M. Ma S. Maxwell J.P. In vitro resistance studies of hepatitis C virus serine protease inhibitors, VX-950 and BILN 2061: structural analysis indicates different resistance mechanisms J. Biol. Chem. 279 17 2004 17508 17514 10.1074/jbc.M313020200 14766754
Lin T.I. Lenz O. Fanning G. Verbinnen T. Delouvroy F. Scholliers A. Vermeiren K. Rosenquist A. Edlund M. Samuelsson B. Vrang L. de Kock H. Wigerinck P. Raboisson P. Simmen K. In vitro activity and preclinical profile of TMC435350, a potent hepatitis C virus protease inhibitor Antimicrob. Agents Chemother. 53 4 2009 1377 1385 10.1128/aac.01058-08 19171797
Lin C.W. Dutta S. Asatryan A. Chiu Y.L. Wang H. Clifton J. 2nd Campbell A. Liu W. Pharmacokinetics, safety, and tolerability of single and multiple doses of ABT-493: a first-in-human study J. Pharmacol. Sci. 106 2 2017 645 651 10.1016/j.xphs.2016.10.007
Liu C. Shi W. Becker S.T. Schatz D.G. Liu B. Yang Y. Structural basis of mismatch recognition by a SARS-CoV-2 proofreading enzyme Science 373 6559 2021 1142 1146 10.1126/science.abi9310 34315827
Loharamtaweethong K. Vinyuvat S. Thammasiri J. Chitpakdee S. Supakatitham C. Puripat N. Impact of antiretroviral drugs on PD-L1 expression and copy number gains with clinical outcomes in HIV-positive and -negative locally advanced cervical cancers Oncol. Lett. 18 6 2019 5747 5758 10.3892/ol.2019.10963 31788048
Lohmann V. Koch J.O. Bartenschlager R. Processing pathways of the hepatitis C virus proteins J. Hepatol. 24 2 Suppl. l 1996 11 19 8836884
Louis J.M. Aniana A. Weber I.T. Sayer J.M. Inhibition of autoprocessing of natural variants and multidrug resistant mutant precursors of HIV-1 protease by clinical inhibitors Proc. Natl. Acad. Sci. U. S. A. 108 22 2011 9072 9077 10.1073/pnas.1102278108 21576495
Love R.A. Parge H.E. Wickersham J.A. Hostomsky Z. Habuka N. Moomaw E.W. Adachi T. Hostomska Z. The crystal structure of hepatitis C virus NS3 proteinase reveals a trypsin-like fold and a structural zinc binding site Cell 87 2 1996 331 342 10.1016/s0092-8674(00)81350-1 8861916
Luo D. Vasudevan S.G. Lescar J. The flavivirus NS2B-NS3 protease-helicase as a target for antiviral drug development Antivir. Res. 118 2015 148 158 10.1016/j.antiviral.2015.03.014 25842996
Lv Z. Cano K.E. Jia L. Drag M. Huang T.T. Olsen S.K. Targeting SARS-CoV-2 proteases for COVID-19 antiviral development Front. Chem. 9 2022 10.3389/fchem.2021.819165
Ma C. Sacco M.D. Hurst B. Townsend J.A. Hu Y. Szeto T. Zhang X. Tarbet B. Marty M.T. Chen Y. Wang J. Boceprevir, GC-376, and calpain inhibitors II, XII inhibit SARS-CoV-2 viral replication by targeting the viral main protease Cell Res. 30 8 2020 678 692 10.1038/s41422-020-0356-z 32541865
Madison V. Prongay A.J. Guo Z. Yao N. Pichardo J. Fischmann T. Strickland C. Myers J. Jr. Weber P.C. Beyer B.M. Ingram R. Hong Z. Prosise W.W. Ramanathan L. Taremi S.S. Yarosh-Tomaine T. Zhang R. Senior M. Yang R.S. Malcolm B. Arasappan A. Bennett F. Bogen S.L. Chen K. Jao E. Liu Y.T. Lovey R.G. Saksena A.K. Venkatraman S. Girijavallabhan V. Njoroge F.G. Key steps in the structure-based optimization of the hepatitis C virus NS3/4A protease inhibitor SCH5 03034 J. Synchrotron Radiat. 15 Pt 3 2008 204 207 10.1107/s0909049507064229 18421139
Majerová T. Konvalinka J. Enyzmes | HIV protease Jez J. Encyclopedia of Biological Chemistry III third ed. 2021 Elsevier Oxford 264 269
Majerová T. Novotný P. Precursors of viral proteases as distinct drug targets Viruses 13 10 2021 1981 34696411
Majerová T. Hoffman H. Majer F. Therapeutic targets for influenza-Perspectives in drug development Collect. Czech Chem. Commun. 75 1 2010 81 103 10.1135/cccc2009087
Majerová T. Novotný P. Krýsová E. Konvalinka J. Exploiting the unique features of Zika and Dengue proteases for inhibitor design Biochimie 166 2019 132 141 10.1016/j.biochi.2019.05.004 31077760
Majerová-Uhlíková T. Dantuma N.P. Lindsten K. Masucci M.G. Konvalinka J. Non-infectious fluorimetric assay for phenotyping of drug-resistant HIV proteinase mutants J. Clin. Virol. 36 1 2006 50 59 10.1016/j.jcv.2006.01.014 16527535
Mancuso P. Chen C. Kaminski R. Gordon J. Liao S. Robinson J.A. Smith M.D. Liu H. Sariyer I.K. Sariyer R. Peterson T.A. Donadoni M. Williams J.B. Siddiqui S. Bunnell B.A. Ling B. MacLean A.G. Burdo T.H. Khalili K. CRISPR based editing of SIV proviral DNA in ART treated non-human primates Nat. Commun. 11 1 2020 6065 10.1038/s41467-020-19821-7 33247091
Manns M.P. Buti M. Gane E. Pawlotsky J.-M. Razavi H. Terrault N. Younossi Z. Hepatitis C virus infection Nat. Rev. Dis. Prim. 3 1 2017 17006 10.1038/nrdp.2017.6
Marcelin A.G. Affolabi D. Lamotte C. Mohand H.A. Delaugerre C. Wirden M. Voujon D. Bossi P. Ktorza N. Bricaire F. Costagliola D. Katlama C. Peytavin G. Calvez V. Resistance profiles observed in virological failures after 24 weeks of amprenavir/ritonavir containing regimen in protease inhibitor experienced patients J. Med. Virol. 74 1 2004 16 20 10.1002/jmv.20140 15258963
Marcellin P. Forns X. Goeser T. Ferenci P. Nevens F. Carosi G. Drenth J.P. Serfaty L. De Backer K. Van Heeswijk R. Luo D. Picchio G. Beumont M. Telaprevir is effective given every 8 or 12 hours with ribavirin and peginterferon alfa-2a or -2b to patients with chronic hepatitis C Gastroenterology 140 2 2011 459 468 10.1053/j.gastro.2010.10.046 e451; quiz e414 21034744
Markham A. Keam S.J. Danoprevir: first global approval Drugs 78 12 2018 1271 1276 10.1007/s40265-018-0960-0 30117020
Marks K.M. Jacobson I.M. The first wave: HCV NS3 protease inhibitors telaprevir and boceprevir Antivir. Ther. 17 6 Pt B 2012 1119 1131 10.3851/imp2424 23188750
Marzolini C. Kuritzkes D.R. Marra F. Boyle A. Gibbons S. Flexner C. Pozniak A. Boffito M. Waters L. Burger D. Back D. Khoo S. Prescribing nirmatrelvir-ritonavir: how to recognize and manage drug-drug interactions Ann. Intern. Med. 175 5 2022 744 746 10.7326/m22-0281 35226530
Mathias A.A. German P. Murray B.P. Wei L. Jain A. West S. Warren D. Hui J. Kearney B.P. Pharmacokinetics and pharmacodynamics of GS-9350: a novel pharmacokinetic enhancer without anti-HIV activity Clin. Pharmacol. Ther. 87 3 2010 322 329 10.1038/clpt.2009.228 20043009
Mattei S. Anders M. Konvalinka J. Kräusslich H.G. Briggs J.A. Müller B. Induced maturation of human immunodeficiency virus J. Virol. 88 23 2014 13722 13731 10.1128/jvi.02271-14 25231305
Matthew A.N. Zephyr J. Nageswara Rao D. Henes M. Kamran W. Kosovrasti K. Hedger A.K. Lockbaum G.J. Timm J. Ali A. Kurt Yilmaz N. Schiffer C.A. Avoiding drug resistance by substrate envelope-guided design: toward potent and robust HCV NS3/4A protease inhibitors mBio 11 2 2020 10.1128/mBio.00172-20
Matthew A.N. Leidner F. Lockbaum G.J. Henes M. Zephyr J. Hou S. Rao D.N. Timm J. Rusere L.N. Ragland D.A. Paulsen J.L. Prachanronarong K. Soumana D.I. Nalivaika E.A. Kurt Yilmaz N. Ali A. Schiffer C.A. Drug design strategies to avoid resistance in direct-acting antivirals and beyond Chem. Rev. 121 6 2021 3238 3270 10.1021/acs.chemrev.0c00648 33410674
Maxwell E. Treatment of herpes keratitis with 5-iodo-2-deoxyuridine (IDU). A clinical evaluation of 1,500 cases Am. J. Ophthalmol. 56 1963 571 573 10.1016/0002-9394(63)90006-0 14070708
McArthur D.B. Emerging infectious diseases Nurs. Clin. 54 2 2019 297 311 10.1016/j.cnur.2019.02.006
Meganck R.M. Baric R.S. Developing therapeutic approaches for twenty-first-century emerging infectious viral diseases Nat. Med. 27 3 2021 401 410 10.1038/s41591-021-01282-0 33723456
Mehta R.M. Bansal S. Bysani S. Kalpakam H. A shorter symptom onset to remdesivir treatment (SORT) interval is associated with a lower mortality in moderate-to-severe COVID-19: a real-world analysis Int. J. Infect. Dis. 106 2021 71 77 10.1016/j.ijid.2021.02.092 33647517
Mensa F.J. Lovell S. Pilot-Matias T. Liu W. Glecaprevir/pibrentasvir for the treatment of chronic hepatitis C virus infection Future Microbiol. 14 2019 89 110 10.2217/fmb-2018-0233 30499343
Merolli A. Kasaei L. Ramasamy S. Kolloli A. Kumar R. Subbian S. Feldman L.C. An intra-cytoplasmic route for SARS-CoV-2 transmission unveiled by Helium-ion microscopy Sci. Rep. 12 1 2022 3794 10.1038/s41598-022-07867-0 35260703
Miao M. Jing X. De Clercq E. Li G. Danoprevir for the treatment of hepatitis C virus infection: design, development, and place in therapy Drug Des. Dev. Ther. 14 2020 2759 2774 10.2147/dddt.S254754
Miller S.L. Amprenavir approved for HIV treatment Am. J. Health Syst. Pharm. 56 11 1999 1057 1058 10.1093/ajhp/56.11.1057
Mitsuya H. Weinhold K.J. Furman P.A. St Clair M.H. Lehrman S.N. Gallo R.C. Bolognesi D. Barry D.W. Broder S. 3'-Azido-3'-deoxythymidine (BW A509U): an antiviral agent that inhibits the infectivity and cytopathic effect of human T-lymphotropic virus type III/lymphadenopathy-associated virus in vitro Proc. Natl. Acad. Sci. U. S. A. 82 20 1985 7096 7100 10.1073/pnas.82.20.7096 2413459
Mody V. Ho J. Wills S. Mawri A. Lawson L. Ebert M. Fortin G.M. Rayalam S. Taval S. Identification of 3-chymotrypsin like protease (3CLPro) inhibitors as potential anti-SARS-CoV-2 agents Commun. Biol. 4 1 2021 93 10.1038/s42003-020-01577-x 33473151
Morcilla V. Bacchus-Souffan C. Fisher K. Horsburgh B.A. Hiener B. Wang X.Q. Schlub T.E. Fitch M. Hoh R. Hecht F.M. Martin J.N. Deeks S.G. Hellerstein M.K. McCune J.M. Hunt P.W. Palmer S. HIV-1 genomes are enriched in memory CD4(+) T-cells with short half-lives mBio 12 5 2021 e0244721 10.1128/mBio.02447-21
Morello J. Rodriguez-Novoa S. Jimenez-Nacher I. Soriano V. Drug interactions of tipranavir, a new HIV protease inhibitor Drug Metabol. Lett. 1 1 2007 81 84 10.2174/187231207779814256
Morris A. McCorkindale W. Consortium T.C.M. Drayman N. Chodera J.D. Tay S. London N. Lee A.A. Discovery of SARS-CoV-2 main protease inhibitors using a synthesis-directed de novo design model Chem. Commun. 57 48 2021 5909 5912 10.1039/d1cc00050k
Murphy R.L. Gulick R.M. DeGruttola V. D'Aquila R.T. Eron J.J. Sommadossi J.P. Currier J.S. Smeaton L. Frank I. Caliendo A.M. Gerber J.G. Tung R. Kuritzkes D.R. Treatment with amprenavir alone or amprenavir with zidovudine and lamivudine in adults with human immunodeficiency virus infection. AIDS Clinical Trials Group 347 Study Team J. Infect. Dis. 179 4 1999 808 816 10.1086/314668 10068575
Nalam M.N.L. Ali A. Altman M.D. Reddy G.S.K.K. Chellappan S. Kairys V. Ozen A. Cao H. Gilson M.K. Tidor B. Rana T.M. Schiffer C.A. Evaluating the substrate-envelope hypothesis: structural analysis of novel HIV-1 protease inhibitors designed to be robust against drug resistance J. Virol. 84 10 2010 5368 5378 10.1128/JVI.02531-09 20237088
Napoli C. Benincasa G. Criscuolo C. Faenza M. Liberato C. Rusciano M. Immune reactivity during COVID-19: implications for treatment Immunol. Lett. 231 2021 28 34 10.1016/j.imlet.2021.01.001 33421440
Nguyen T. McNicholl I. Custodio J.M. Szwarcberg J. Piontkowsky D. Drug interactions with cobicistat- or ritonavir-boosted elvitegravir AIDS Rev. 18 2 2016 101 111 27196356
Nie S. Yao Y. Wu F. Wu X. Zhao J. Hua Y. Wu J. Huo T. Lin Y.L. Kneubehl A.R. Vogt M.B. Ferreon J. Rico-Hesse R. Song Y. Synthesis, structure-activity relationships, and antiviral activity of allosteric inhibitors of flavivirus NS2B-NS3 protease J. Med. Chem. 64 5 2021 2777 2800 10.1021/acs.jmedchem.0c02070 33596380
Nijhuis M. van Maarseveen N.M. Lastere S. Schipper P. Coakley E. Glass B. Rovenska M. de Jong D. Chappey C. Goedegebuure I.W. Heilek-Snyder G. Dulude D. Cammack N. Brakier-Gingras L. Konvalinka J. Parkin N. Kräusslich H.G. Brun-Vezinet F. Boucher C.A. A novel substrate-based HIV-1 protease inhibitor drug resistance mechanism PLoS Med. 4 1 2007 e36 10.1371/journal.pmed.0040036 17227139
Nijhuis M. van Maarseveen N.M. Boucher C.A. Antiviral resistance and impact on viral replication capacity: evolution of viruses under antiviral pressure occurs in three phases Handb. Exp. Pharmacol. 189 2009 299 320 10.1007/978-3-540-79086-0_11
Njoroge F.G. Chen K.X. Shih N.Y. Piwinski J.J. Challenges in modern drug discovery: a case study of boceprevir, an HCV protease inhibitor for the treatment of hepatitis C virus infection Acc. Chem. Res. 41 1 2008 50 59 10.1021/ar700109k 18193821
Oerlemans R. Ruiz-Moreno A.J. Cong Y. Dinesh Kumar N. Velasco-Velazquez M.A. Neochoritis C.G. Smith J. Reggiori F. Groves M.R. Dömling A. Repurposing the HCV NS3–4A protease drug boceprevir as COVID-19 therapeutics RSC Med. Chem. 12 3 2021 370 379 10.1039/D0MD00367K
Okubo K. Sato A. Isono M. Asano T. Asano T. Nelfinavir induces endoplasmic reticulum stress and sensitizes renal cancer cells to TRAIL Anticancer Res. 38 8 2018 4505 4514 10.21873/anticanres.12754 30061216
Osipiuk J. Azizi S.A. Dvorkin S. Endres M. Jedrzejczak R. Jones K.A. Kang S. Kathayat R.S. Kim Y. Lisnyak V.G. Maki S.L. Nicolaescu V. Taylor C.A. Tesar C. Zhang Y.A. Zhou Z. Randall G. Michalska K. Snyder S.A. Dickinson B.C. Joachimiak A. Structure of papain-like protease from SARS-CoV-2 and its complexes with non-covalent inhibitors Nat. Commun. 12 1 2021 743 10.1038/s41467-021-21060-3 33531496
Owen D.R. Allerton C.M.N. Anderson A.S. Aschenbrenner L. Avery M. Berritt S. Boras B. Cardin R.D. Carlo A. Coffman K.J. Dantonio A. Di L. Eng H. Ferre R. Gajiwala K.S. Gibson S.A. Greasley S.E. Hurst B.L. Kadar E.P. Kalgutkar A.S. Lee J.C. Lee J. Liu W. Mason S.W. Noell S. Novak J.J. Obach R.S. Ogilvie K. Patel N.C. Pettersson M. Rai D.K. Reese M.R. Sammons M.F. Sathish J.G. Singh R.S.P. Steppan C.M. Stewart A.E. Tuttle J.B. Updyke L. Verhoest P.R. Wei L. Yang Q. Zhu Y. An oral SARS-CoV-2 M(pro) inhibitor clinical candidate for the treatment of COVID-19 Science 374 6575 2021 1586 1593 10.1126/science.abl4784 34726479
Pablos I. Machado Y. de Jesus H.C.R. Mohamud Y. Kappelhoff R. Lindskog C. Vlok M. Bell P.A. Butler G.S. Grin P.M. Cao Q.T. Nguyen J.P. Solis N. Abbina S. Rut W. Vederas J.C. Szekely L. Szakos A. Drag M. Kizhakkedathu J.N. Mossman K. Hirota J.A. Jan E. Luo H. Banerjee A. Overall C.M. Mechanistic insights into COVID-19 by global analysis of the SARS-CoV-2 3CL(pro) substrate degradome Cell Rep. 37 4 2021 109892 10.1016/j.celrep.2021.109892
Palich R. Makinson A. Veyri M. Guihot A. Valantin M.A. Brégigeon-Ronot S. Poizot-Martin I. Solas C. Grabar S. Martin-Blondel G. Spano J.P. Kaposi's sarcoma in virally suppressed people living with HIV: an emerging condition Cancers 13 22 2021 10.3390/cancers13225702
Palumbo E. Pegylated interferon and ribavirin treatment for hepatitis C virus infection Ther. Adv. Chronic. Dis. 2 1 2011 39 45 10.1177/2040622310384308 23251740
Pan Y.-Y. Wang S.-M. Huang K.-J. Chiang C.-C. Wang C.-T. Placement of leucine zipper motifs at the carboxyl terminus of HIV-1 protease significantly reduces virion production PLoS One 7 3 2012 e32845 10.1371/journal.pone.0032845
Pankey G.A. Sabath L.D. Clinical relevance of bacteriostatic versus bactericidal mechanisms of action in the treatment of gram-positive bacterial infections Clin. Infect. Dis. 38 6 2004 864 870 10.1086/381972 14999632
Pardesbio https://www.pardesbio.com/pipeline/
Paredes R. Clotet B. Clinical management of HIV-1 resistance Antivir. Res. 85 1 2010 245 265 10.1016/j.antiviral.2009.09.015 19808056
Park J.H. Sayer J.M. Aniana A. Yu X. Weber I.T. Harrison R.W. Louis J.M. Binding of clinical inhibitors to a model precursor of a rationally selected multidrug resistant HIV-1 protease is significantly weaker than that to the released mature enzyme Biochemistry 55 16 2016 2390 2400 10.1021/acs.biochem.6b00012 27039930
Park Y.J. Woo H.Y. Heo J. Park S.G. Hong Y.M. Yoon K.T. Kim D.U. Kim G.H. Kim H.H. Song G.A. Cho M. Real-life effectiveness and safety of glecaprevir/pibrentasvir for Korean patients with chronic hepatitis C at a single institution Gut Liver. 15 3 2021 440 450 10.5009/gnl19393 32839365
Patick A.K. Duran M. Cao Y. Shugarts D. Keller M.R. Mazabel E. Knowles M. Chapman S. Kuritzkes D.R. Markowitz M. Genotypic and phenotypic characterization of human immunodeficiency virus type 1 variants isolated from patients treated with the protease inhibitor nelfinavir Antimicrob. Agents Chemother. 42 10 1998 2637 2644 10.1128/AAC.42.10.2637 9756769
Pawar S.D. Freas C. Weber I.T. Harrison R.W. Analysis of drug resistance in HIV protease BMC bioinform. 19 Suppl. 11 2018 362 10.1186/s12859-018-2331-y 362
Payne S. Introduction to RNA viruses Viruses 2017 97 105 10.1016/B978-0-12-803109-4.00010-6 28468299
Perni R.B. Britt S.D. Court J.C. Courtney L.F. Deininger D.D. Farmer L.J. Gates C.A. Harbeson S.L. Kim J.L. Landro J.A. Levin R.B. Luong Y.P. O'Malley E.T. Pitlik J. Rao B.G. Schairer W.C. Thomson J.A. Tung R.D. Van Drie J.H. Wei Y. Inhibitors of hepatitis C virus NS3.4A protease 1. Non-charged tetrapeptide variants Bioorg. Med. Chem. Lett. 13 22 2003 4059 4063 10.1016/j.bmcl.2003.08.050 14592508
Perni R.B. Farmer L.J. Cottrell K.M. Court J.J. Courtney L.F. Deininger D.D. Gates C.A. Harbeson S.L. Kim J.L. Lin C. Lin K. Luong Y.P. Maxwell J.P. Murcko M.A. Pitlik J. Rao B.G. Schairer W.C. Tung R.D. Van Drie J.H. Wilson K. Thomson J.A. Inhibitors of hepatitis C virus NS3.4A protease. Part 3: P2 proline variants Bioorg. Med. Chem. Lett. 14 8 2004 1939 1942 10.1016/j.bmcl.2004.01.078 15050632
Perni R.B. Pitlik J. Britt S.D. Court J.J. Courtney L.F. Deininger D.D. Farmer L.J. Gates C.A. Harbeson S.L. Levin R.B. Lin C. Lin K. Moon Y.C. Luong Y.P. O'Malley E.T. Rao B.G. Thomson J.A. Tung R.D. Van Drie J.H. Wei Y. Inhibitors of hepatitis C virus NS3.4A protease 2. Warhead SAR and optimization Bioorg. Med. Chem. Lett. 14 6 2004 1441 1446 10.1016/j.bmcl.2004.01.022 15006379
Perni R.B. Chandorkar G. Cottrell K.M. Gates C.A. Lin C. Lin K. Luong Y.P. Maxwell J.P. Murcko M.A. Pitlik J. Rao G. Schairer W.C. Van Drie J. Wei Y. Inhibitors of hepatitis C virus NS3.4A protease. Effect of P4 capping groups on inhibitory potency and pharmacokinetics Bioorg. Med. Chem. Lett. 17 12 2007 3406 3411 10.1016/j.bmcl.2007.03.090 17482818
Pettit S.C. Gulnik S. Everitt L. Kaplan A.H. The dimer interfaces of protease and extra-protease domains influence the activation of protease and the specificity of GagPol cleavage J. Virol. 77 1 2003 366 374 10.1128/jvi.77.1.366-374.2003 12477841
Pettit S.C. Everitt L.E. Choudhury S. Dunn B.M. Kaplan A.H. Initial cleavage of the human immunodeficiency virus type 1 GagPol precursor by its activated protease occurs by an intramolecular mechanism J. Virol. 78 16 2004 8477 8485 10.1128/jvi.78.16.8477-8485.2004 15280456
Pettit S.C. Clemente J.C. Jeung J.A. Dunn B.M. Kaplan A.H. Ordered processing of the human immunodeficiency virus type 1 GagPol precursor is influenced by the context of the embedded viral protease J. Virol. 79 16 2005 10601 10607 10.1128/jvi.79.16.10601-10607.2005 16051852
Pia L. Rowland-Jones S. Omicron entry route Nat. Rev. Immunol. 22 144 2022 10.1038/s41577-022-00681-9
Pierson T.C. Diamond M.S. The continued threat of emerging flaviviruses Nat. Microbiol. 5 6 2020 796 812 10.1038/s41564-020-0714-0 32367055
Piliero P.J. Atazanavir: a novel HIV-1 protease inhibitor Expet Opin. Invest. Drugs 11 9 2002 1295 1301 10.1517/13543784.11.9.1295
Pokorná J. Machala L. Rezáčová P. Konvalinka J. Current and novel inhibitors of HIV protease Viruses 1 3 2009 1209 1239 10.3390/v1031209 21994591
Pol S. Lagaye S. The remarkable history of the hepatitis C virus Gene Immun. 20 5 2019 436 446 10.1038/s41435-019-0066-z
Preston B.D. Poiesz B.J. Loeb L.A. Fidelity of HIV-1 reverse transcriptase Science 242 4882 1988 1168 1171 10.1126/science.2460924 2460924
Priya S. Tripathi G. Singh D.B. Jain P. Kumar A. Machine learning approaches and their applications in drug discovery and design Chem. Biol. Drug Des. 100 1 2022 136 153 10.1111/cbdd.14057 35426249
Proia E. Ragno A. Antonini L. Sabatino M. Mladenovič M. Capobianco R. Ragno R. Ligand-based and structure-based studies to develop predictive models for SARS-CoV-2 main protease inhibitors through the 3d-qsar.com portal J. Comput. Aided Mol. Des. 36 7 2022 483 505 10.1007/s10822-022-00460-7 35716228
Prongay A.J. Guo Z. Yao N. Pichardo J. Fischmann T. Strickland C. Myers J. Jr. Weber P.C. Beyer B.M. Ingram R. Hong Z. Prosise W.W. Ramanathan L. Taremi S.S. Yarosh-Tomaine T. Zhang R. Senior M. Yang R.S. Malcolm B. Arasappan A. Bennett F. Bogen S.L. Chen K. Jao E. Liu Y.T. Lovey R.G. Saksena A.K. Venkatraman S. Girijavallabhan V. Njoroge F.G. Madison V. Discovery of the HCV NS3/4A protease inhibitor (1R,5S)-N-[3-amino-1-(cyclobutylmethyl)-2,3-dioxopropyl]-3- [2(S)-[[[(1,1-dimethylethyl)amino]carbonyl]amino]-3,3-dimethyl-1-oxobutyl]- 6,6-dimethyl-3-azabicyclo[3.1.0]hexan-2(S)-carboxamide (Sch 503034) II. Key steps in structure-based optimization J. Med. Chem. 50 10 2007 2310 2318 10.1021/jm060173k 17444623
Prusoff W.H. Synthesis and biological activities of iododeoxyuridine, an analog of thymidine Biochim. Biophys. Acta 32 1 1959 295 296 10.1016/0006-3002(59)90597-9 13628760
Qin E. Zhu Q. Yu M. Fan B. Chang G. Si B. Yang B. Peng W. Jiang T. Liu B. Deng Y. Liu H. Zhang Y. Wang C. Li Y. Gan Y. Li X. Lü F. Tan G. Cao W. Yang R. Wang J. Li W. Xu Z. Li Y. Wu Q. Lin W. Chen W. Tang L. Deng Y. Han Y. Li C. Lei M. Li G. Li W. Lü H. Shi J. Tong Z. Zhang F. Li S. Liu B. Liu S. Dong W. Wang J. Wong G.K. Yu J. Yang H. A complete sequence and comparative analysis of a SARS-associated virus (Isolate BJ01) Chin. Sci. Bull. 48 10 2003 941 948 10.1007/bf03184203 32214698
Raboisson P. de Kock H. Rosenquist A. Nilsson M. Salvador-Oden L. Lin T.I. Roue N. Ivanov V. Wähling H. Wickström K. Hamelink E. Edlund M. Vrang L. Vendeville S. Van de Vreken W. McGowan D. Tahri A. Hu L. Boutton C. Lenz O. Delouvroy F. Pille G. Surleraux D. Wigerinck P. Samuelsson B. Simmen K. Structure-activity relationship study on a novel series of cyclopentane-containing macrocyclic inhibitors of the hepatitis C virus NS3/4A protease leading to the discovery of TMC435350 Bioorg. Med. Chem. Lett. 18 17 2008 4853 4858 10.1016/j.bmcl.2008.07.088 18678486
Ramanathan S. Custodio J.M. Wei X. Wang H. Fordyce M. Dave A. Ling K.H. Szwarcberg J. Kearney B.P. Pharmacokinetics of Co-formulated elvitegravir/cobicistat/emtricitabine/tenofovir disoproxil fumarate after switch from efavirenz/emtricitabine/tenofovir disoproxil fumarate in healthy subjects J. Acquir. Immune Defic. Syndr. 72 3 2016 281 288 10.1097/qai.0000000000000959 26885802
Ratia K. Pegan S. Takayama J. Sleeman K. Coughlin M. Baliji S. Chaudhuri R. Fu W. Prabhakar B.S. Johnson M.E. Baker S.C. Ghosh A.K. Mesecar A.D. A noncovalent class of papain-like protease/deubiquitinase inhibitors blocks SARS virus replication Proc. Natl. Acad. Sci. U. S. A. 105 42 2008 16119 16124 10.1073/pnas.0805240105 18852458
Reesink H.W. Fanning G.C. Farha K.A. Weegink C. Van Vliet A. Van 't Klooster G. Lenz O. Aharchi F. Mariën K. Van Remoortere P. de Kock H. Broeckaert F. Meyvisch P. Van Beirendonck E. Simmen K. Verloes R. Rapid HCV-RNA decline with once daily TMC435: a phase I study in healthy volunteers and hepatitis C patients Gastroenterology 138 3 2010 913 921 10.1053/j.gastro.2009.10.033 19852962
Ribeiro R.M. Li H. Wang S. Stoddard M.B. Learn G.H. Korber B.T. Bhattacharya T. Guedj J. Parrish E.H. Hahn B.H. Shaw G.M. Perelson A.S. Quantifying the diversification of hepatitis C virus (HCV) during primary infection: estimates of the in vivo mutation rate PLoS Pathog. 8 8 2012 e1002881 10.1371/journal.ppat.1002881
Rittweger M. Arastéh K. Clinical pharmacokinetics of darunavir Clin. Pharmacokinet. 46 9 2007 739 756 10.2165/00003088-200746090-00002 17713972
Roberts J.D. Bebenek K. Kunkel T.A. The accuracy of reverse transcriptase from HIV-1 Science 242 4882 1988 1171 1173 10.1126/science.2460925 2460925
Roberts N.A. Martin J.A. Kinchington D. Broadhurst A.V. Craig J.C. Duncan I.B. Galpin S.A. Handa B.K. Kay J. Kröhn A. Rational design of peptide-based HIV proteinase inhibitors Science 248 4953 1990 358 361 10.1126/science.2183354 2183354
Robins M.J. Robins R.K. The synthesis of 2′,3′-dideoxyadenosine from 2′-deoxyadenosine J. Am. Chem. Soc. 86 17 1964 3585 3586
Rodriguez-Torres M. Glass S. Hill J. Freilich B. Hassman D. Di Bisceglie A.M. Taylor J.G. Kirby B.J. Dvory-Sobol H. Yang J.C. An D. Stamm L.M. Brainard D.M. Kim S. Krefetz D. Smith W. Marbury T. Lawitz E. GS-9857 in patients with chronic hepatitis C virus genotype 1-4 infection: a randomized, double-blind, dose-ranging phase 1 study J. Viral Hepat. 23 8 2016 614 622 10.1111/jvh.12527 26957110
Rosenberg S. Recent advances in the molecular biology of hepatitis C virus J. Mol. Biol. 313 3 2001 451 464 10.1006/jmbi.2001.5055 11676530
Rosenquist Å. Samuelsson B. Johansson P.O. Cummings M.D. Lenz O. Raboisson P. Simmen K. Vendeville S. de Kock H. Nilsson M. Horvath A. Kalmeijer R. de la Rosa G. Beumont-Mauviel M. Discovery and development of simeprevir (TMC435), a HCV NS3/4A protease inhibitor J. Med. Chem. 57 5 2014 1673 1693 10.1021/jm401507s 24446688
Rubin R. Baricitinib is first approved COVID-19 immunomodulatory treatment JAMA 327 23 2022 2281 10.1001/jama.2022.9846 2281
Rubin R. From positive to negative to positive again-the mystery of why COVID-19 rebounds in some patients who take Paxlovid JAMA 327 24 2022 2380 2382 10.1001/jama.2022.9925 35675094
Rut W. Lv Z. Zmudzinski M. Patchett S. Nayak D. Snipas S.J. El Oualid F. Huang T.T. Bekes M. Drag M. Olsen S.K. Activity profiling and crystal structures of inhibitor-bound SARS-CoV-2 papain-like protease: a framework for anti-COVID-19 drug design Sci. Adv. 6 42 2020 10.1126/sciadv.abd4596
Rut W. Groborz K. Zhang L. Sun X. Zmudzinski M. Pawlik B. Wang X. Jochmans D. Neyts J. Młynarski W. Hilgenfeld R. Drag M. SARS-CoV-2 M(pro) inhibitors and activity-based probes for patient-sample imaging Nat. Chem. Biol. 17 2 2021 222 228 10.1038/s41589-020-00689-z 33093684
Saab S. Gonzalez Y.S. Huber C. Wang A. Juday T. Cost-effectiveness of ombitasvir/paritaprevir/ritonavir, Dasabuvir+Ribavirin for US post-liver transplant recurrent genotype 1 HCV Liver Int. 36 4 2016 515 521 10.1111/liv.13033 26610059
Sadybekov A.A. Sadybekov A.V. Liu Y. Iliopoulos-Tsoutsouvas C. Huang X.P. Pickett J. Houser B. Patel N. Tran N.K. Tong F. Zvonok N. Jain M.K. Savych O. Radchenko D.S. Nikas S.P. Petasis N.A. Moroz Y.S. Roth B.L. Makriyannis A. Katritch V. Synthon-based ligand discovery in virtual libraries of over 11 billion compounds Nature 601 7893 2022 452 459 10.1038/s41586-021-04220-9 34912117
Santolini E. Pacini L. Fipaldini C. Migliaccio G. Monica N. The NS2 protein of hepatitis C virus is a transmembrane polypeptide J. Virol. 69 12 1995 7461 7471 10.1128/jvi.69.12.7461-7471.1995 7494252
Saribas A.S. Coric P. Bouaziz S. Safak M. Expression of novel proteins by polyomaviruses and recent advances in the structural and functional features of agnoprotein of JC virus, BK virus, and simian virus 40 J. Cell. Physiol. 234 6 2019 8295 8315 10.1002/jcp.27715 30390301
Sarrazin C. Hézode C. Zeuzem S. Pawlotsky J.M. Antiviral strategies in hepatitis C virus infection J. Hepatol. 56 Suppl. 1 2012 S88 S100 10.1016/s0168-8278(12)60010-5 22300469
Sasaki M. Tabata K. Kishimoto M. Itakura Y. Kobayashi H. Ariizumi T. Uemura K. Toba S. Kusakabe S. Maruyama Y. Iida S. Nakajima N. Suzuki T. Yoshida S. Nobori H. Sanaki T. Kato T. Shishido T. Hall W.W. Orba Y. Sato A. Sawa H. S-217622, a SARS-CoV-2 main protease inhibitor, decreases viral load and ameliorates COVID-19 severity in hamsters Sci. Transl. Med. 2022 10.1126/scitranslmed.abq4064 eabq4064
Sasková K.G. Kozísek M. Lepsík M. Brynda J. Rezácová P. Václavíková J. Kagan R.M. Machala L. Konvalinka J. Enzymatic and structural analysis of the I47A mutation contributing to the reduced susceptibility to HIV protease inhibitor lopinavir Protein Sci. 17 9 2008 1555 1564 10.1110/ps.036079.108 18560011
Sasková K.G. Kozísek M. Rezácová P. Brynda J. Yashina T. Kagan R.M. Konvalinka J. Molecular characterization of clinical isolates of human immunodeficiency virus resistant to the protease inhibitor darunavir J. Virol. 83 17 2009 8810 8818 10.1128/jvi.00451-09 19535439
Schneider J. Kent S.B. Enzymatic activity of a synthetic 99 residue protein corresponding to the putative HIV-1 protease Cell 54 3 1988 363 368 10.1016/0092-8674(88)90199-7 3293801
Schrödinger L. DeLano W. PyMOL 2020 Retrieved from http://www.pymol.org/pymol
Seelmeier S. Schmidt H. Turk V. von der Helm K. Human immunodeficiency virus has an aspartic-type protease that can be inhibited by pepstatin A Proc. Natl. Acad. Sci. U. S. A. 85 18 1988 6612 6616 10.1073/pnas.85.18.6612 3045820
Seiwert S.D. Andrews S.W. Jiang Y. Serebryany V. Tan H. Kossen K. Rajagopalan P.T. Misialek S. Stevens S.K. Stoycheva A. Hong J. Lim S.R. Qin X. Rieger R. Condroski K.R. Zhang H. Do M.G. Lemieux C. Hingorani G.P. Hartley D.P. Josey J.A. Pan L. Beigelman L. Blatt L.M. Preclinical characteristics of the hepatitis C virus NS3/4A protease inhibitor ITMN-191 (R7227) Antimicrob. Agents Chemother. 52 12 2008 4432 4441 10.1128/aac.00699-08 18824605
Shah D. Freas C. Weber I.T. Harrison R.W. Evolution of drug resistance in HIV protease BMC Bioinf. 21 Suppl. 18 2020 497 10.1186/s12859-020-03825-7
Sham H.L. Kempf D.J. Molla A. Marsh K.C. Kumar G.N. Chen C.M. Kati W. Stewart K. Lal R. Hsu A. Betebenner D. Korneyeva M. Vasavanonda S. McDonald E. Saldivar A. Wideburg N. Chen X. Niu P. Park C. Jayanti V. Grabowski B. Granneman G.R. Sun E. Japour A.J. Leonard J.M. Plattner J.J. Norbeck D.W. ABT-378, a highly potent inhibitor of the human immunodeficiency virus protease Antimicrob. Agents Chemother. 42 12 1998 3218 3224 10.1128/aac.42.12.3218 9835517
Shaqra A.M. Zvornicanin S.N. Huang Q.Y.J. Lockbaum G.J. Knapp M. Tandeske L. Bakan D.T. Flynn J. Bolon D.N.A. Moquin S. Dovala D. Kurt Yilmaz N. Schiffer C.A. Defining the substrate envelope of SARS-CoV-2 main protease to predict and avoid drug resistance Nat. Commun. 13 1 2022 3556 10.1038/s41467-022-31210-w 35729165
Sharp P.M. Hahn B.H. Origins of HIV and the AIDS pandemic Cold Spring Harb. Perspect. Med. 1 1 2011 a006841 10.1101/cshperspect.a006841 22229120
Sheahan T.P. Sims A.C. Graham R.L. Menachery V.D. Gralinski L.E. Case J.B. Leist S.R. Pyrc K. Feng J.Y. Trantcheva I. Bannister R. Park Y. Babusis D. Clarke M.O. Mackman R.L. Spahn J.E. Palmiotti C.A. Siegel D. Ray A.S. Cihlar T. Jordan R. Denison M.R. Baric R.S. Broad-spectrum antiviral GS-5734 inhibits both epidemic and zoonotic coronaviruses Sci. Transl. Med. 9 396 2017 10.1126/scitranslmed.aal3653
Shehu-Xhilaga M. Crowe S.M. Mak J. Maintenance of the gag/gag-pol ratio is important for human immunodeficiency virus type 1 RNA dimerization and viral infectivity J. Virol. 75 4 2001 1834 1841 10.1128/JVI.75.4.1834-1841.2001 11160682
Sheik Amamuddy O. Afriyie Boateng R. Barozi V. Wavinya Nyamai D. Tastan Bishop Ö. Novel dynamic residue network analysis approaches to study allosteric modulation: SARS-CoV-2 Mpro and its evolutionary mutations as a case study Comput. Struct. Biotechnol. J. 19 2021 6431 6455 10.1016/j.csbj.2021.11.016 34849191
Shekhar C. Nasam R. Paipuri S.R. Kumar P. Nayani K. Pabbaraja S. Mainkar P.S. Chandrasekhar S. Total synthesis of antiviral drug, nirmatrelvir (PF-07321332) Tetrahedron Chem. 4 2022 100033 10.1016/j.tchem.2022.100033
Shepherd S.J. Abdelrahman T. MacLean A.R. Thomson E.C. Aitken C. Gunson R.N. Prevalence of HCV NS3 pre-treatment resistance associated amino acid variants within a Scottish cohort J. Clin. Virol. 65 2015 50 53 10.1016/j.jcv.2015.02.005 25766988
Sherman E.M. Worley M.V. Unger N.R. Gauthier T.P. Schafer J.J. Cobicistat: review of a pharmacokinetic enhancer for HIV infection Clin. Therapeut. 37 9 2015 1876 1893 10.1016/j.clinthera.2015.07.022
Shetty B.V. Kosa M.B. Khalil D.A. Webber S. Preclinical pharmacokinetics and distribution to tissue of AG1343, an inhibitor of human immunodeficiency virus type 1 protease Antimicrob. Agents Chemother. 40 1 1996 110 114 10.1128/aac.40.1.110 8787890
Shimba N. Nomura A.M. Marnett A.B. Craik C.S. Herpesvirus protease inhibition by dimer disruption J. Virol. 78 12 2004 6657 6665 10.1128/jvi.78.12.6657-6665.2004 15163756
Shin D. Mukherjee R. Grewe D. Bojkova D. Baek K. Bhattacharya A. Schulz L. Widera M. Mehdipour A.R. Tascher G. Geurink P.P. Wilhelm A. van der Heden van Noort G.J. Ovaa H. Müller S. Knobeloch K.-P. Rajalingam K. Schulman B.A. Cinatl J. Hummer G. Ciesek S. Dikic I. Papain-like protease regulates SARS-CoV-2 viral spread and innate immunity Nature 587 7835 2020 657 662 10.1038/s41586-020-2601-5 32726803
Shuai H. Chan J.F. Hu B. Chai Y. Yuen T.T. Yin F. Huang X. Yoon C. Hu J.C. Liu H. Shi J. Liu Y. Zhu T. Zhang J. Hou Y. Wang Y. Lu L. Cai J.P. Zhang A.J. Zhou J. Yuan S. Brindley M.A. Zhang B.Z. Huang J.D. To K.K. Yuen K.Y. Chu H. Attenuated replication and pathogenicity of SARS-CoV-2 B.1.1.529 Omicron Nature 603 7902 2022 693 699 10.1038/s41586-022-04442-5 35062016
Siegel R.D. Classification of Human Viruses. Principles and Practice of Pediatric Infectious Diseases 2018 e1041 10.1016/B978-0-323-40181-4.00201-2 1044-1048
Simmonds P. Aiewsakun P. Katzourakis A. Prisoners of war - host adaptation and its constraints on virus evolution Nat. Rev. Microbiol. 17 5 2019 321 328 10.1038/s41579-018-0120-2 30518814
Singh T.U. Parida S. Lingaraju M.C. Kesavan M. Kumar D. Singh R.K. Drug repurposing approach to fight COVID-19 Pharmacol. Rep. 72 6 2020 1479 1508 10.1007/s43440-020-00155-6 32889701
Skwarecki A.S. Nowak M.G. Milewska M.J. Amino acid and peptide-based antiviral agents ChemMedChem 16 20 2021 3106 3135 10.1002/cmdc.202100397 34254457
Smart T. FDA triple header. Food and Drug Administration GMHC Treat. Iss. 9 11 1995 1 3
Smith D.B. Bukh J. Kuiken C. Muerhoff A.S. Rice C.M. Stapleton J.T. Simmonds P. Expanded classification of hepatitis C virus into 7 genotypes and 67 subtypes: updated criteria and genotype assignment web resource Hepatology 59 1 2014 318 327 10.1002/hep.26744 24115039
Sorbo M.C. Cento V. Di Maio V.C. Howe A.Y.M. Garcia F. Perno C.F. Ceccherini-Silberstein F. Hepatitis C virus drug resistance associated substitutions and their clinical relevance: update 2018 Drug Resist. Updates 37 2018 17 39 10.1016/j.drup.2018.01.004
Spira S. Wainberg M.A. Loemba H. Turner D. Brenner B.G. Impact of clade diversity on HIV-1 virulence, antiretroviral drug sensitivity and drug resistance J. Antimicrob. Chemother. 51 2 2003 229 240 10.1093/jac/dkg079 12562686
Steinkühler C. Biasiol G. Brunetti M. Urbani A. Koch U. Cortese R. Pessi A. De Francesco R. Product inhibition of the hepatitis C virus NS3 protease Biochemistry 37 25 1998 8899 8905 10.1021/bi980313v 9636031
Subeha M.R. Telleria C.M. The anti-cancer properties of the HIV protease inhibitor nelfinavir Cancers 12 11 2020 10.3390/cancers12113437
Sudo S. Haraguchi H. Hirai Y. Gatanaga H. Sakuragi J.-i. Momose F. Morikawa Y. Efavirenz enhances HIV-1 gag processing at the plasma membrane through gag-pol dimerization J. Virol. 87 6 2013 3348 3360 10.1128/JVI.02306-12 23302874
Summa V. Ludmerer S.W. McCauley J.A. Fandozzi C. Burlein C. Claudio G. Coleman P.J. Dimuzio J.M. Ferrara M. Di Filippo M. Gates A.T. Graham D.J. Harper S. Hazuda D.J. Huang Q. McHale C. Monteagudo E. Pucci V. Rowley M. Rudd M.T. Soriano A. Stahlhut M.W. Vacca J.P. Olsen D.B. Liverton N.J. Carroll S.S. MK-5172, a selective inhibitor of hepatitis C virus NS3/4a protease with broad activity across genotypes and resistant variants Antimicrob. Agents Chemother. 56 8 2012 4161 4167 10.1128/aac.00324-12 22615282
Sun L. Niu L. Zhu X. Hao J. Wang P. Wang H. Antitumour effects of a protease inhibitor, nelfinavir, in hepatocellular carcinoma cancer cells J. Chemother. 24 3 2012 161 166 10.1179/1973947812y.0000000011 22759761
Sutton S.S. Magagnoli J. Cummings T. Hardin J. Association between the use of antibiotics, antivirals, and hospitalizations among patients with laboratory-confirmed influenza Clin. Infect. Dis. 72 4 2021 566 573 10.1093/cid/ciaa074 31974543
Svicher V. Ceccherini-Silberstein F. Erba F. Santoro M. Gori C. Bellocchi M.C. Giannella S. Trotta M.P. Monforte A. Antinori A. Perno C.F. Novel human immunodeficiency virus type 1 protease mutations potentially involved in resistance to protease inhibitors Antimicrob. Agents Chemother. 49 5 2005 2015 2025 10.1128/aac.49.5.2015-2025.2005 15855527
Tabler C.O. Wegman S.J. Chen J. Shroff H. Alhusaini N. Tilton J.C. The HIV-1 viral protease is activated during assembly and budding prior to particle release J. Virol. 96 9 2022 e0219821 10.1128/jvi.02198-21
Tanramluk D. Pakotiprapha D. Phoochaijaroen S. Chantravisut P. Thampradid S. Vanichtanankul J. Narupiyakul L. Akavipat R. Yuvaniyama J. MANORAA: a machine learning platform to guide protein-ligand design by anchors and influential distances Structure 30 1 2022 181 189 10.1016/j.str.2021.09.004 e185 34614393
Taylor J.G. Zipfel S. Ramey K. Vivian R. Schrier A. Karki K.K. Katana A. Kato D. Kobayashi T. Martinez R. Sangi M. Siegel D. Tran C.V. Yang Z.-Y. Zablocki J. Yang C.Y. Wang Y. Wang K. Chan K. Barauskas O. Cheng G. Jin D. Schultz B.E. Appleby T. Villaseñor A.G. Link J.O. Discovery of the pan-genotypic hepatitis C virus NS3/4A protease inhibitor voxilaprevir (GS-9857): a component of Vosevi Bioorg. Med. Chem. Lett 29 16 2019 2428 2436 10.1016/j.bmcl.2019.03.037 31133531
Thaisrivongs S. Strohbach J.W. Structure-based discovery of Tipranavir disodium (PNU-140690E): a potent, orally bioavailable, nonpeptidic HIV protease inhibitor Biopolymers 51 1 1999 51 58 10.1002/(sici)1097-0282(1999)51:1<51::Aid-bip6>3.0.Co;2-u 10380352
Thaisrivongs S. Skulnick H.I. Turner S.R. Strohbach J.W. Tommasi R.A. Johnson P.D. Aristoff P.A. Judge T.M. Gammill R.B. Morris J.K. Romines K.R. Chrusciel R.A. Hinshaw R.R. Chong K.-T. Tarpley W.G. Poppe S.M. Slade D.E. Lynn J.C. Horng M.-M. Tomich P.K. Seest E.P. Dolak L.A. Howe W.J. Howard G.M. Schwende F.J. Toth L.N. Padbury G.E. Wilson G.J. Shiou L. Zipp G.L. Wilkinson K.F. Rush B.D. Ruwart M.J. Koeplinger K.A. Zhao Z. Cole S. Zaya R.M. Kakuk T.J. Janakiraman M.N. Watenpaugh K.D. Structure-based design of HIV protease inhibitors: sulfonamide-containing 5,6-Dihydro-4-hydroxy-2-pyrones as non-peptidic inhibitors J. Med. Chem. 39 22 1996 4349 4353 10.1021/jm960541s 8893827
Thompson W.J. Ghosh A.K. Holloway M.K. Lee H.Y. Munson P.M. Schwering J.E. Wai J. Darke P.L. Zugay J. Emini E.A. Schleif W.A. Huff J.R. Anderson P.S. 3'-Tetrahydrofuranylglycine as a novel, unnatural amino acid surrogate for asparagine in the design of inhibitors of the HIV protease J. Am. Chem. Soc. 115 2 1993 801 803 10.1021/ja00055a069 30364398
Todosmedical https://todosmedical.com/tollovir
Tomei L. Failla C. Vitale R.L. Bianchi E. De Francesco R. A central hydrophobic domain of the hepatitis C virus NS4A protein is necessary and sufficient for the activation of the NS3 protease J. Gen. Virol. 77 Pt 5 1996 1065 1070 10.1099/0022-1317-77-5-1065 8609472
Tong L. Qian C. Massariol M.J. Bonneau P.R. Cordingley M.G. Lagacé L. A new serine-protease fold revealed by the crystal structure of human cytomegalovirus protease Nature 383 6597 1996 272 275 10.1038/383272a0 8805706
Tong L. Qian C. Massariol M.J. Déziel R. Yoakim C. Lagacé L. Conserved mode of peptidomimetic inhibition and substrate recognition of human cytomegalovirus protease Nat. Struct. Biol. 5 9 1998 819 826 10.1038/1860 9731777
Tong X. Chase R. Skelton A. Chen T. Wright-Minogue J. Malcolm B.A. Identification and analysis of fitness of resistance mutations against the HCV protease inhibitor SCH 503034 Antivir. Res. 70 2 2006 28 38 10.1016/j.antiviral.2005.12.003 16448708
Tong X. Arasappan A. Bennett F. Chase R. Feld B. Guo Z. Hart A. Madison V. Malcolm B. Pichardo J. Prongay A. Ralston R. Skelton A. Xia E. Zhang R. Njoroge F.G. Preclinical characterization of the antiviral activity of SCH 900518 (narlaprevir), a novel mechanism-based inhibitor of hepatitis C virus NS3 protease Antimicrob. Agents Chemother. 54 6 2010 2365 2370 10.1128/aac.00135-10 20308381
Toots M. Yoon J.J. Cox R.M. Hart M. Sticher Z.M. Makhsous N. Plesker R. Barrena A.H. Reddy P.G. Mitchell D.G. Shean R.C. Bluemling G.R. Kolykhalov A.A. Greninger A.L. Natchus M.G. Painter G.R. Plemper R.K. Characterization of orally efficacious influenza drug with high resistance barrier in ferrets and human airway epithelia Sci. Transl. Med. 11 515 2019 10.1126/scitranslmed.aax5866
Toots M. Yoon J.J. Hart M. Natchus M.G. Painter G.R. Plemper R.K. Quantitative efficacy paradigms of the influenza clinical drug candidate EIDD-2801 in the ferret model Transl. Res. 218 2020 16 28 10.1016/j.trsl.2019.12.002 31945316
Traynor K. Two drugs approved for chronic hepatitis C infection Am. J. Health Syst. Pharm. 68 13 2011 1176 10.2146/news110042 21690418
Tremblay C.L. Combating HIV resistance - focus on darunavir Therapeut. Clin. Risk Manag. 4 4 2008 759 766 10.2147/tcrm.s1709
Trinité B. Zhang H. Levy D.N. NNRTI-induced HIV-1 protease-mediated cytotoxicity induces rapid death of CD4 T cells during productive infection and latency reversal Retrovirology 16 1 2019 17 10.1186/s12977-019-0479-9 31242909
Trovato M. Sartorius R. D'Apice L. Manco R. De Berardinis P. Viral emerging diseases: challenges in developing vaccination strategies Front. Immunol. 11 2020 2130 10.3389/fimmu.2020.02130 33013898
Tseng A. Seet J. Phillips E.J. The evolution of three decades of antiretroviral therapy: challenges, triumphs and the promise of the future Br. J. Clin. Pharmacol. 79 2 2015 182 194 10.1111/bcp.12403 24730660
Tsu B.V. Fay E.J. Nguyen K.T. Corley M.R. Hosuru B. Dominguez V.A. Daugherty M.D. Running with scissors: evolutionary conflicts between viral proteases and the host immune system Front. Immunol. 12 2021 769543 10.3389/fimmu.2021.769543
Unoh Y. Uehara S. Nakahara K. Nobori H. Yamatsu Y. Yamamoto S. Maruyama Y. Taoda Y. Kasamatsu K. Suto T. Kouki K. Nakahashi A. Kawashima S. Sanaki T. Toba S. Uemura K. Mizutare T. Ando S. Sasaki M. Orba Y. Sawa H. Sato A. Sato T. Kato T. Tachibana Y. Discovery of S-217622, a noncovalent oral SARS-CoV-2 3CL protease inhibitor clinical candidate for treating COVID-19 J. Med. Chem. 65 9 2022 6499 6512 10.1021/acs.jmedchem.2c00117 35352927
Utkina-Sosunova I.V. Niatsetskaya Z.V. Sosunov S.A. Ratner V.I. Matsiukevich D. Ten V.S. Nelfinavir inhibits intra-mitochondrial calcium influx and protects brain against hypoxic-ischemic injury in neonatal mice PLoS One 8 4 2013 e62448 10.1371/journal.pone.0062448
V'Kovski P. Kratzel A. Steiner S. Stalder H. Thiel V. Coronavirus biology and replication: implications for SARS-CoV-2 Nat. Rev. Microbiol. 19 3 2021 155 170 10.1038/s41579-020-00468-6 33116300
Vacca J.P. Guare J.P. deSolms S.J. Sanders W.M. Giuliani E.A. Young S.D. Darke P.L. Zugay J. Sigal I.S. Schleif W.A. L-687,908, a potent hydroxyethylene-containing HIV protease inhibitor J. Med. Chem. 34 3 1991 1225 1228 10.1021/jm00107a050 2002465
Vacca J.P. Dorsey B.D. Schleif W.A. Levin R.B. McDaniel S.L. Darke P.L. Zugay J. Quintero J.C. Blahy O.M. Roth E. L-735,524: an orally bioavailable human immunodeficiency virus type 1 protease inhibitor Proc. Natl. Acad. Sci. U. S. A. 91 9 1994 4096 4100 10.1073/pnas.91.9.4096 8171040
van Boheemen S. de Graaf M. Lauber C. Bestebroer T.M. Raj V.S. Zaki A.M. Osterhaus A.D. Haagmans B.L. Gorbalenya A.E. Snijder E.J. Fouchier R.A. Genomic characterization of a newly discovered coronavirus associated with acute respiratory distress syndrome in humans mBio 3 6 2012 10.1128/mBio.00473-12
van Heeswijk R.P. Veldkamp A. Mulder J.W. Meenhorst P.L. Lange J.M. Beijnen J.H. Hoetelmans R.M. Combination of protease inhibitors for the treatment of HIV-1-infected patients: a review of pharmacokinetics and clinical experience Antivir. Ther. 6 4 2001 201 229 11878403
Vangeel L. Chiu W. De Jonghe S. Maes P. Slechten B. Raymenants J. André E. Leyssen P. Neyts J. Jochmans D. Remdesivir, Molnupiravir and Nirmatrelvir remain active against SARS-CoV-2 Omicron and other variants of concern Antivir. Res. 198 2022 105252 10.1016/j.antiviral.2022.105252
Venkatraman S. Discovery of boceprevir, a direct-acting NS3/4A protease inhibitor for treatment of chronic hepatitis C infections Trends Pharmacol. Sci. 33 5 2012 289 294 10.1016/j.tips.2012.03.012 22521415
Venkatraman S. Bogen S.L. Arasappan A. Bennett F. Chen K. Jao E. Liu Y.T. Lovey R. Hendrata S. Huang Y. Pan W. Parekh T. Pinto P. Popov V. Pike R. Ruan S. Santhanam B. Vibulbhan B. Wu W. Yang W. Kong J. Liang X. Wong J. Liu R. Butkiewicz N. Chase R. Hart A. Agrawal S. Ingravallo P. Pichardo J. Kong R. Baroudy B. Malcolm B. Guo Z. Prongay A. Madison V. Broske L. Cui X. Cheng K.C. Hsieh Y. Brisson J.M. Prelusky D. Korfmacher W. White R. Bogdanowich-Knipp S. Pavlovsky A. Bradley P. Saksena A.K. Ganguly A. Piwinski J. Girijavallabhan V. Njoroge F.G. Discovery of (1R,5S)-N-[3-amino-1-(cyclobutylmethyl)-2,3-dioxopropyl]- 3-[2(S)-[[[(1,1-dimethylethyl)amino]carbonyl]amino]-3,3-dimethyl-1-oxobutyl]- 6,6-dimethyl-3-azabicyclo[3.1.0]hexan-2(S)-carboxamide (SCH 503034), a selective, potent, orally bioavailable hepatitis C virus NS3 protease inhibitor: a potential therapeutic agent for the treatment of hepatitis C infection J. Med. Chem. 49 20 2006 6074 6086 10.1021/jm060325b 17004721
Vitoria M. Rangaraj A. Ford N. Doherty M. Current and future priorities for the development of optimal HIV drugs Curr. Opin. HIV AIDS 14 2 2019
Vogel M. Rockstroh J.K. Safety of lopinavir/ritonavir for the treatment of HIV-infection Expet Opin. Drug Saf. 4 3 2005 403 420 10.1517/14740338.4.3.403
von Hentig N. Atazanavir/ritonavir: a review of its use in HIV therapy Drugs Today 44 2 2008 103 132 10.1358/dot.2008.44.2.1137107
Wang Q. Gao H. Clark K.M. Mugisha C.S. Davis K. Tang J.P. Harlan G.H. DeSelm C.J. Presti R.M. Kutluay S.B. Shan L. CARD8 is an inflammasome sensor for HIV-1 protease activity Science 371 6535 2021 10.1126/science.abe1707
Ward A.R. Mota T.M. Jones R.B. Immunological approaches to HIV cure Semin. Immunol. 51 2021 101412 10.1016/j.smim.2020.101412
Warren T.K. Jordan R. Lo M.K. Ray A.S. Mackman R.L. Soloveva V. Siegel D. Perron M. Bannister R. Hui H.C. Larson N. Strickley R. Wells J. Stuthman K.S. Van Tongeren S.A. Garza N.L. Donnelly G. Shurtleff A.C. Retterer C.J. Gharaibeh D. Zamani R. Kenny T. Eaton B.P. Grimes E. Welch L.S. Gomba L. Wilhelmsen C.L. Nichols D.K. Nuss J.E. Nagle E.R. Kugelman J.R. Palacios G. Doerffler E. Neville S. Carra E. Clarke M.O. Zhang L. Lew W. Ross B. Wang Q. Chun K. Wolfe L. Babusis D. Park Y. Stray K.M. Trancheva I. Feng J.Y. Barauskas O. Xu Y. Wong P. Braun M.R. Flint M. McMullan L.K. Chen S.-S. Fearns R. Swaminathan S. Mayers D.L. Spiropoulou C.F. Lee W.A. Nichol S.T. Cihlar T. Bavari S. Therapeutic efficacy of the small molecule GS-5734 against Ebola virus in rhesus monkeys Nature 531 7594 2016 381 385 10.1038/nature17180 26934220
Weber J. Mesters J.R. Lepsík M. Prejdová J. Svec M. Sponarová J. Mlcochová P. Skalická K. Strísovský K. Uhlíková T. Soucek M. Machala L. Stanková M. Vondrásek J. Klimkait T. Kraeusslich H.G. Hilgenfeld R. Konvalinka J. Unusual binding mode of an HIV-1 protease inhibitor explains its potency against multi-drug-resistant virus strains J. Mol. Biol. 324 4 2002 739 754 10.1016/s0022-2836(02)01139-7 12460574
Weber I.T. Kneller D.W. Wong-Sam A. Highly resistant HIV-1 proteases and strategies for their inhibition Future Med. Chem. 7 8 2015 1023 1038 10.4155/fmc.15.44 26062399
Weichseldorfer M. Reitz M. Latinovic O.S. Past HIV-1 medications and the current status of combined antiretroviral therapy options for HIV-1 patients Pharmaceutics 13 11 2021 1798 10.3390/pharmaceutics13111798 34834213
Weikl T.R. Hemmateenejad B. Accessory mutations balance the marginal stability of the HIV-1 protease in drug resistance Proteins 88 3 2020 476 484 10.1002/prot.25826 31599014
Wen W. Chen C. Tang J. Wang C. Zhou M. Cheng Y. Zhou X. Wu Q. Zhang X. Feng Z. Wang M. Mao Q. Efficacy and safety of three new oral antiviral treatment (molnupiravir, fluvoxamine and Paxlovid) for COVID-19：a meta-analysis Ann. Med. 54 1 2022 516 523 10.1080/07853890.2022.2034936 35118917
Wensing A.M. van Maarseveen N.M. Nijhuis M. Fifteen years of HIV Protease Inhibitors: raising the barrier to resistance Antivir. Res. 85 1 2010 59 74 10.1016/j.antiviral.2009.10.003 19853627
WHO https://www.who.int/news-room/fact-sheets/detail/hepatitis-c
Williamson B.N. Feldmann F. Schwarz B. Meade-White K. Porter D.P. Schulz J. van Doremalen N. Leighton I. Yinda C.K. Pérez-Pérez L. Okumura A. Lovaglio J. Hanley P.W. Saturday G. Bosio C.M. Anzick S. Barbian K. Cihlar T. Martens C. Scott D.P. Munster V.J. de Wit E. Clinical benefit of remdesivir in rhesus macaques infected with SARS-CoV-2 Nature 585 7824 2020 273 276 10.1038/s41586-020-2423-5 32516797
Wire M.B. Shelton M.J. Studenberg S. Fosamprenavir : clinical pharmacokinetics and drug interactions of the amprenavir prodrug Clin. Pharmacokinet. 45 2 2006 137 168 10.2165/00003088-200645020-00002 16485915
Wutzler P. Thust R. Genetic risks of antiviral nucleoside analogues--a survey Antivir. Res. 49 2 2001 55 74 10.1016/s0166-3542(00)00139-x 11248359
Xia C. He Z. Liang S. Chen R. Xu W. Yang J. Xiao G. Jiang S. Metformin combined with nelfinavir induces SIRT3/mROS-dependent autophagy in human cervical cancer cells and xenograft in nude mice Eur. J. Pharmacol. 848 2019 62 69 10.1016/j.ejphar.2019.01.045 30695683
Yang K.S. Leeuwon S.Z. Xu S. Liu W.R. Evolutionary and structural insights about potential SARS-CoV-2 evasion of nirmatrelvir J. Med. Chem. 2022 10.1021/acs.jmedchem.2c00404
You D.M. Pockros P.J. Simeprevir for the treatment of chronic hepatitis C Expet Opin. Pharmacother. 14 18 2013 2581 2589 10.1517/14656566.2013.850074
Yousefi H. Mashouri L. Okpechi S.C. Alahari N. Alahari S.K. Repurposing existing drugs for the treatment of COVID-19/SARS-CoV-2 infection: a review describing drug mechanisms of action Biochem. Pharmacol. 183 2021 114296 10.1016/j.bcp.2020.114296
Zehbe I. Richard C. Lee K.F. Campbell M. Hampson L. Hampson I.N. Lopinavir shows greater specificity than zinc finger ejecting compounds as a potential treatment for human papillomavirus-related lesions Antivir. Res. 91 2 2011 161 166 10.1016/j.antiviral.2011.05.016 21669231
Zephyr J. Kurt Yilmaz N. Schiffer C.A. Viral proteases: structure, mechanism and inhibition Enzymes 50 2021 301 333 10.1016/bs.enz.2021.09.004 34861941
Zephyr J. Nageswara Rao D. Vo S.V. Henes M. Kosovrasti K. Matthew A.N. Hedger A.K. Timm J. Chan E.T. Ali A. Kurt Yilmaz N. Schiffer C.A. Deciphering the molecular mechanism of HCV protease inhibitor fluorination as a general approach to avoid drug resistance J. Mol. Biol. 434 9 2022 167503 10.1016/j.jmb.2022.167503
Zeuzem S. Ghalib R. Reddy K.R. Pockros P.J. Ben Ari Z. Zhao Y. Brown D.D. Wan S. DiNubile M.J. Nguyen B.Y. Robertson M.N. Wahl J. Barr E. Butterton J.R. Grazoprevir-elbasvir combination therapy for treatment-naive cirrhotic and noncirrhotic patients with chronic hepatitis C virus genotype 1, 4, or 6 infection: a randomized trial Ann. Intern. Med. 163 1 2015 1 13 10.7326/m15-0785 25909356
Zhang Y. Lee A.A. Bayesian semi-supervised learning for uncertainty-calibrated prediction of molecular properties and active learning Chem. Sci. 10 35 2019 8154 8163 10.1039/c9sc00616h 31857882
Zhang Y.M. Imamichi H. Imamichi T. Lane H.C. Falloon J. Vasudevachari M.B. Salzman N.P. Drug resistance during indinavir therapy is caused by mutations in the protease gene and in its Gag substrate cleavage sites J. Virol. 71 9 1997 6662 6670 10.1128/jvi.71.9.6662-6670.1997 9261388
Zhang L. Lin D. Sun X. Curth U. Drosten C. Sauerhering L. Becker S. Rox K. Hilgenfeld R. Crystal structure of SARS-CoV-2 main protease provides a basis for design of improved α-ketoamide inhibitors Science 368 6489 2020 409 412 10.1126/science.abb3405 32198291
Zhang T.-h. Dai L. Barton J.P. Du Y. Tan Y. Pang W. Chakraborty A.K. Lloyd-Smith J.O. Sun R. Predominance of positive epistasis among drug resistance-associated mutations in HIV-1 protease PLoS Genet. 16 10 2020 e1009009 10.1371/journal.pgen.1009009
Zhang X. Shang L. Fan G. Gu X. Xu J. Wang Y. Huang L. Cao B. The efficacy and safety of Janus kinase inhibitors for patients with COVID-19: a living systematic review and meta-analysis Front. Med. 8 2021 800492 10.3389/fmed.2021.800492
Zhang Y. Luo M. Wu P. Wu S. Lee T.Y. Bai C. Application of computational biology and artificial intelligence in drug design Int. J. Mol. Sci. 23 21 2022 10.3390/ijms232113568
Zhao Y. Fang C. Zhang Q. Zhang R. Zhao X. Duan Y. Wang H. Zhu Y. Feng L. Zhao J. Shao M. Yang X. Zhang L. Peng C. Yang K. Ma D. Rao Z. Yang H. Crystal Structure of SARS-CoV-2 Main Protease in Complex with Protease Inhibitor PF-07321332. Protein & Cell 2021 10.1007/s13238-021-00883-2
Zhou S. Hill C.S. Sarkar S. Tse L.V. Woodburn B.M.D. Schinazi R.F. Sheahan T.P. Baric R.S. Heise M.T. Swanstrom R. β-d-N4-hydroxycytidine inhibits SARS-CoV-2 through lethal mutagenesis but is also mutagenic to mammalian cells J. Infect. Dis. 224 3 2021 415 419 10.1093/infdis/jiab247 33961695
Zhou Y. Gammeltoft K.A. Ryberg L.A. Pham L.V. Fahnøe U. Binderup A. Hernandez C.R.D. Offersgaard A. Fernandez-Antunez C. Peters G.H.J. Ramirez S. Bukh J. Gottwein J.M. Nirmatrelvir Resistant SARS-CoV-2 Variants with High Fitness in Vitro 2022 bioRxiv 10.1101/2022.06.06.494921 2022.2006.2006.494921
| 36459838 | PMC9706241 | NO-CC CODE | 2022-12-02 23:17:25 | no | Mol Aspects Med. 2022 Dec 29; 88:101159 | utf-8 | Mol Aspects Med | 2,022 | 10.1016/j.mam.2022.101159 | oa_other |
==== Front
Am J Cardiol
Am J Cardiol
The American Journal of Cardiology
0002-9149
1879-1913
Elsevier Inc.
S0002-9149(22)01058-X
10.1016/j.amjcard.2022.09.030
Article
Mechanical Circulatory Support in Patients With COVID-19 Presenting With Myocardial Infarction
Guddeti Raviteja R. MD a
Sanina Cristina MD b
Jauhar Rajiv MD c
Henry Timothy D. MD d
Dehghani Payam MD e
Garberich Ross MS, MBA a
Schmidt Christian W. MS b
Nayak Keshav R. MD f
Shavadia Jay S. MD g
Bagai Akshay MD, MHS h
Alraies Chadi MD i
Mehra Aditya MD j
Bagur Rodrigo MD, PhD k
Grines Cindy MD l
Singh Avneet MD c
Patel Rajan A.G. MD m
Htun Wah Wah MD n
Ghasemzadeh Nima MD o
Davidson Laura MD p
Acharya Deepak MD q
Kabour Ameer MD r
Hafiz Abdul Moiz MD s
Amlani Shy MD t
Wasserman Hal S. MD u
Smith Timothy MD d
Kapur Navin K. MD v
Garcia Santiago MD d⁎
a Minneapolis Heart Institute Foundation at Abbott Northwestern Hospital, Minneapolis, Minnesota
b Montefiore Medical Center, Albert Einstein College of Medicine, Bronx, New York
c North Shore University Hospital, Manhasset, New York
d The Lindner Center for Research and Education, The Christ Hospital, Cincinnati, Ohio
e Prairie Vascular Research, Regina, Saskatchewan, Canada
f Department of Cardiology, Scripps Mercy Hospital, San Diego, California
g Royal University Hospital, University of Saskatchewan Saskatoon, Saskatchewan, Canada
h St Michael's Hospital, Toronto, Ontario, Canada
i DMC Harper University Hospital, Detroit, Michigan
j Jersey Shore University Medical Center, Neptune, New Jersey
k London Health Sciences Centre, London, Ontario, Canada
l Northside Hospital Cardiovascular Institute, Atlanta, Georgia
m Ochsner Health, University of Queensland Ochsner Clinical School, New Orleans, Louisiana
n Gunderson Health, Onalaska, Wisconsin
o Georgia Heart Institute, Gainesville, Georgia
p Northwestern University, Evanston, Illinois
q University of Arizona Sarver Heart Center, Tuczon, Arizona
r Mercy St Vincent's Medical Center, Toledo, Ohio
s Southern Illinois University School of Medicine. Springfiled, Illinois
t William Osler Health System, Ontario, Canada
u Nuvance Health, Danbury Hospital. Danbury, Connecticut
v Tufts Medical Center, Boston, Massachusetts
⁎ Corresponding author: Tel: 513-321-0875; fax: 513-585-1510.
29 11 2022
15 1 2023
29 11 2022
187 7683
17 5 2022
26 8 2022
© 2022 Elsevier Inc. All rights reserved.
2022
Elsevier Inc.
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
ST-segment elevation myocardial infarction (STEMI) complicating COVID-19 is associated with an increased risk of cardiogenic shock and mortality. However, little is known about the frequency of use and clinical impact of mechanical circulatory support (MCS) in these patients. We sought to define patterns of MCS utilization, patient characteristics, and outcomes in patients with COVID-19 with STEMI. The NACMI (North American COVID-19 Myocardial Infarction) is an ongoing prospective, observational registry of patients with COVID-19 positive (COVID-19+) with STEMI with a contemporary control group of persons under investigation who subsequently tested negative for COVID-19 (COVID-19−). We compared the baseline characteristics and in-hospital outcomes of COVID-19+ and patients with COVID-19− according to the use of MCS. The primary outcome was a composite of in-hospital mortality, stroke, recurrent MI, and repeat unplanned revascularization. A total of 1,379 patients (586 COVID-19+ and 793 COVID-19−) enrolled in the NACMI registry between January 2020 and November 2021 were included in this analysis; overall, MCS use was 12.3% (12.1% [n = 71] COVID-19+/MCS positive [MCS+] vs 12.4% [n = 98] COVID-19−/MCS+). Baseline characteristics were similar between the 2 groups. The use of percutaneous coronary intervention was similar between the groups (84% vs 78%; p = 0.404). Intra-aortic balloon pump was the most frequently used MCS device in both groups (53% in COVID-19+/MCS+ and 75% in COVID-19−/MCS+). The primary outcome was significantly higher in COVID-19+/MCS+ patients (60% vs 30%; p = 0.001) because of very high in-hospital mortality (59% vs 28%; p = 0.001). In conclusion, patients with COVID-19+ with STEMI requiring MCS have very high in-hospital mortality, likely related to the significantly higher pulmonary involvement compared with patients with COVID-19− with STEMI requiring MCS.
==== Body
pmcCOVID-19 continues to have a drastic impact on all aspects of acute myocardial infarction (MI) care.1 , 2 Myocardial injury is present in 8% to 62% of patients hospitalized with COVID-19 infection and is associated with poor clinical outcomes.3, 4, 5, 6 Patients with ST-segment elevation myocardial infarction (STEMI) and COVID-19 infection have unique demographic and adverse clinical characteristics, including frequent in-hospital presentation, extra-cardiac manifestations such as lung infiltrates, and cardiogenic shock.2 , 7 STEMI patients with COVID-19 represent a high-risk group with greater odds of in-hospital mortality, stroke, recurrent MI, and repeat unplanned revascularization than those without COVID-19.2 , 7 In addition, patients with COVID-19 and STEMI are less likely to receive invasive angiography and reperfusion therapy, all of which may contribute to increased adverse outcomes.2 Mechanical circulatory support (MCS) is being used with increasing frequency for the management of patients with acute MI and cardiogenic shock, and clinical outcomes of MCS use have been previously reported.8, 9, 10 However, little is known about the use of MCS in patients with COVID-19 presenting with STEMI. Although data on the use of veno-venous extracorporeal membrane oxygenation (VV-ECMO) in patients with COVID-19 with severe acute respiratory failure has been reported previously, no study to date has reported on the use of MCS for cardiogenic shock in these patients. In this study, we assessed patterns of MCS utilization in patients with COVID-19 presenting with STEMI and its association with in-hospital outcomes compared with patients who are COVID-19 negative (COVID-19−).
Methods
The NACMI (North American COVID-19 Myocardial Infarction) registry is a prospective, investigator-initiated, multicenter, observational registry of hospitalized patients with STEMI with COVID-19 infection (confirmed or suspected) created in collaboration with the Society for Cardiovascular Angiography and Interventions (SCAI) and the Canadian Association of Interventional Cardiology in conjunction with the American College of Cardiology Interventional Council.11 A detailed description of the study rationale and design has previously been published.11 Standardized data collection forms modeled after the American College of Cardiology National Cardiovascular Data Registry definitions were used with a secure web-based application (REDCap [Research Electronic Data Capture]) to manage the dataset.
In this sub-study, all patients with COVID-19+ enrolled in the NACMI registry between January 1, 2020, and November 22, 2021, were included and compared with persons under investigation for COVID-19 subsequently confirmed to be negative. Cardiogenic shock was defined as systolic blood pressure ≤90 mm Hg and cardiac index <2.2 L/min/m2 and/or need for vasopressors or inotropes for hemodynamic stabilization. The type of temporary MCS devices (intra-aortic balloon pump, Impella, tandem heart, or extracorporeal membrane oxygenation) used to treat hemodynamically unstable patients was left to the discretion of the operator performing the procedure. The study cohort was divided in 4 groups according to COVID-19 and MCS status (1) COVID-19 positive (COVID-19+)/MCS positive (MCS+), (2) COVID-19+/MCS negative (MCS−), (3) COVID-19−/MCS+, and (4) COVID-19−/MCS−. Baseline characteristics, clinical presentation, angiographic findings, and in-hospital outcomes were compared between the groups.
The primary outcome of interest was a composite of in-hospital death, stroke, recurrent MI, and unplanned revascularization. The secondary outcomes included individual components of the primary end point. The study protocol was approved by the Institutional Review Boards at each of the respective hospitals.
Statistical analysis was performed at the data coordinating center at the Minneapolis Heart Institute Foundation. Continuous variables are reported as mean ± SD if normally distributed or as median (interquartile range) if skewed and were analyzed using Student's t test or Wilcoxon rank-sum test depending on the distribution. Discrete variables are reported as counts and percentages and were analyzed using chi-square or Fisher's exact test, where appropriate. All analyses were performed using Stata version 15.1 (StataCorp, College Station, Texas).
Results
A total of 1,379 patients (COVID-19+/MCS+ [n = 71], COVID-19+/MCS− [n = 515], COVID-19−/MCS+ [n = 98] and COVID-19−/MCS− [n = 695]) were enrolled in the NACMI registry during the study period and were included in the present analysis. Patients with COVID-19+ (n = 586) commonly had minority ethnicity (51% [African-American 14%, Asian 7%, Hispanic 17%, Other 9%]), a high prevalence of diabetes mellitus (40%), and presented with respiratory symptoms such as dyspnea with infiltrates on chest-x-ray. In patients with COVID-19+, those who received MCS were predominantly Caucasian (58%). The baseline clinical characteristics are listed in Table 1 . Of the 1,379 patients with STEMI, MCS was used in 169 patients (12.3%) with no difference between patients with COVID-19+ (12.1%) and COVID-19− (12.4%) (Figure 1 ).Table 1 Baseline characteristics of the study cohort according to COVID-19 and MCS status
Table 1Baseline characteristics COVID+/MCS+ (n = 71) COVID-/MCS+ (n = 98) p Value COVID+/MCS-(n = 515) COVID-/MCS- (n = 695) p Value
Male, n (%) 56 (79%) 82 (84%) 0.4 375 (73%) 489 (70%) 0.3
Age group, n (%)
18–55
56–65
66–75
76–85
>85
22 (31%)
28 (39%)
14 (20%)
6 (8.5%)
1 (1.4%)
23 (23%)
35 (36%)
25 (26%)
14 (14%)
1 (1.0%) 0.6
127 (25%)
150 (29%)
127 (25%)
86 (17%)
24 (4.7%)
200 (29%)
198 (28%)
162 (23%)
93 (13%)
42 (6.0%) 0.3
Race, n (%)
Caucasian
African-American
Asian
Hispanic
Indigenous
Other
39 (58%)
7 (10%)
4 (6.0%)
14 (21%)
1 (1.5%)
2 (3.0%)
63 (67%)
11 (12%)
6 (6.4%)
8 (8.5%)
1 (1.1%)
5 (5.3%) 0.3
247 (50%)
76 (15%)
36 (7.2%)
88 (18%)
9 (1.8%)
42 (8.4%)
496 (75%)
60 (9.1%)
28 (4.2%)
46 (7.0%)
7 (1.1%)
23 (3.5%) <0.001
Non-Caucasian, n (%) 28 (42%) 31 (33%) 0.3 251 (50%) 164 (25%) <0.001
Weight, kg, mean±SD 89±25 85±23 0.6 87±24 87±23 0.6
BMI, mean±SD 28±10 27±12 0.8 28±10 26±11 0.11
CAD, n (%) 19 (29%) 34 (37%) 0.3 120 (26%) 173 (26%) 0.8
Prior PCI, n (%) 10 (16%) 28 (30%) 0.041 73 (16%) 113 (17%) 0.7
Prior MI, n (%) 10 (15%) 23 (25%) 0.15 64 (15%) 104 (16%) 0.5
Prior CABG, n (%) 2 (3.2%) 5 (5.4%) 0.7 27 (6.0%) 25 (3.8%) 0.093
Hypertension, n (%) 43 (62%) 74 (78%) 0.029 345 (70%) 481 (72%) 0.4
Dyslipidemia, n (%) 31 (47%) 55 (59%) 0.13 212 (45%) 382 (59%) <0.001
DM, n (%) 28 (43%) 39 (41%) 0.8 209 (44%) 206 (31%) <0.001
Previous stroke, n (%) 4 (6.3%) 14 (15%) 0.10 44 (9.6%) 64 (9.8%) >0.9
Smoking history, n (%)
Current
Former
Never
39 (66%)
11 (19%)
9 (15%)
46 (48%)
30 (32%)
19 (20%) 0.091
252 (53%)
87 (18%)
135 (28%)
255 (39%)
246 (38%)
152 (23%) <0.001
History of heart failure, n (%) 14 (22%) 14 (16%) 0.3 70 (15%) 65 (10%) 0.011
Family history of CAD, n (%) 10 (25%) 22 (35%) 0.3 85 (26%) 133 (29%) 0.5
Aspirin on admission, n (%) 21 (30%) 38 (39%) 0.2 194 (38%) 193 (28%) <0.001
Statin on admission, n (%) 18 (25%) 40 (41%) 0.037 193 (37%) 225 (32%) 0.065
CABG = coronary artery bypass grafting; CAD = coronary artery disease; CHF = congestive heart failure; COVID = coronavirus disease; DM = diabetes mellitus; MCS = mechanical circulatory support; MI = myocardial infarction; PCI = percutaneous coronary intervention; SD = standard deviation; TIA = transient ischemic attack.
Figure 1 Flow chart showing patient inclusion.
Figure 1
A summary of clinical and angiographic characteristics is listed in Table 2 . Patients with COVID-19+ more frequently presented with dyspnea and infiltrates on chest x-ray. Cardiogenic shock pre-percutaneous coronary intervention (pre-PCI) was present in 42.6% of patients requiring MCS compared with 7.7% in those not needing MCS (p <0.001). Similarly, 27.8% of patients in the MCS group had cardiac arrest pre-PCI compared with 8.4% in the non-MCS group (p <0.001). Although all patients in the COVID-19−/MCS+ group underwent coronary angiography, 3% of patients with COVID-19+/MCS+ with STEMI did not undergo angiography. In patients who underwent coronary angiography and PCI, median door-to-balloon times were similar between patients with COVID-19+/MCS+ and COVID-19−/MCS+, and overall PCI rates (primary and rescue) were not different between the 2 groups (77% vs 76.5%, p = 0.1). Coronary artery bypass grafting was performed more frequently in patients with COVID-19−/MCS+ than patients with COVID-19+/MCS+ (13% vs 3%, p = 0.031). The distribution of the culprit vessel was similar between patients with COVID-19+/MCS+ and COVID-19−/MCS+.Table 2 Clinical presentation and angiographic findings
Table 2 COVID+/MCS+ (n = 71) COVID-/MCS+ (n = 98) p Value COVID+/MCS- (n = 515) COVID-/MCS- (n = 695) p Value
Dyspnea 37 (52%) 32 (33%) 0.011 241 (47%) 223 (32%) <0.001
Chest pain 36 (51%) 72 (73%) 0.002 291 (57%) 566 (81%) <0.001
Syncope 7 (9.9%) 4 (4.1%) 0.2 15 (2.9%) 34 (4.9%) 0.084
Infiltrates 31 (44%) 22 (22%) 0.003 195 (38%) 84 (12%) <0.001
Pleural effusion 9 (13%) 13 (13%) >0.9 42 (8.2%) 41 (5.9%) 0.12
Cardiomegaly 6 (8.5%) 4 (4.1%) 0.3 45 (8.7%) 42 (6.0%) 0.073
Arrest pre-PCI 13 (21%) 34 (35%) 0.056 34 (7.6%) 68 (10%) 0.13
Shock pre-PCI 33 (52%) 39 (41%) 0.2 42 (9.7%) 51 (7.8%) 0.3
In-house presentation 6 (8.6%) 1 (1.0%) 0.021 33 (6.5%) 10 (1.5%) <0.001
Ejection fraction 30+/−13 32+/−14 0.4 46+/−13 45+/−12 0.3
D2B, median (IQR) 82 (56, 122) 77 (48, 137) 0.8 73 (50, 109) 73 (51, 102) 0.9
D2B (primary PCI only) 94 (59, 124) 78 (52, 120) 0.5 71 (48, 106) 73 (52, 100) 0.6
D2B <90 17 (49%) 33 (61%) 0.2 120 (64%) 317 (68%) 0.3
No angiography 2 (2.9%) 0 (0%) 0.2 68 (14%) 25 (3.7%) <0.001
Reperfusion strategy
Thrombolytics
Primary PCI
Facilitated/rescue PCI
Medical therapy
CABG
0 (0%)
53 (79%)
2 (3.0%)
10 (15%)
2 (3.0%)
0 (0%)
72 (76%)
3 (3.2%)
8 (8.4%)
12 (13%) 0.10
16 (3.8%)
278 (67%)
16 (3.8%)
102 (24%)
6 (1.4%)
5 (0.8%)
526 (80%)
18 (2.7%)
93 (14%)
15 (2.3%) <0.001
Any PCI 55 (82%) 75 (79%) 0.6 294 (70%) 544 (83%) <0.001
Normal coronaries (no culprit) 7 (10%) 4 (4.2%) 0.2 99 (24%) 74 (11%) <0.001
Culprit artery*
LMCA
LAD/diagonal
LCx/OM/PDA
RCA/PDA
Graft
Ramus
Multiple
No culprit
0 (0%)
29 (44%)
5 (7.6%)
10 (15%)
0 (0%)
0 (0%)
15 (23%)
7 (11%)
2 (2.1%)
33 (35%)
4 (4.3%)
15 (16%)
0 (0%)
1 (1.1%)
35 (37%)
4 (4.3%) 0.2
2 (0.5%)
107 (26%)
23 (5.6%)
113 (28%)
0 (0%)
1 (0.2%)
65 (16%)
99 (24%)
3 (0.5%)
217 (33%)
54 (8.2%)
211 (32%)
5 (0.8%)
3 (0.5%)
89 (14%)
74 (11%) <0.001
CABG = coronary artery bypass grafting; COVID = coronavirus disease; D2B = door-to-balloon; IQR = interquartile range; LAD = left anterior descending artery; LCx = left circumflex artery; LMCA = left main coronary artery; MCS = mechanical circulatory support; PCI = percutaneous coronary intervention; PDA = posterior descending artery; RCA = right coronary artery; SD = standard deviation.
⁎ In patients who underwent coronary angiography.
Right-sided cardiac catheterization was performed in about 39% of patients with COVID-19+/MCS+ compared with <25% in patients with COVID-19−/MCS+ (Table 3 ). All the patients that received MCS were classified as either SCAI class D shock (62% of patients with COVID-19+/MCS+ vs 51% of patients with COVID-19−/MCS+) or class E SCAI shock classification (38% in COVID-19+/MCS+ group vs 49% in COVID-19−/MCS+ group).Table 3 Hemodynamic data
Table 3 COVID+/MCS+ (n = 71) COVID-/MCS+ (n = 98) p Value COVID+/MCS- (n = 515) COVID-/MCS- (n = 695) p Value
LVEDP, mm Hg 26 (20, 36) 23 (15, 34) 0.4 17 (12, 24) 18 (12, 24) 0.4
Swan-Ganz inserted, n (%) 22 (39%) 17 (22%) 0.035 16 (5.0%) 18 (3.3%) 0.2
RA (mean), mm Hg 13 (10, 19) 15 (12, 19) 0.5 12.0 (7.8, 15.0) 15.0 (10.0, 20.5) 0.2
RV (systolic), mm Hg 39 (32, 55) 36 (30, 42) 0.6 41 (33, 47) 47 (37, 50) 0.3
Mean PAP mm Hg 30 (25, 33) 30 (22, 41) >0.9 28 (21, 35) 30 (28, 45) 0.12
Wedge (mean) mm Hg 22 (17, 34) 19 (16, 28) 0.5 15 (9, 24) 23 (17, 29) 0.051
Cardiac output, L/min 3.15 (2.35, 3.94) 3.90 (3.14, 6.16) 0.068 4.20 (3.49, 5.44) 4.09 (3.40, 5.79) 0.7
Cardiac index, L/min/m2 1.94 (1.59, 2.56) 1.90 (1.30, 2.62) 0.7 2.04 (1.75, 2.82) 2.25 (1.89, 3.12) 0.4
Intubated, n (%) 45 (74%) 50 (55%) 0.019 102 (23%) 79 (12%) <0.001
SCAI cardiogenic class, n (%)*
A
B
C
D
E
0 (0%)
0 (0%)
0 (0%)
40 (62%)
25 (38%)
0 (0%)
0 (0%)
0 (0%)
50 (51%)
48 (49%)
COVID = coronavirus disease; IQR = interquartile range; LAD = left anterior descending artery; LVEDP = left ventricular end-diastolic pressure; MCS = mechanical circulatory support; PAP = pulmonary artery pressure; RA = right atrium; RV = right ventricle; SCAI = Society of Cardiovascular Angiography and Intervention.
⁎ SCAI cardiogenic shock classification data were available for all patients in the COVID-19–/MCS+ group and 65 patients in the COVID-19+/MCS+ group.
Intra-aortic balloon pump (IABP) was the most common type of MCS used in both groups (74% and 62% in COVID-19−/MCS+ and COVID-19+/MCS+, respectively; Figure 2 ). Although Impella use was comparable between the 2 groups (21% in COVID-19−/MCS+ and 28% in COVID-19+/MCS+), the use of ECMO was numerically higher in the COVID-19+/MCS+ group (COVID-19+ 7% vs COVID-19− 3%, p = 0.11).Figure 2 Bar graph demonstrating frequency of use of various MCS devices stratified by COVID-19 status.
Figure 2
The primary outcome occurred in 58% of patients with COVID-19+/MCS+ and 28% of patients with COVID-19−/MCS+ (p <0.001). The difference was driven by higher in-hospital mortality (55% in the COVID-19+/MCS+ group vs 27% in the COVID-19−/MCS+ group; p = 0.001) with no difference in stroke, recurrent MI, or unplanned revascularization. Length of intensive care unit and total length of hospital stay were not significantly different between COVID-19+/MCS+ and COVID-19−/MCS+ groups, although significantly longer when compared with patients with COVID-19+/MCS− and COVID-19−/MCS− (Figure 3 ).Figure 3 In-hospital outcomes, including composite outcome and its individual components, total length of hospital stay, and length of ICU stay. ICU = intensive care unit.
Figure 3
Left ventricular ejection fraction was significantly lower in the COVID-19+/MCS+ group compared with patients with COVID-19+/MCS− (30 ± 13% vs 46 ± 13%, p <0.001; Supplementary Tables 1 and 2). Coronary angiography was not performed in 2.9% and 14% of patients with COVID-19+/MCS+ and COVID-19+/MCS−, respectively (p = 0.009). In contrast, PCI (both primary and rescue) was performed in 82% of patients in the COVID-19+/MCS+ group compared with 70.8% in the patients with COVID-19+/MCS− (p = 0.10). Left anterior descending coronary artery was the predominant culprit vessel in patients with COVID-19+/MCS+, whereas the right coronary artery was the major culprit vessel in patients with COVID-19+/MCS−.
The primary end point occurred in 25% of patients with COVID-19+/MCS− in comparison with 58% in patients with COVID-19+/MCS+ (p <0.001) (Figure 3, Supplementary Table 3).
Discussion
We used the NACMI registry to describe clinical and angiographic characteristics, patterns of MCS utilization, and in-hospital outcomes of patients with STEMI and concomitant COVID-19 infection. Several important findings are noted. First, MCS was used in 12.3% of patients with COVID-19+ presenting with STEMI, which is comparable with patients with STEMI without COVID-19. Second, in patients treated with MCS, ECMO use was significantly higher in patients with COVID-19+/MCS+ compared with COVID-19−/MCS+. Third, despite use of patients with MCS, COVID-19+ had significantly higher in-hospital mortality rates than patients with COVID-19−/MCS+. Fourth, when compared with a control group composed of patients with COVID-19 needing MCS devices, patients with COVID-19+ had a similar proportion of high-risk pre-PCI conditions (cardiogenic shock and cardiac arrest). Fifth, although most patients with STEMI and COVID-19+ infection in the NACMI registry were from ethnic minorities,2 those who received MCS were predominantly Caucasian. Patients with COVID-19+ had significantly higher rates of the composite primary end point driven primarily by very high rates of in-hospital mortality and abnormal lung findings (infiltrates), which suggest that extra-cardiac (i.e., pulmonary) involvement may account for these differential outcomes. Currently, there is no data on MCS use in patients with COVID-19 presenting with STEMI, and this is the first study to address this challenging yet crucial aspect of STEMI management in these high-risk patients.
COVID-19 has had a drastic impact on the overall management of STEMI. Globally, there was a decrease in the number of STEMI activations and primary PCI, an increase in door-to-balloon times, total ischemic times, and higher in-hospital mortality because of multiple reasons, including fear of contracting the disease in the hospital, delayed presentations, and a switch to pharmacological reperfusion.2 , 12, 13, 14
Cardiogenic shock complicates about 5% to 10% of STEMI cases and is associated with an in-hospital mortality rate of 23% if isolated, but as high as 44% when combined with pre-PCI cardiac arrest.15 The rationale for temporary MCS use in cardiogenic shock complicating acute MI includes maintaining organ perfusion, thereby preventing systemic shock syndrome, and reducing left ventricular filling pressures and wall stress, all of which reduce myocardial oxygen consumption, thereby limiting ischemia and infarct size.16 , 17 In patients presenting with acute MI and cardiogenic shock, MCS initiated before PCI, coupled with hemodynamic assessment, has shown to improve outcomes, including survival in observational studies, although randomized controlled trials employing this strategy are lacking.9 , 10 , 18 , 19
At the beginning of the pandemic, MCS, in the form of VV-ECMO, was primarily used in the setting of severe acute respiratory distress syndrome caused by COVID-19 infection.20 However, in patients with cardiovascular compromise, MCS may also be required for hemodynamic stabilization. Prepandemic data from the National Cardiovascular Disease Registry demonstrated that MCS devices were used in about 45% of patients presenting with cardiogenic shock and MI.21 , 22 Although the overall trend in the use of MCS remained stable over the past several years, IABP use has seen a significant decrease, whereas Impella use has more than doubled.21 In our study, most patients received IABP (62% and 74% in patients with COVID-19+ and COVID-19−, respectively). Not surprisingly, the frequency of ECMO use in patients with COVID-19+/MCS+ was numerically higher than that in the COVID-19−/MCS+ group (7% vs 3%), and most of them were VV-ECMO. Data from the National Cardiovascular Disease Registry CathPCI and Chest Pain-MI registries showed that between 2015 and 2017, the frequency of ECMO use in patients with acute MI complicated by cardiogenic shock who underwent PCI was about 2.6%.21
Patients who received MCS were predominantly Caucasian, representing 58% and 67% of COVID-19+ and COVID-19–, respectively, although 51% of patients in NACMI had minority ethnicity. On the contrary, data from the NACMI registry showed that patients with COVID-19 and STEMI were predominantly ethnic minorities.2 Previous reports have indicated that women and ethnic minorities are less likely to receive MCS and have high mortality.23 Differences in socioeconomic or education status and timely access to care are some of the factors associated with disparities in healthcare. It is currently unclear from our study whether Caucasian patients had more unstable presentations compared with other ethnicities or if there was any implicit bias toward the management approach in this high-risk patient population because of the previously mentioned factors. Future studies addressing this key question might shed some light on whether racial and ethnic disparities exist in COVID-19 and STEMI care.
The most striking finding of this study is the very high in-hospital mortality of patients with STEMI with COVID-19 infection who require MCS support, with nearly 2 of 3 patients dying in the hospital. In-hospital mortality rate in patients with STEMI with cardiogenic shock has trended downward over the years from 44.6% to 33.8%, which is partly attributed to early revascularization, early recognition of cardiac shock, and use of MCS devices.24 Despite similar rates of revascularization and hemodynamic profile, patients with COVID-19+/MCS+ with STEMI had significantly higher rates of in-hospital mortality of 55% compared with 27% in patients with COVID-19−/MCS+. It is likely that COVID-19 infection confers additional risk of mortality to patients with STEMI through concomitant pulmonary involvement, as evidenced by abnormal findings on chest x-rays and more mechanical ventilation in this group, and systemic hypercoagulable state.25, 26, 27
There are important limitations to our study. First, the NACMI registry is a prospective registry that includes only patients with STEMI. Patients with cardiogenic shock requiring MCS from etiologies other than STEMI are not represented in our study. Second, only in-hospital outcomes are reported in the present study. Long-term follow-up data on patients who survived to hospital discharge is ongoing. Third, although the type of MCS device was known, the decision to implant a particular device was left to individual operator discretion and not randomized. Therefore, we are unable to compare different MCS devices. The type of ECMO support was not recorded in the NACMI registry. In addition, the duration of MCS support in these patients was not available. Fourth, whereas a sizable number of patients in the COVID-19+/MCS+ and COVID-19−/MCS+ groups presented with cardiac arrest or had shock pre-PCI, we speculate hemodynamic compromise as a potential indication for MCS used in the rest of the patients and that decision was left to the operator discretion. Finally, we do not have data on the relation between the timing of initiation of MCS and PCI in our registry.
In conclusion, patients with STEMI with COVID-19 infection requiring MCS support have very high in-hospital mortality compared with patients with STEMI without COVID-19 requiring MCS.
Disclosures
The authors have no conflicts of interest to declare.
Appendix Supplementary materials
Image, application 1
Image, application 2
Image, application 3
This work was supported by the Society for Cardiovascular Angiography and Interventions (Washington, District of Columbia), American College of Cardiology Accreditation Grant (Washington, District of Columbia), Canadian Association of Interventional Cardiology (Ottawa, Ontario, Canada), Medtronic (Dublin, Ireland), and Abbott Vascular (Chicago, Illinois).
Supplementary material associated with this article can be found in the online version at https://doi.org/10.1016/j.amjcard.2022.09.030.
==== Refs
References
1 Garcia S Albaghdadi MS Meraj PM Schmidt C Garberich R Jaffer FA Dixon S Rade JJ Tannenbaum M Chambers J Huang PP Henry TD. Reduction in ST-segment elevation cardiac catheterization laboratory activations in the United States during COVID-19 pandemic J Am Coll Cardiol 75 2020 2871 2872 32283124
2 Garcia S Dehghani P Grines C Davidson L Nayak KR Saw J Waksman R Blair J Akshay B Garberich R Schmidt C Ly HQ Sharkey S Mercado N Alfonso CE Misumida N Acharya D Madan M Hafiz AM Javed N Shavadia J Stone J Alraies MC Htun W Downey W Bergmark BA Ebinger J Alyousef T Khalili H Hwang CW Purow J Llanos A McGrath B Tannenbaum M Resar J Bagur R Cox-Alomar P Stefanescu Schmidt AC Cilia LA Jaffer FA Gharacholou M Salinger M Case B Kabour A Dai X Elkhateeb O Kobayashi T Kim HH Roumia M Aguirre FV Rade J Chong AY Hall HM Amlani S Bagherli A Patel RAG Wood DA Welt FG Giri J Mahmud E Henry TD Society for Cardiac Angiography and Interventions, the Canadian Association of Interventional Cardiology, and the American College of Cardiology Interventional Council. Initial findings from the North American COVID-19 Myocardial Infarction registry J Am Coll Cardiol 77 2021 1994 2003 33888249
3 Franks CE Scott MG Farnsworth CW. Elevated cardiac troponin I is associated with poor outcomes in COVID-19 patients at an Academic Medical Center in Midwestern USA J Appl Lab Med 5 2020 1137 1139 32484893
4 Shi S Qin M Shen B Cai Y Liu T Yang F Gong W Liu X Liang J Zhao Q Huang H Yang B Huang C. Association of cardiac injury with mortality in hospitalized patients with COVID-19 in Wuhan, China JAMA Cardiol 5 2020 802 810 32211816
5 Guo T Fan Y Chen M Wu X Zhang L He T Wang H Wan J Wang X Lu Z. Cardiovascular implications of fatal outcomes of patients with coronavirus disease 2019 (COVID-19) JAMA Cardiol 5 2020 811 818 32219356
6 Metkus TS Sokoll LJ Barth AS Czarny MJ Hays AG Lowenstein CJ Michos ED Nolley EP Post WS Resar JR Thiemann DR Trost JC Hasan RK. Myocardial injury in severe COVID-19 compared with non-COVID-19 acute respiratory distress syndrome Circulation 143 2021 553 565 33186055
7 Kite TA Ludman PF Gale CP Wu J Caixeta A Mansourati J Sabate M Jimenez-Quevedo P Candilio L Sadeghipour P Iniesta AM Hoole SP Palmer N Ariza-Solé A Namitokov A Escutia-Cuevas HH Vincent F Tica O Ngunga M Meray I Morrow A Arefin MM Lindsay S Kazamel G Sharma V Saad A Sinagra G Sanchez FA Roik M Savonitto S Vavlukis M Sangaraju S Malik IS Kean S Curzen N Berry C Stone GW Gersh BJ Gershlick AH International COVID-ACS Registry Investigators. International prospective registry of acute coronary syndromes in patients with COVID-19 J Am Coll Cardiol 77 2021 2466 2476 34016259
8 Shah M Patnaik S Patel B Ram P Garg L Agarwal M Agrawal S Arora S Patel N Wald J Jorde UP. Trends in mechanical circulatory support use and hospital mortality among patients with acute myocardial infarction and non-infarction related cardiogenic shock in the United States Clin Res Cardiol 107 2018 287 303 29134345
9 Basir MB Kapur NK Patel K Salam MA Schreiber T Kaki A Hanson I Almany S Timmis S Dixon S Kolski B Todd J Senter S Marso S Lasorda D Wilkins C Lalonde T Attallah A Larkin T Dupont A Marshall J Patel N Overly T Green M Tehrani B Truesdell AG Sharma R Akhtar Y McRae T, 3rd O'Neill B Finley J Rahman A Foster M Askari R Goldsweig A Martin S Bharadwaj A Khuddus M Caputo C Korpas D Cawich I McAllister D Blank N Alraies MC Fisher R Khandelwal A Alaswad K Lemor A Johnson T Hacala M O'Neill WW National Cardiogenic Shock Initiative Investigators. Improved outcomes associated with the use of shock protocols: updates from the National cardiogenic shock initiative Catheter Cardiovasc Interv 93 2019 1173 1183 31025538
10 Basir MB Schreiber T Dixon S Alaswad K Patel K Almany S Khandelwal A Hanson I George A Ashbrook M Blank N Abdelsalam M Sareen N Timmis SBH O'Neill Md WW. Feasibility of early mechanical circulatory support in acute myocardial infarction complicated by cardiogenic shock: the Detroit cardiogenic shock initiative Catheter Cardiovasc Interv 91 2018 454 461 29266676
11 Dehghani P Davidson LJ Grines CL Nayak K Saw J Kaul P Bagai A Garberich R Schmidt C Ly HQ Giri J Meraj P Shah B Garcia S Sharkey S Wood DA Welt FG Mahmud E Henry TD North American COVID-19 ST-segment-elevation myocardial infarction (NACMI) registry: rationale, design, and implications Am Heart J 227 2020 11 18 32425198
12 Chew NW Ow ZGW Teo VXY Heng RRY Ng CH Lee CH Low AF Chan MY Yeo TC Tan HC Loh PH. The global impact of the COVID-19 pandemic on STEMI care: a systematic review and meta-analysis Can J Cardiol 37 2021 1450 1459 33848599
13 De Luca G Verdoia M Cercek M Jensen LO Vavlukis M Calmac L Johnson T Ferrer GR Ganyukov V Wojakowski W Kinnaird T van Birgelen C Cottin Y, A IJ Tuccillo B Versaci F Royaards KJ Berg JT Laine M Dirksen M Siviglia M Casella G Kala P Diez Gil JL Banning A Becerra V De Simone C Santucci A Carrillo X Scoccia A Amoroso G Lux A Kovarnik T Davlouros P Mehilli J Gabrielli G Rios XF Bakraceski N Levesque S Cirrincione G Guiducci V Kidawa M Spedicato L Marinucci L Ludman P Zilio F Galasso G Fabris E Menichelli M Garcia-Touchard A Manzo S Caiazzo G Moreu J Fores JS Donazzan L Vignali L Teles R Benit E Agostoni P Bosa Ojeda F Lehtola H Camacho-Freiere S Kraaijeveld A Antti Y Boccalatte M Deharo P Martinez-Luengas IL Scheller B Alexopoulos D Moreno R Kedhi E Uccello G Faurie B Gutierrez Barrios A Di Uccio FS Wilbert B Smits P Cortese G Parodi G Dudek D Impact of COVID-19 pandemic on mechanical reperfusion for patients with STEMI J Am Coll Cardiol 76 2020 2321 2330 33183506
14 Pessoa-Amorim G Camm CF Gajendragadkar P De Maria GL Arsac C Laroche C Zamorano JL Weidinger F Achenbach S Maggioni AP Gale CP Poppas A Casadei B. Admission of patients with STEMI since the outbreak of the COVID-19 pandemic: a survey by the European Society of Cardiology Eur Heart J Qual Care Clin Outcomes 6 2020 210 216 32467968
15 Omer MA Tyler JM Henry TD Garberich R Sharkey SW Schmidt CW Henry JT Eckman P Megaly M Brilakis ES Chavez I Burke N Gossl M Mooney M Sorajja P Traverse JH Wang Y Hryniewicz K Garcia S. Clinical characteristics and outcomes of STEMI patients with cardiogenic shock and cardiac arrest JACC Cardiovasc Interv 13 2020 1211 1219 32438992
16 Burkhoff D Sayer G Doshi D Uriel N. Hemodynamics of mechanical circulatory support J Am Coll Cardiol 66 2015 2663 2674 26670067
17 Rihal CS Naidu SS Givertz MM Szeto WY Burke JA Kapur NK Kern M Garratt KN Goldstein JA Dimas V Tu T Society for Cardiovascular Angiography and Interventions (SCAI), Heart Failure Society of America (HFSA), Society of Thoracic Surgeons (STS), American Heart Association (AHA), and American College of Cardiology (ACC). 2015 SCAI/ACC/HFSA/STS Clinical Expert Consensus Statement on the Use of Percutaneous Mechanical Circulatory Support Devices in Cardiovascular Care: Endorsed by the American Heart Assocation, the Cardiological Society of India, and Sociedad Latino Americana de Cardiologia Intervencion; Affirmation of Value by the Canadian Association of Interventional Cardiology-Association Canadienne de Cardiologie d'intervention J Am Coll Cardiol 65 2015 e7 e26 25861963
18 Henry TD Tomey MI Tamis-Holland JE Thiele H Rao SV Menon V Klein DG Naka Y Pina IL Kapur NK Dangas GD American Heart Association Interventional Cardiovascular Care Committee of the Council on Clinical Cardiology; Council on Arteriosclerosis, Thrombosis and Vascular Biology; and Council on Cardiovascular and Stroke Nursing. Invasive management of acute myocardial infarction complicated by cardiogenic shock: a scientific statement from the American Heart Association Circulation 143 2021 e815 e829 33657830
19 O'Neill WW Schreiber T Wohns DH Rihal C Naidu SS Civitello AB Dixon SR Massaro JM Maini B Ohman EM. The current use of Impella 2.5 in acute myocardial infarction complicated by cardiogenic shock: results from the USpella Registry J Interv Cardiol 27 2014 1 11 24329756
20 Barbaro RP MacLaren G Boonstra PS Iwashyna TJ Slutsky AS Fan E Bartlett RH Tonna JE Hyslop R Fanning JJ Rycus PT Hyer SJ Anders MM Agerstrand CL Hryniewicz K Diaz R Lorusso R Combes A Brodie D Extracorporeal Life Support Organization. Extracorporeal membrane oxygenation support in COVID-19: an international cohort study of the Extracorporeal Life Support Organization registry Lancet 396 2020 1071 1078 32987008
21 Dhruva SS Ross JS Mortazavi BJ Hurley NC Krumholz HM Curtis JP Berkowitz AP Masoudi FA Messenger JC Parzynski CS Ngufor CG Girotra S Amin AP Shah ND Desai NR. Use of mechanical circulatory support devices among patients with acute myocardial infarction complicated by cardiogenic shock JAMA Netw Open 4 2021 e2037748
22 Sandhu A McCoy LA Negi SI Hameed I Atri P Al'Aref SJ Curtis J McNulty E Anderson HV Shroff A Menegus M Swaminathan RV Gurm H Messenger J Wang T Bradley SM. Use of mechanical circulatory support in patients undergoing percutaneous coronary intervention: insights from the National Cardiovascular Data Registry Circulation 132 2015 1243 1251 26286905
23 Ya'qoub L Lemor A Dabbagh M O'Neill W Khandelwal A Martinez SC Ibrahim NE Grines C Voeltz M Basir MB. Racial, ethnic, and sex disparities in patients with STEMI and cardiogenic shock JACC Cardiovasc Interv 14 2021 653 660 33736772
24 Kolte D Khera S Aronow WS Mujib M Palaniswamy C Sule S Jain D Gotsis W Ahmed A Frishman WH Fonarow GC. Trends in incidence, management, and outcomes of cardiogenic shock complicating ST-elevation myocardial infarction in the United States J Am Heart Assoc 3 2014 e000590
25 Guzik TJ Mohiddin SA Dimarco A Patel V Savvatis K Marelli-Berg FM Madhur MS Tomaszewski M Maffia P D'Acquisto F Nicklin SA Marian AJ Nosalski R Murray EC Guzik B Berry C Touyz RM Kreutz R Wang DW Bhella D Sagliocco O Crea F Thomson EC McInnes IB. COVID-19 and the cardiovascular system: implications for risk assessment, diagnosis, and treatment options Cardiovasc Res 116 2020 1666 1687 32352535
26 Woolf SH Chapman DA Sabo RT Weinberger DM Hill L Taylor DDH. Excess deaths from COVID-19 and other causes, March–July 2020 JAMA 324 2020 1562 1564 33044483
27 Dehghani P Schmidt CW Garcia S Okeson B Grines CL Singh A Patel RAG Wiley J Htun WW Nayak KR Alraies MC Ghasemzadeh N Davidson LJ Acharya D Stone J Alyousef T Case BC Dai X Hafiz AM Madan M Jaffer FA Shavadia JS Garberich R Bagai A Singh J Aronow HD Mercado N Henry TD. North American COVID-19 Myocardial Infarction (NACMI) risk score for prediction of in-hospital mortality J Soc CardioVasc Angiogr Interv 1 2022 100404
| 36459751 | PMC9706494 | NO-CC CODE | 2022-12-01 23:20:15 | no | Am J Cardiol. 2023 Jan 15; 187:76-83 | utf-8 | Am J Cardiol | 2,022 | 10.1016/j.amjcard.2022.09.030 | oa_other |
==== Front
Am J Cardiol
Am J Cardiol
The American Journal of Cardiology
0002-9149
1879-1913
The Author(s). Published by Elsevier Inc.
S0002-9149(22)01133-X
10.1016/j.amjcard.2022.10.043
Article
Clinical Outcomes of Telehealth in Patients With Coronary Artery Disease and Heart Failure During the COVID-19 Pandemic
Woo Pauline MD a⁎
Chung Joanie MPH b
Shi Jiaxiao M. PhD b
Tovar Stephanie MS b
Lee Ming-Sum MD, PhD c
Adams Annette L. PhD b
a Department of Cardiology, Kaiser Permanente Baldwin Park Medical Center, Baldwin Park, California
b Department of Research & Evaluation, Kaiser Permanente Southern California, Pasadena, California
c Department of Cardiology, Kaiser Permanente Los Angeles Medical Center, Los Angeles, California
⁎ Corresponding author: Tel: 626-851-5986.
29 11 2022
15 1 2023
29 11 2022
187 171178
25 6 2022
21 9 2022
© 2022 The Author(s)
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
The COVID-19 pandemic necessitated a rapid adoption of telehealth (TH); however, its safety in subspecialty clinical practice remains uncertain. To assess the clinical outcomes associated with TH use in patients with coronary artery disease and/or heart failure during the initial phase of the COVID-19 pandemic, eligible adult patients who saw cardiologists from March 1, 2020, to August 31, 2020 (TH period) were identified. Patients were divided into two 3-month subcohorts (TH1, TH2) and compared with corresponding 2019 prepandemic subcohorts. The primary outcome was cardiovascular (CV) events within 3 months after index visits. Secondary analysis was CV events in patients aged ≥75 years within 3-month follow-up associated with TH use. Multivariable logistic regression was used to evaluate the association between TH use and CV outcomes. The study cohort included 6,485 TH and 7,557 prepandemic patients. The mean age was 70 years, with 40% of patients aged ≥75 years and 35% women. TH visits accounted for 0% of visits during the prepandemic period, compared with 68% during the TH period. Telephone visits comprised ≥92% of all TH encounters. Compared with the prepandemic period, patients seen during the TH period had fewer overall CV events (adjusted odds ratio 0.78, 95% confidence interval 0.67 to 0.90). Patients aged ≥75 years had similar findings (adjusted odds ratio 0.70, 95% confidence interval 0.55 to 0.89). Additional analysis of CV outcome events within 6 months after index visits showed similar findings. In conclusion, TH largely by way of telephone encounters can be safely incorporated into the ambulatory cardiology practice regardless of age.
==== Body
pmcThe COVID-19 pandemic has altered medical practice dramatically, forcing the adoption of telehealth (TH) at an unprecedented pace.1 Patients with cardiovascular (CV) disease are especially vulnerable to COVID-19 infections and have increased mortality risks, yet they are also at increased risk for CV events if routine clinical assessments are suboptimal.2 Early research suggested the potential benefits of telehealth in increasing care access, patient satisfaction, and cost savings,3 in addition to preserving personal protective equipment and mitigating infections.4, 5, 6 Despite new guidance from cardiology publications on how to perform TH visits,7, 8, 9 residual hesitancy among patients10 and providers11 in adopting TH stems from the paucity of evidence evaluating its impact on clinical outcomes compared with in-person visits. In this study, we sought to investigate the clinical outcomes of patients with coronary artery disease (CAD) and/or heart failure (HF) who were evaluated in cardiology clinics through TH visits during the first 6 months of the 2020 COVID-19 pandemic in comparison with prepandemic (PP) patients seen in the first 6 months of 2019. Our primary objective was to compare overall CV outcome events occurring within 3 months after patients’ index visits for the TH and PP cohorts. The secondary objective was to apply the same comparisons on patients with advanced age (≥75 years) to assess whether TH use adversely affects the clinical outcomes for this vulnerable subgroup. Additional analysis extending the follow-up period to 6 months after index visits was performed to assess the impact of TH care over a longer term.
Methods
This retrospective observational cohort study included adult (aged ≥18 years) patients with documented diagnoses of CAD and/or HF who were seen by cardiologists in clinics during the 6-month periods in March 2019 to August 2019 and March 2020 to August 2020 at 3 Kaiser Permanente Southern California (KPSC) medical centers which collectively served a patient population of approximately 889,000 members with median household income between $50,000 to $75,000. Eligible patients were identified from the KPSC region-wide electronic health record using the International Classification of Diseases, tenth edition (ICD-10) diagnosis codes. The index visit for each subject was defined as the first clinic encounter during the study periods. Patients were excluded from the study if they were seen solely for device checks, on hemodialysis, pregnant, enrolled in hospice or palliative care, terminated KPSC membership, or tested positive for COVID-19 within 6 months of their index clinic visits. Nonclinical telephone encounters lasting ≤10 minutes to discuss test results were also excluded.
Patients who had at least 1 index visit in the period from March 1 to August 31 in both the 2020 TH and 2019 PP periods were counted and compared for baseline characteristics such as age, gender, race, and body mass index (BMI). The overall co-morbid burden was quantified using the Charlson Co-morbidity Index (CCI) score12 with the associated chronic conditions identified in the 12 months before each patient's index visit. Co-morbid medical conditions known to be associated with CV events such as hypertension, diabetes mellitus, stroke (CV accident [CVA]), chronic kidney disease (CKD 3 to 5), anemia, pulmonary disease, and the presence of implantable cardioverter defibrillator (AICD) were also compared. In-person visit types included return and consultation encounters. TH visit types were telephone and video encounters which included both consultations and return visits. As the “stay at home” order was lifted in the latter part of the TH study period, the proportion of TH use also shifted, although TH visits still accounted for the majority of clinic encounters. To address the potential difference in outcomes because of this shift in TH use, our TH cohort in 2020 was divided into two 3-month subcohorts: TH1 (from March 1 to May 31) and TH2 (from June 1 to August 31) and compared with the two same 3-month subcohorts (PP1 and PP2) in 2019 when patients were seen exclusively in person.
CV outcome events were identified using the principal or primary discharge diagnoses with ICD-10 codes corresponding to CVA, transient ischemic attack, myocardial infarction, angina, HF exacerbations, or cardiac arrest. CV death was identified by ICD-10 codes in death registries and/or by chart review by a cardiologist (Supplementary Appendix 1). CV outcome events were determined by tracking emergency department (ED) visits, urgent care visits, hospitalizations, and CV deaths that occurred at least 24 hours after and within 3 and 6 months from index visits. Patients with more than 1 clinic visit spanning both 3-month study periods in 2019 or 2020 would have their qualifying outcome events count toward the first index visit. ED visits leading to CV hospitalization were counted as hospitalization encounters.
This study was approved by the KPSC Institutional Review Board and was internally funded by the KPSC Regional Research Committee. Informed consent was waived for this retrospective study utilizing existing clinical data.
Categorical variables were described using counts and proportions. Continuous variables were summarized using means and SDs. Comparison of the distributions of demographic, baseline clinical variables, and visit types between PP and TH cohorts were performed using chi-square for categorical variables and Kruskal–Wallis tests for continuous variables. Logistic regression analyses were performed to compare the outcome events between PP and TH cohorts. The multivariable models were adjusted for age, gender, race/ethnicity, BMI, and baseline medical co-morbidities. All analyses were two-sided and performed using SAS version 9.4 (Cary, North Carolina). A p <0.05 was considered statistically significant.
Results
A total of 8,816 patients in the 2020 TH periods and 8,490 patients in the 2019 PP periods were initially identified. After applying all exclusion criteria, 6,485 TH and 7,557 PP patient encounters were included in the final analysis (Figure 1 ). Since 875 patients in TH and 1,038 patients in PP cohorts were seen in both 3-month study periods, 5,610 patients in TH and 6,519 patients in PP periods were identified for baseline comparisons after eliminating duplicate encounters. Among the TH cohort, 2,659 patients (46%) were also seen in the PP period. There was no apparent increase in clinic encounters in TH2 after heavy TH use in the TH1 period as the proportion of patients who were seen in both 3-month study periods in PP and TH was identical at 16%. There was no apparent increase in total clinic encounters in TH periods, even after accounting for the 372 patients who were excluded because of COVID-19 infections.Figure 1 Cardiology patients and adverse cardiovascular clinical event identification flow diagram: prepandemic 2019 versus 2020 telehealth cohorts.
Figure 1
Baseline characteristics comparisons between the entire PP and TH cohorts showed similar distributions of age, gender, race, BMI, and CCI scores. The mean patient age at index visit was 70 years, with 40% of each cohort aged ≥75 years. The mean BMI was 30 kg/m2, 35% were women, and 64% had a CCI score of ≥3. Approximately 37% were Hispanic patients, while 17% and 15% were Asian and Black patients, respectively. Among the medical co-morbidities, patients with hypertension and diabetes had proportionally fewer encounters while patients with AICD and CVA had more clinic encounters in the TH than the PP periods. (Table 1 ). Baseline characteristics in the subcohort comparisons showed similar distributions of age, gender, race, BMI, CCI score ≥3, and encounter proportions for patients with hypertension, diabetes, and AICD (Table 2 ).Table 1 Baseline characteristics of cardiology patients seen in pre pandemic (2019) and telehealth (2020) periods
Table 1(No. of patients) 2019 (n = 6519) 2020 (n = 5610) p-Value
Hospital, n (%) 0.0212*
Baldwin Park 2095 (32.1%) 1775 (31.6%)
Downey 2461 (37.8%) 2019 (36.0%)
South Bay 1963 (30.1%) 1816 (32.4%)
Age mean (SD) 70.4 (12.0) 70.3 (12.5) 0.6335†
Age ≥75, n (%) 2597 (39.8%) 2226 (39.7%) 0.8591*
Gender, F n (%) 2304 (35.3%) 1966 (35.0%) 0.7316*
Race/Ethnicity, n (%)
White 1913 (29.3%) 1717 (30.6%) 0.4205*
Hispanic 2471 (37.9%) 2064 (36.8%)
Asian 1112 (17.1%) 920 (16.4%)
Black 930 (14.3%) 828 (14.8%)
Other/unknown 93 (1.4%) 81 (1.4%)
BMI
Mean (SD) 29.7 (6.4) 29.8 (6.7) 0.2387†
BMI ≥30, n (%) 2644 (40.6%) 2341 (41.8%) 0.1674*
Charlson co-morbidity score n (%)
0 369 (5.7%) 354 (6.3%) 0.2116*
1-2 2014 (30.9%) 1679 (29.9%)
3+ 4136 (63.4%) 3577 (63.8%)
Coronary artery disease, n (%) 4618 (70.8%) 3832 (68.3%) 0.0025*
Heart failure, n (%) 3274 (50.2%) 2989 (53.3%) 0.0008*
Anemia, n (%) 111 (1.7%) 111 (2.0%) 0.2584*
Chronic kidney disease 3-5, n (%) 1120 (17.2%) 936 (16.7%) 0.4679*
Stroke, n (%) 43 (0.7%) 58 (1.0%) 0.0237*
Diabetes, n (%) 2358 (36.2%) 1866 (33.3%) 0.0008*
Hypertension, n (%) 4465 (68.5%) 3484 (62.1%) <.0001*
Presence of AICD, n (%) 357 (5.5%) 401 (7.1%) 0.0001*
Pulmonary disease, n (%) 278 (4.3%) 277 (4.9%) 0.0769*
AICD= implantable cardioverter defibrillator; BMI = body mass index.
⁎ Chi-Square p-value.
† Kruskal–Wallis p-value.
Table 2 Baseline characteristics of cardiology subcohorts seen in prepandemic (2019) and telehealth (2020) periods
Table 2 PP 1 TH 1 PP 2 TH 2
(Patient Counts) (n = 3970) (n = 3508) p-Value* (n = 3587) (n = 2977) p-Value*
Hospital 0.004 <0.001
Baldwin Park 1244 (31.3%) 977 (27.9%) 1149 (32%) 1110 (37.3%)
Downey 1531 (38.6%) 1431 (40.8%) 1353 (37.7%) 948 (31.8%)
South Bay 1195 (30.1%) 1100 (31.4%) 1085 (30.2%) 919 (30.9%)
Age mean (SD) 70.4 (12.05) 70.1 (12.60) 0.45 70.6 (12.07) 70.4 (12.36) 0.63
Age ≥75 1576 (39.7%) 1380 (39.3%) 0.75 1446 (40.3%) 1197 (40.2%) 0.93
Gender F 1395 (35.1%) 1242 (35.4%) 0.81 1291 (36%) 1030 (34.6%) 0.24
Race/ethnicity 0.25 0.21
White 1151 (29%) 1053 (30%) 1046 (29.2%) 932 (31.3%)
Asian 664 (16.7%) 547 (15.6%) 634 (17.7%) 506 (17%)
Black 584 (14.7%) 565 (16.1%) 525 (14.6%) 392 (13.2%)
Hispanic 1512 (38.1%) 1290 (36.8%) 1340 (37.4%) 1107 (37.2%)
Other/unknown 59 (1.5%) 53 (1.5%) 42 (1.2%) 40 (1.3%)
BMI mean (SD) 29.7 (6.45) 30.0 (6.79) 0.29 29.6 (6.54) 29.7 (6.53) 0.95
BMI ≥30 1626 (41%) 1495 (42.7%) 0.13 1446 (40.3%) 1196 (40.2%) 0.93
Charlson co-morbidity score 0.32 0.26
0 194 (4.9%) 185 (5.3%) 203 (5.7%) 186 (6.2%)
1-2 1212 (30.5%) 1019 (29%) 1023 (28.5%) 887 (29.8%)
3+ 2564 (64.6%) 2304 (65.7%) 2361 (65.8%) 1904 (64%)
Coronary artery disease 2826 (71.2%) 2399 (68.4%) 0.009 2495 (69.6%) 2020 (67.9%) 0.14
Heart failure 2033 (51.2%) 1921 (54.8%) 0.002 1879 (52.4%) 1612 (54.1%) 0.15
Anemia 73 (1.8%) 66 (1.9%) 0.89 49 (1.4%) 67 (2.3%) 0.007
Chronic kidney disease 3-5 727 (18.3%) 548 (15.6%) 0.002 577 (16.1%) 555 (18.6%) 0.006
Stroke 25 (0.6%) 33 (0.9%) 0.13 27 (0.8%) 34 (1.1%) 0.10
Diabetes 1475 (37.2%) 1143 (32.6%) <0.001 1302 (36.3%) 1031 (34.6%) 0.16
Hypertension 2751 (69.3%) 2126 (60.6%) <0.001 2417 (67.4%) 1909 (64.1%) 0.006
Presence of AICD 217 (5.5%) 277 (7.9%) <0.001 209 (5.8%) 231 (7.8%) 0.002
Pulmonary disease 188 (4.7%) 176 (5%) 0.57 129 (3.6%) 147 (4.9%) 0.007
Index visit type (All) <0.001 <0.001
Consult in person 449 (11.3%) 109 (3.1%) 449 (12.5%) 195 (6.6%)
Return in person 3521 (88.7%) 695 (19.8%) 3137 (87.5%) 1060 (35.6%)
Telephone 0 (0%) 2670 (76.1%) 1 (0%) 1578 (53%)
Video 0 (0%) 34 (1%) 0 (0%) 144 (4.8%)
Index visit type Age≥75 <0.001 <0.001
Consult in person 139 (8.8%) 27 (2.0%) 143 (9.9%) 61 (5.1%)
Return in person 1437 (91.2%) 268 (19.4%) 1302 (90.0%) 435 (36.3%)
Telephone 0 1077 (78%) 1 (0.1%) 663 (55.4%)
Video 0 8 (0.6%) 0 38 (3.2%)
⁎ Chi-Square for categorical variables and Kruskal–Wallis for continuous variables.
AICD = implantable cardioverter defibrillator; BMI = body mass index; PP1 = prepandemic 1; PP2 = prepandemic 2; TH 1 = telehealth 1; TH 2 = telehealth 2.
Clinic encounters were exclusively in person in the PP periods. Consultations comprised 11% of all PP1 and 12% of all PP2 encounters, thus 89% of all PP1 and 88% of all PP2 encounters were return visits. Clinic encounters in TH periods were mostly done by way of telephone which comprised 76.1% of all TH1 and 53% of all TH2 encounters. Video visits comprised only 1% of all TH1 and 4.8% of all TH2 encounters. In-person consultations decreased in TH periods to 3.1% of all TH1 (a 73% decrease from PP1) and 6.6% of all TH2 encounters. In-person return visits also decreased to 19.8% of all TH1 (a 78% decrease from PP1) and 35.6% of all TH2 encounters (Table 2).
Patients aged ≥75 were seen in TH periods mostly by telephone which comprised 78.0% of all TH1 and 55.4% of all TH2 encounters. Video encounters comprised 0.6% of all TH1 and 3.2% of all TH2 encounters. A larger decrease in the proportions of in-person consultations and return visits were noted in TH1 for this group which comprised 2% (vs 8.8% in PP1) and 19.4% (vs 91.2% in PP1) of all encounters respectively. By the TH2 period, the proportional decrease in both in-person consultations and return visits for patients aged ≥75 was like the total cohort at 48% and 60% respectively when compared with the PP2 period (Table 2).
Outcome comparisons for all study cohorts showed no increase in overall CV outcome events including all ED and urgent care visits and hospitalizations between the TH and PP periods (part A of Table 3 ). After adjustments for age, gender, race/ethnicity, BMI, and medical co-morbid conditions, overall event rates were lower in TH1 [adjusted odds ratio [aOR] 0.75, 95% confidence interval [CI] 0.62 to 0.91] and TH2 (aOR 0.78, 95% CI 0.62 to 0.98) periods when compared with PP1 and PP2 subcohorts. CV hospitalization rates were significantly lower in TH1 (aOR 0.72, 95% CI 0.57 to 0.92) and trended lower in TH2 (aOR 0.77, 95% CI 0.58 to 1.01). ED visits trended lower in TH1 (aOR 0.73, 95% CI 0.52 to 1.02) and were significantly lower in TH2 (aOR 0.54, 95% CI 0.35 to 0.83). Urgent care visits were of very low volume and were not significantly different between the TH and PP subcohorts. Unexpected CV death counts in all 4 subcohorts were too few (≤5 each) to analyze reliably.Table 3 Outcome event rates within 3 months after index visits - 3-month subcohort comparisons
Table 3 A. All patients B. Patients aged ≥75 years
Event rate % (counts) Adjusted OR* (95%CI) Event rate % (counts) Adjusted OR* (95%CI)
PP 1 (3970) TH 1 (3508) PP 1 (1575) TH 1 (1380)
ED 2.37 (94) 1.65 (58) 0.73 (0.52-1.02) 2.92 (46) 1.81 (25) 0.65 (0.39-1.09)
Hospital 5.37 (213) 3.93 (138) 0.72 (0.57-0.92) 5.77 (91) 3.77 (52) 0.66 (0.45-0.95)
Urgent Care 0.76 (30) 0.68 (24) 0.86 (0.50-1.50) 0.51 (8) 0.51 (7) 0.96 (0.34-2.69)
Overall 8.49 (337) 6.27 (220) 0.75 (0.62-0.91) 9.2 (145) 6.09 (84) 0.69 (0.51-0.93)
PP 2 (3587) TH 2 (2977) PP 2 (1446) TH 2 (1197)
ED 2.09 (75) 1.11 (33) 0.54 (0.35-0.83) 2.28 (33) 1.25 (15) 0.53 (0.28-1.00)
Hospital 3.9 (140) 3.09 (92) 0.77 (0.58-1.01) 4.15 (60) 2.92 (35) 0.65 (0.41-1.02)
Urgent Care 0.36 (13) 0.64 (19) 1.86 (0.89-3.88) 0.14 (2) 0.50 (6) 3.28 (0.65-16.62)
Overall 6.36 (228) 4.84 (144) 0.78 (0.62-0.98) 6.57 (95) 4.68 (56) 0.67 (0.46-0.96)
BMI = body mass index; CI = confidence interval; ED = emergency visits; PP1 = prepandemic 1; PP2 = prepandemic 2; TH1 = telehealth 1; TH2 = telehealth 2.
⁎ Adjusted for age, gender, race, BMI, co-morbidities.
For the subgroup of patients aged ≥75 years, results mirrored those of the larger cohorts. There was no increase in CV event rates in both TH subcohorts when compared with the PP subcohorts (part B of Table 3). Overall outcome event rates were significantly lower for TH1 (aOR 0.69, 95% CI 0.51 to 0.93) and TH2 (aOR 0.67, 95% CI 0.46 to 0.96) when compared with PP1 and PP2 subcohorts. CV hospitalization rates were significantly lower in TH1 (aOR 0.66, 95% CI 0.45 to 0.95) and trended lower in TH2 (aOR 0.65, 95% CI 0.41 to 1.02). ED visits trended lower in TH1 (aOR 0.65, 95% CI 0.39 to 1.09) and in TH2 (aOR 0.53, 95% CI 0.28 to 1.0). Again, the volumes of urgent care visits and CV deaths were too low for meaningful comparisons.
The multivariable logistic regression model was used to further analyze how study cohorts, patient characteristics, and medical co-morbid conditions were associated with overall CV outcomes within 3 months after index visits. For all study patients, the TH cohort was less likely (aOR 0.78, 95% CI 0.67 to 0.90, p = 0.001) than the PP cohort to have outcome events (Figure 2 ). Black patients had a much higher likelihood of outcome events (aOR 1.54, 95% CI 1.23 to 1.92, p <0.001) while outcome events for Hispanic patients (aOR 1.20, 95% CI 0.99 to 1.44, p = 0.06) trended higher than White patients. Among the medical co-morbidities, patients with CKD 3 to 5 had a 26% higher likelihood of outcome events (aOR 1.26, 95% CI 1.04 to 1.54, p = 0.019). Applying the same analysis model to patients aged ≥75 years showed similar results (Figure 3 ). The TH cohort was less likely to have outcome events (aOR 0.70, 95% CI 0.55 to 0.89, p = 0.003), and Black patients had a higher outcome event likelihood (aOR 1.53, 95% CI 1.09 to 2.15, p = 0.013). Different than the larger cohort, Asian patients trended lower in likelihood for outcome events (aOR 0.66, 95% CI 0.44 to 1.0, p = 0.049) while patients with the pulmonary disease were associated with higher outcome event likelihood (aOR 1.55, 95% CI 1.02 to 2.35, p = 0.039).Figure 2 Likelihood of overall 3-month outcome event rates in association with all patient cohorts, characteristics, and medical co-morbidities.
Figure 2
Figure 3 Likelihood of overall 3-month outcome event rates in association with advanced age patient cohorts, characteristics, and medical co-morbidities.
Figure 3
Extending the analysis of CV outcome events to 6 months after index visits showed similar findings. The overall adverse event rates in both study periods roughly doubled the 3-month rates for all subcohorts with a similar decrease in overall CV adverse events, ED visits, and hospitalization rates for all TH subcohorts (Supplementary Table 1). Patients aged ≥75 years had no increase in event rates as well (Supplementary Table 1) when compared with the PP subcohorts. The lower likelihood for overall CV event rates was again seen for the entire TH cohort (aOR 0.77, 95% CI 0.68 to 0.87, p <0.0001) and for patients aged ≥75 years (aOR 0.74, 95% CI 0.62 to 0.89, p = 0.001) when compared with the PP cohort using the same multivariable logistic regression model for analysis.
Discussion
This study is novel in analyzing clinical outcomes of a high-risk cardiology patient group with CAD and/or HF after TH was adopted for a sustained period of 6 months during the early phase of the COVID-19 pandemic. Despite the continuous effort to promote video visits, TH visits were mostly conducted by telephone which comprised ≥92% of all TH encounters in this study. Overall CV outcome event counts from ED visits, urgent care visits, and hospitalizations were tracked for up to 6 months after patients’ index visits. The TH and PP cohorts had similar baseline patient characteristics, demographics, and co-morbidity burdens. Almost half (46%) of the TH patients were also seen in the PP period. Subcohort analyses were performed to address the potential outcome differences as the proportions of TH, and in-person visits shifted during the second half of our study period. The safety of TH use for advanced age (≥75 years) patients who were known to be challenged by TH visits13 was also studied.
Our results showed that TH care primarily by way of telephone encounters did not lead to any excess CV outcome events within 3 or 6 months after index visits, even when 77% of all encounters were conducted by way of TH visits during the first 3 months of the COVID-19 pandemic shutdown. No apparent increase in follow-up clinic encounters because of TH use was detected since the total clinic encounter counts were not increased in the TH period, and the proportion of patients (16%) who were seen in both 3-month subcohorts in the TH period was identical to that of the PP period. There was no outcome difference as the proportion of TH use shifted to 58% of all encounters in the second half of our TH study period. Those who were aged ≥75 years also had no excess outcome events utilizing TH largely by way of telephone contacts like the larger cohort. The favorable outcome in our TH cohorts that persisted over a 6-month study period from March 2020 and a follow-up period of up to 6 months could not be solely influenced by patients’ avoidance of medical care.5 The similar volume of patients who were seen in both TH and PP cohorts supports the potential benefit of TH in maintaining care access and reducing healthcare disparities for vulnerable patients as suggested by other investigators.14, 15, 16
As described in other studies, patients of Black and Hispanic ethinicity17 tend to have an increased likelihood of CV outcome events compared with White patients. Patients aged ≥75 years as a group were less willing to be seen in person during the early phase of the pandemic shutdown.18 Inequities in telemedicine use were a challenge to our cohorts who were older13 and had difficulties with video encounters during the TH periods.19 , 20 Lack of digital literacy and Internet access were potential factors limiting the adoption of video encounters to no more than 8% of our TH visits.21
This study is different from previous reports in several respects. We focused on clinical outcomes in a high-risk subspecialty patient group associated with TH use.22 Our patient population is drawn from a large, community-based integrated-care health system that serves a diverse population with an overall household median income of <$75,000, where Hispanic patients were the most prevalent ethnicity. Our non-fee-for-service reimbursement structure provided for a significantly larger percentage of telephone encounters being conducted than other studies, thus potentially narrowing the gap in TH inequity.13 This may be an important factor contributing to the favorable outcome seen in our older patient subgroup (≥75 years) who were able to access care by telephone despite the pandemic shutdown. By establishing a longer study and follow-up period of 6 months, the question of the clinical safety of TH can be better verified.19 , 20
The remaining challenge for TH is to devise evidence-based metrics to triage the appropriate patient demographics, disease diagnoses, and clinic visit types for TH care.23 As observed in our study, triage decisions for TH were subjective and likely based upon cardiologists’ clinical judgment that physical assessment was needed and/or patients’ desire to be seen in person. Evidence from this study showed that other patient characteristics such as Black ethnicity, diagnoses such as CKD 3 to 5 for the total cohort, and pulmonary disease among patients aged ≥75 years were associated with a higher likelihood for outcome events which can be factored into future triage decisions for TH use. Despite reimbursement policies favoring the use of video encounters,24 this digital platform could not be easily adopted by our older cardiology cohort. Further clinical outcome studies comparing the use of different TH modalities will be important since many vulnerable digitally challenged patient subgroups can potentially increase their access to care if telephone TH is economically incentivized by reimbursement policy.
This study has some potential limitations. We have excluded patients diagnosed with COVID-19 which comprised 4% of the TH cohort, and patients on hemodialysis because of a potential increase in CV outcomes associated with noncardiac factors. Owing to the significant surge in COVID-19 cases during our TH study periods, we did not compare the groups regarding all-cause mortality.25 Direct comparison of CV outcome among the different clinic and TH visit types was not performed because of the constraints of this observational study but necessitates further studies. Event outcome comparisons beyond 6 months after index visits were not analyzed because of the significant decrease in the proportion of TH visits and the increase in in-person clinic visits among our study cohort during the latter part of 2020 to 2021.
In conclusion, this study showed that incorporating TH largely by way of telephone encounter into the ambulatory cardiology clinical practice across all adult cardiology patients is clinically acceptable with no excess CV outcome events within 3 or 6 months after index visits. Since most cardiologists will still maintain TH as a fraction of their practice after COVID-1911 and the hybrid model of TH care as a supplement to in-person visits will persist,26 this outcome study will help to support the use of TH in the cardiology clinics. Outcome events in association with patient characteristics, demographics, and co-morbid diagnoses were presented to potentially provide evidence for future triage decisions for TH care.
Disclosures
The authors have no conflicts of interest to declare.
Appendix Supplementary materials
Image, application 1
Image, application 2
Acknowledgment
The authors would like to acknowledge the invaluable contributions of our colleagues who reviewed our study plan, provided suggestions to strengthen the study and the final manuscript, and Timothy M. Cotter, MD, CPC; Martha I Mercado, MD; Ramin F Shadman, MD; Maria T Taitano, MD; Cheng-Wei Huang, MD; and Alejandra E Montano, who provided other forms of support for this project.
This research was supported by a Grant KP-RRC-20201009 from the Regional Research Committee of Kaiser Permanente Southern California, Pasadena, California.
Supplementary material associated with this article can be found in the online version at https://doi.org/10.1016/j.amjcard.2022.10.043.
==== Refs
References
1 Wosik J Fudim M Cameron B Gellad ZF Cho A Phinney D Curtis S Roman M Poon EG Ferranti J Katz JN Tcheng J. Telehealth transformation: COVID-19 and the rise of virtual care J Am Med Inform Assoc 27 2020 957 962 32311034
2 Tessitore E Carballo D Poncet A Perrin N Follonier C Assouline B Carballo S Girardin F Mach F. Mortality and high risk of major adverse events in patients with COVID-19 and history of cardiovascular disease Open Heart 8 2021 e001526
3 Oseran AS Wasfy JH. Early experiences with cardiology electronic consults: a systematic review Am Heart J 215 2019 139 146 31325772
4 Mehrotra A, Ray K, Brockmeyer DM, Barnet ML, Bender JA. Rapidly converting to “virtual practices”: outpatient care in the era of Covid-19. NEJM Catalyst. Available at: https://catalyst.nejm.org/doi/full/10.1056/CAT.20.0091. Accessed on April 1, 2020.
5 Koonin LM Hoots B Tsang CA Leroy Z Farris K Jolly T Antall P McCabe B Zelis CBR Tong I Harris AM. Trends in the use of telehealth During the emergence of the COVID-19 pandemic- United States, January-March 2020 MMWR Morb Mortal Wkly Rep 69 2020 1595 1599 33119561
6 Hollander JE Carr BG. Virtually perfect? Telemedicine for COVID-19 N Engl J Med 382 2020 1679 1681 32160451
7 Gorodeski EZ Goyal P Cox ZL Thibodeau JT Reay RE Rasmusson K Rogers JG Starling RC. Virtual visits for care of patients with heart failure in the era of Covid-19: A Statement from the Heart Failure Society of America J Card Fail 26 2020 448 456 32315732
8 Kelly SA Schesing KB Thibodeau JT Ayers CR Drazner MH. Feasibility of remote video assessment of jugular venous pressure and implications for telehealth JAMA Cardiol 5 2020 1194 1195 32609293
9 Bhatt AB Nagale S. Telemedicine for cardiovascular disease care Bhatt AB Healthcare Information Technology for Cardiovascular Medicine. Health Informatics 2021 Springer Cham 1 11
10 Singh A Mountjoy N McElroy D Mittal S Al Hemyari B Coffey N Miller K Gaines K Patient perspectives with telehealth visits in cardiology During COVID-19: online patient survey study JMIR Cardio 5 2021 e25074 33385234
11 Lee M Luna P Lynch S Nagpal S Dominguez YC Ahmed Z Brice A Arham A Smolderen K Hurtado CM. Telehealth provider perspectives during COVID-19: insights from an academic cardiology practice J Am Coll Cardiol 77 2021 3212
12 Deyo RA Cherkin DC Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases J Clin Epidemiol 45 1992 613 619 1607900
13 Eberly LA Kallan MJ Julien HM Haynes N Khatana SA Nathan AS Snider C Chokshi NP Eneanya ND Takvorian SU Anastos-Wallen R.Chaiyachati Kri Ambrose M O'Quinn R Seigerman M Goldberg LR Leri D Choi K Gitelman Y Kolansky DM Cappola TP Ferrari VA Hanson CW Deleener ME Adusumalli S Patient Characteristics Associated with Telemedicine Access for Primary and Specialty Ambulatory Care During the COVID-19 Pandemic JAMA Netw Open 3 2020 e2031640
14 Kuehn BM. Telemedicine helps cardiologists extend their reach Circulation 134 2016 1189 1191 27754948
15 Poppas A Rumsfeld JS Wessler JD. Telehealth is having a moment: will it last? J Am Coll Cardiol 75 2020 2989 2991 32475633
16 Yuan N Pevnick JM Botting PG Elad Y Miller SJ Cheng S Ebinger JE. Patient use and clinical practice patterns of remote cardiology clinic visits in the era of COVID-19 JAMA Netw Open 4 2021 e214157
17 Lewey J Choudhry NK. The current state of ethnic and racial disparities in cardiovascular care: lessons from the past and opportunities for the future Curr Cardiol Rep 16 2014 530 25135343
18 Sammour Y Spertus JA Austin BA Magalski A Gupta SK Shatla I Dean E Kennedy KF Jones PG Nassif ME Main ML Sperry BW. Outpatient management of heart failure during the COVID-19 pandemic after adoption of a telehealth model JACC Heart Fail 9 2021 916 924 34857175
19 Sammour Y Shatla I Miller L Dean E Nassif M Gupta S Magalski A Main M Spertus J Sperry B. Comparison of video and telephone virtual visits in outpatients with heart failure Am J Cardiol 158 2021 153 156 34470705
20 Rowe SJ Paratz ED Fahy L Prior DL MacIsaac AI. Telehealth in Australian cardiology: insight into factors predicting the use of telephone versus video during the COVID-19 pandemic Intern Med J 51 2021 1229 1235 34227713
21 Boriani G Maisano A Bonini N Albini A Imberti JF Venturelli A Menozzi M Ziveri V Morgante V Camaioni G Passiatore M De Mitri G Nanni G Girolami D Fontanesi R Siena V Sgreccia D Malavasi VL Valenti AC Vitolo M Digital literacy as a potential barrier to implementation of cardiology tele-visits after COVID-19 pandemic: the INFO-COVID survey J Geriatr Cardiol 18 2021 739 747 34659380
22 Reed M Huang J Graetz I Muelly E Millman A Lee C. Treatment and follow-up care associated with patient-scheduled primary care telemedicine and in-person visits in a large integrated health system JAMA Netw Open 4 2021 e2132793
23 Reeves JJ Ayers JW Longhurst CA Telehealth in the COVID-19 era: a balancing act to avoid harm J Med Internet Res 23 2021 e24785 33477104
24 Medicare and Medicaid programs, basic health program, and exchanges; additional policy and regulatory revisions in response to the COVID-19 public health emergency and delay of certain reporting requirements for the skilled nursing facility quality reporting program. Federal Register. Available at: https://www.federalregister.gov/documents/2020/05/08/2020-09608/medicare-and-medicaid-programs-basic-health-program-and-exchanges-additional-policy-and-regulatory. Accessed on May 25, 2020.
25 Woolf SH Chapman DA Sabo RT Zimmerman EB. Excess deaths From COVID-19 and other causes in the US, March 1, 2020, to January 2, 2021 JAMA 325 2021 1786 1789 33797550
26 Hollander J, Sharma R. Telemedicine is mainstream care delivery.NEJM Catalyst Innovations in Care Delivery. Available at: https://catalyst.nejm.org/doi/full/10.1056/CAT.22.0176. Accessed on June 15, 2022.
| 36459742 | PMC9706495 | NO-CC CODE | 2022-12-02 23:17:35 | no | Am J Cardiol. 2023 Jan 15; 187:171-178 | utf-8 | Am J Cardiol | 2,022 | 10.1016/j.amjcard.2022.10.043 | oa_other |
==== Front
Sci Total Environ
Sci Total Environ
The Science of the Total Environment
0048-9697
1879-1026
Elsevier B.V.
S0048-9697(22)07647-1
10.1016/j.scitotenv.2022.160544
160544
Corrigendum
Corrigendum to “Delta SARS-CoV-2 variant is entirely substituted by the omicron variant during the fifth COVID-19 wave in Attica region” [Sci. Total Environ., 856(Pt 1) (2023)/159062]
Galani Aikaterini a
Markou Athina a
Dimitrakopoulos Lampros a
Kontou Aikaterini a
Kostakis Marios a
Kapes Vasileios a
Diamantopoulos Marios A. b
Adamopoulos Panagiotis G. b
Avgeris Margaritis cd
Lianidou Evi a
Scorilas Andreas b
Paraskevis Dimitrios e
Tsiodras Sotirios f
Dimopoulos Meletios-Athanasios g
Thomaidis Nikolaos a⁎
a Laboratory of Analytical Chemistry, Department of Chemistry, National and Kapodistrian University of Athens, University Campus, Zografou, 15771 Athens, Greece
b Department of Biochemistry and Molecular Biology, Faculty of Biology, National and Kapodistrian University of Athens, Athens, Greece
c Department of Biochemistry and Molecular Biology, Faculty of Biology, National and Kapodistrian University of Athens, Athens, Greece
d Laboratory of Clinical Biochemistry - Molecular Diagnostics, Second Department of Pediatrics, "P. & A. Kyriakou" Children's Hospital, School of Medicine, National and Kapodistrian University of Athens, Athens, Greece
e Department of Hygiene Epidemiology and Medical Statistics, School of Medicine, National and Kapodistrian University of Athens, 11527 Athens, Greece
f Fourth Department of Internal Medicine, School of Medicine, University General Hospital Attikon, National and Kapodistrian University of Athens, Greece
g Department of Clinical Therapeutics, School of Medicine, National and Kapodistrian University of Athens, Greece
⁎ Corresponding author at: Laboratory of Analytical Chemistry, Department of Chemistry, National and Kapodistrian University of Athens, Panepistimiopolis Zografou, 15771 Athens, Greece.
29 11 2022
29 11 2022
160544© 2022 Elsevier B.V. All rights reserved.
2022
Elsevier B.V.
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
==== Body
pmcThe authors state that the printed version of the above article missed the contribution of an author, which was that the third author had contributed to the writing of the original draft in addition to methodology. The correct and final version follows.
CRediT authorship contribution statement
Aikaterini Galani: Methodology, Validation, Writing – original draft. Athina Markou: Supervision, Writing – review & editing, Project administration. Lampros Dimitrakopoulos: Methodology, Writing – original draft. Aikaterini Kontou: Validation. Marios Kostakis: Validation. Vasileios Kapes: Methodology. Marios A. Diamantopoulos: Formal analysis, Software. Panagiotis G. Adamopoulos: Formal analysis. Margaritis Avgeris: Formal analysis, Writing– review & editing. Evi Lianidou: Writing – review & editing. Andreas Scorilas: Formal analysis. Dimitrios Paraskevis: Writing – review & editing. Sotirios Tsiodras: Writing – review & editing. Meletios-Athanasios Dimopoulos: Funding acquisition, Writing – review & editing. Nikolaos Thomaidis: Conceptualization, Project administration, Visualization, Resources.
| 36460542 | PMC9706548 | NO-CC CODE | 2022-12-01 23:19:46 | no | Sci Total Environ. 2022 Nov 29;:160544 | utf-8 | Sci Total Environ | 2,022 | 10.1016/j.scitotenv.2022.160544 | oa_other |
==== Front
J Clin Epidemiol
J Clin Epidemiol
Journal of Clinical Epidemiology
0895-4356
1878-5921
Elsevier Inc.
S0895-4356(22)00261-X
10.1016/j.jclinepi.2022.08.017
Letter to the Editor
Understanding how deferred consent affects patient characteristics and outcomes: an exploratory analysis of a clinical trial of prone positioning for COVID-19
Colacci Michael Physician
General Internal Medicine, Department of Medicine, University of Toronto, Toronto, Ontario, Canada
Raissi Afsaneh
Unity Health Toronto, Toronto, Ontario, Canada
Bhasin Ajay Physician
Division of Hospital Medicine, Department of Medicine, Northwestern University, Feinberg School of Medicine, Chicago, IL, USA
Branfield Day Leora Physician
General Internal Medicine, Department of Medicine, University of Toronto, Toronto, Ontario, Canada
Bregger Melissa Physician
Division of Hospital Medicine, Department of Medicine, Northwestern University, Feinberg School of Medicine, Chicago, IL, USA
Carpenter Travis Physician
Division of General Internal Medicine, St Joseph's Health Centre, Unity Health Toronto, Department of Medicine, University of Toronto, Toronto, Ontario, Canada
Castellucci Lana Associate Professor
Department of Medicine, Ottawa Hospital Research Institute at the University of Ottawa, Ottawa, Ontario, Canada
Cheung Angela M. Professor
Department of Medicine, University Health Network, University of Toronto, Toronto, Ontario, Canada
Dragoi Laura Physician
Interdepartmental Division of Critical Care, University of Toronto, Toronto, Ontario, Canada
Dunbar-Yaffe Richard Assistant Professor
Division of General Internal Medicine and Geriatrics, Sinai Health System and University Health Network, Toronto, Ontario, Canada
Fidler Lee Assistant Professor
Division of Respirology, Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada
Fowler Rob Professor
Interdepartmental Division of Critical Care Medicine, University Health Network, Toronto, Ontario, Canada
Gosset Alexi Medical Student
Department of Medicine, University of Toronto, Toronto, Ontario, Canada
Hensel Rachel Physician
Department of Medicine, Beth Israel Deaconess Medical Center, Boston, MA, USA
Herridge Margaret Professor
Interdepartmental Division of Critical Care Medicine, University of Toronto, Toronto, Ontario, Canada
Hussein Haseena Physician
Department of Medicine, William Osler Health System, Brampton, Ontario, Canada
Kapral Moira Professor
General Internal Medicine, Department of Medicine, University of Toronto, Toronto, Ontario, Canada
Department of Medicine, University Health Network, University of Toronto, Toronto, Ontario, Canada
Munshi Laveena Assistant Professor
Mount Sinai Hospital, Interdepartmental Division of Critical Care Medicine, Toronto, Ontario, Canada
Quinn Kieran Assistant Professor
Sinai Health System, Department of Medicine, University of Toronto, Toronto, Ontario, Canada
Razak Fahad Assistant Professor
Division of General Internal Medicine, St. Michael's Hospital, Unity Health Toronto, Toronto, Ontario, Canada
Roza da Costa Bruno Associate Professor
The Applied Health Research Centre (AHRC), Li Ka Shing Knowledge Institute of St. Michael's Hospital, Toronto, Ontario, Canada
Soong Christine Associate Professor
Divisions of General Internal Medicine and Hospital Medicine, Sinai Health, Toronto, Ontario, Canada
Tang Terence Lecturer
Institute for Better Health, Trillium Health Partners, Mississauga, Ontario, Canada
Venus Kevin Lecturer
Division of General Internal Medicine and Geriatrics, University Health Network, Toronto, Ontario, Canada
Verma Amol Assistant Professor
Division of General Internal Medicine, St. Michael's Hospital, Unity Health Toronto, Toronto, Ontario, Canada
Fralick Michael Assistant Professor ∗
Sinai Health System, Department of Medicine, University of Toronto, Toronto, Ontario, Canada
∗ Corresponding author. Sinai Health System, Department of Internal Medicine, 5th Floor Room L255, 60 Murray Street, Toronto, Ontario, Canada M5B 1W8. Tel.: +1 416 586 4800.
20 10 2022
20 10 2022
© 2022 Elsevier Inc. All rights reserved.
2022
Elsevier Inc.
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
==== Body
pmc What is new?
Key findings
• Differences in characteristics were found between patients in the deferred consent and nondeferred consent groups.
• Primarily it was found that the patients in the deferred consent group were more likely to be enrolled on weekends and were more likely to be older.
What this adds to what was known
• This finding shows that deferred consent, the process of enrolling patients in clinical trials prior to obtaining consent, may allow for the enrollment of a greater number of individuals.
What is the implication and what should change now?
• These findings may have implications for obtaining consent when conducting clinical trials in the future.
1 Introduction
Deferred consent, the process of enrolling patients in a clinical trial before consent is obtained, is often employed in studies of minimal risk [[1], [2], [3], [4]], or when patients are unable to consent at the time of study enrollment [[5], [6], [7], [8]]. Deferred consent has several possible benefits including decreasing the time to study enrollment, allowing for enrollment when study personnel are unavailable [9,10], and allowing the inclusion of populations less represented in clinical trials [4,11,12]. Multiple studies have identified that deferred consent is considered an acceptable substitute for pre-enrollment consent by both participants and clinicians [13,14]. Our objective was to assess patient-level characteristics, adherence, and rate of withdrawal among participants enrolled with consent obtained before (nondeferred consent) vs. after (deferred consent) randomization in the COVID-PRONE randomized trial (NCT04383613).
2 Methods
We conducted a secondary analysis of the COVID-prone international, pragmatic randomized clinical trial, which assessed prone positioning in patients with COVID-19 [15].
Deferred consent was allowed as the benefit of prone positioning was thought to be greatest when immediately implemented, because patients may be in respiratory distress at the time of enrollment, and because we expected potential harms from prone positioning to be minimal. Deferred consent was allowed at all except two hospitals. The decision on timing of consent was made by the site lead at the time of recruitment.
The primary outcome of this analysis was the difference in patient characteristics, time spent prone, and rate of withdrawal among participants who were enrolled with consent obtained before vs. after randomization. We did not compare the primary outcome between the two groups because our study was stopped for futility and our primary outcome was similar between patients randomized to prone positioning compared to standard of care.
3 Results
Among 248 total patients, 125 (50.5%) were enrolled after consent, and 123 (49.5%) were enrolled with deferred consent. The median time between randomization and consent was 1 day (interquartile range [IQR] 0–2 days) in the deferred consent group. Patients in the deferred consent group were more likely to be enrolled on weekends (14.6% vs. 9.6%), male (67.5% vs. 60.8%), and older (median age 60 vs. 54 years) Table 1 . The frequency of deferred consent varied significantly by hospital (median 41.5%; range: 0–100%). Patients in the deferred consent group were more likely to require oxygen via face mask (9.8% vs. 2.4%) and less likely to have received remdesivir (27.6% vs. 56%).Table 1 Baseline characteristics
Characteristic Deferred (N = 123) Nondeferred (N = 125)
COVID status
COVID-19 test result 120 (97.6%) 122 (97.6%)
Randomization timing
Number of days between admission and randomization 1 [1, 1] 1 [1, 1]
Days between randomization and consent 1 [0, 2] 0 [0, 0]
Randomized on a saturday or sunday 18 (14.6%) 12 (9.6%)
Age
Median [IQR] 60 [49.5, 68] 54 [39, 61]
<50 31 (25.2%) 50 (40%)
50–70 69 (56.1%) 64 (51.2%)
>70 23 (18.7%) 11 (8.8%)
Sex
Female 40 (32.5%) 49 (39.2%)
Date of randomization
Before Sept 1, 2020 6 (4.9%) 2 (1.6%)
Sept 1, 2020 to Feb 28, 2021 84 (68.3%) 60 (48%)
After Feb 28, 2021 33 (26.8%) 63 (50.4%)
Comorbid conditions
Diabetes 35 (28.5%) 32 (25.6%)
Hypertension 47 (38.2%) 51 (40.8%)
Current smoker 4 (3.3%) 3 (2.4%)
COPD or asthma 8 (6.5%) 19 (15.2%)
Heart failure 3 (2.4%) 3 (2.4%)
Illness severity
Lymphocyte count 0.82 [0.6, 1.1] 0.9 [0.68, 1.2]
Creatinine 81 [66, 100] 76 [63, 93]
Systolic blood pressure 123 [114, 133] 122.5 [115, 130]
Oxygen saturation 94 [93, 96] 94 [93, 95]
FiO2 32 [28, 36] 32 [28, 36]
S/F ratio 303 [261, 339] 305 [264, 337]
FiO2 delivery method
Nasal prong 109 (88.6%) 113 (90.4%)
High-flow nasal cannula 1 (0.8%) 6 (4.8%)
Face mask 12 (9.8%) 3 (2.4%)
Medication
Dexamethasone 115 (93.5%) 121 (96.8%)
Remdesivir 34 (27.6%) 70 (56%)
Tocilizumab 0 (0%) 2 (1.6%)
Code status
Full code 110 (89.4%) 119 (95.2%)
Do not resuscitate 5 (4.1%) 0 (0%)
Other 8 (6.5%) 5 (4%)
Abbreviations: COPD, chronic obstructive pulmonary disease; FiO2, fraction of inspired oxygen; IQR, interquartile range; S/F ratio, ratio of saturation of oxygen to fraction of inspired oxygen.
Missingness for all variables was <2%.
There was no difference in the rates of study withdrawal ([3%] in each group) or median number of hours spent in prone position within the group randomized to prone positioning (nondeferred = 7 [IQR 2.3–16.8], deferred = 4 [IQR 1.3–12.0]). The rate of serious adverse events was 4.9 % in the deferred consent group and 1.6 % in the nondeferred consent group Table 2 .Table 2 Secondary outcomes and rate of adverse events among patients enrolled with deferred vs. nondeferred consent
Outcome Deferred (N = 123) Non-deferred (N = 125)
Secondary outcomes
S/F ratio after 72 hours 332 [207, 423] 345 [260, 446]
Change in S/F ratio in first 72 hours (median [IQR]) 7 [−58, 73] 60 [−20, 108]
Change in FiO2 (%) in first 72 hours (median [IQR]) 0 [−7, 8] −4 [−8, 4]
Days to discharge (median [IQR]) 6 [4, 10] 4 [3, 6]
Discharged 114 (92.7%) 119 (95.2%)
Serious adverse events
SAE composite 6 (4.9%) 2 (1.6%)
Abbreviations: IQR, interquartile range; SAE, severe adverse event; S/F ratio, ratio of saturation of oxygen to fraction of inspired oxygen.
4 Discussion
In this secondary analysis of an international randomized controlled trial of prone positioning for noncritically ill patients with COVID-19, patients who underwent deferred consent were more likely to be male, older, and be enrolled on a weekend. We observed similar rates of both participant withdrawal and protocol adherence.
Our study has several limitations. First, data were unavailable for why deferred consent was chosen. Second, we did not ascertain patients' and providers’ perspectives on deferred consent, though prior literature has shown that both groups find it to be an acceptable alternative [13,14]. Third, this was an exploratory and post-hoc analysis, and thus we did not test for statistical significance and our findings require replication.
Our results suggest that the use of deferred consent may allow for the inclusion of patients who would not otherwise have been enrolled (e.g., patients hospitalized on weekends), potentially improving the external generalizability of randomized trials.
Conflicts of interest: The authors declare the following financial interests/personal relationships, which may be considered as potential competing interests. Michael Fralick is a consultant for ProofDx, a start-up company that has created a point of care device for COVID-19 using CRISPR
Funding: This study was funded by St Michael's Hospital Innovation fund, the Sinai Health Department of Medicine Research fund, and Sunnybrook Health Sciences Center Alternate Funding Plan Innovation Fund. The funders had no role in considering the study design or in the collection, analysis, or interpretation of data, the writing of the report, or the decision to submit the article for publication.
Author Contributions: Understanding how deferred consent affects patient characteristics and outcome: an exploratory analysis of a clinical trial of prone positioning for COVID-19. Study concept and design: All authors; Acquisition of data: All authors; Analysis/interpretation of data: All authors; Drafting of the manuscript: Colacci M, Fralick M; Critical revision of the manuscript: All authors; Statistical analysis: All authors.
==== Refs
References
1 Grimes D.A. Hubacher D. Nanda K. Schulz K.F. Moher D. Altman D.G. The Good Clinical Practice guideline: a bronze standard for clinical research Lancet 366 2005 172 174 16005342
2 Harron K. Woolfall K. Dwan K. Gamble C. Mok Q. Ramnarayan P. Deferred consent for randomized controlled trials in emergency care settings Pediatrics 136 2015 e1316 e1322 26438711
3 Topolovec-Vranic J. Santos M. Baker A.J. Smith O.M. Burns K.E.A. Deferred consent in a minimal-risk study involving critically ill subarachnoid hemorrhage patients Can Respir J 21 2014 293 296 24914705
4 Honarmand K. Belley-Cote E.P. Ulic D. Khalifa A. Gibson A. McClure G. The deferred consent model in a prospective observational study evaluating myocardial injury in the intensive care unit J Intensive Care Med 33 2018 475 480 29991343
5 The ARISE Investigators and the ANZICS Clinical Trials Group∗ Goal-directed resuscitation for patients with early septic shock N Engl J Med 371 2014 1496 1506 10.1056/NEJMoa1404380 25272316
6 Young P. Saxena M. Bellomo R. Freebairn R. Hammond N. van Haren F. Acetaminophen for fever in critically ill patients with suspected infection N Engl J Med 373 2015 2215 2224 26436473
7 Myburgh J.A. Finfer S. Bellomo R. Billot L. Cass A. Gattas D. Hydroxyethyl starch or saline for fluid resuscitation in intensive care N Engl J Med 367 2012 1901 1911 23075127
8 The NICE-SUGAR Study Investigators Intensive versus conventional glucose control in critically ill patients N Engl J Med 360 2009 1283 1297 19318384
9 Shamy M.C.F. Dewar B. Chevrier S. Wang C.Q. Page S. Goyal M. Deferral of consent in acute stroke trials Stroke 50 2019 1017 1020 30869570
10 Menon K. O’Hearn K. McNally J.D. Acharya A. Wong H.R. Lawson M. Comparison of consent models in a randomized trial of corticosteroids in pediatric septic shock Pediatr Crit Care Med 18 2017 1009 1018 28817507
11 Burns K.E.A. Zubrinich C. Tan W. Raptis S. Xiong W. Smith O. Research recruitment practices and critically ill patients: a multicenter, cross-sectional study (the consent study) Am J Respir Crit Care Med 187 2013 1212 1218 23525935
12 Byrne M.M. Tannenbaum S.L. Glück S. Hurley J. Antoni M. Participation in cancer clinical trials: why are patients not participating? Med Decis Making 34 2014 116 126 23897588
13 Manda-Taylor L. Bickton F.M. Gooding K. Rylance J. A formative qualitative study on the acceptability of deferred consent in adult emergency care research in Malawi J Empir Res Hum Res Ethics 14 2019 318 327 31390941
14 Den Boer M.C. Houtlosser M. Foglia E.E. Lopriore E. de Vries M.C. Engberts D.P. Deferred consent for delivery room studies: the providers’ perspective Arch Dis Child Fetal Neonatal Ed 105 2020 310 315 31427459
15 Fralick M. Colacci M. Munshi L. Venus K. Fidler L. Hussein H. Prone positioning of patients with moderate hypoxaemia due to covid-19: multicentre pragmatic randomised trial (COVID-PRONE) BMJ 376 2022 e068585 35321918
| 36273771 | PMC9706549 | NO-CC CODE | 2022-12-01 23:19:46 | no | J Clin Epidemiol. 2022 Oct 20; doi: 10.1016/j.jclinepi.2022.08.017 | utf-8 | J Clin Epidemiol | 2,022 | 10.1016/j.jclinepi.2022.08.017 | oa_other |
==== Front
Cell Rep Med
Cell Rep Med
Cell Reports Medicine
2666-3791
The Author(s).
S2666-3791(22)00414-1
10.1016/j.xcrm.2022.100850
100850
Report
Longitudinal analysis of serum neutralization of SARS-CoV-2 Omicron BA.2, BA.4, and BA.5 in patients receiving monoclonal antibodies
Bruel Timothée 1213∗
Stéfic Karl 34
Nguyen Yann 5
Toniutti Donatella 1
Staropoli Isabelle 1
Porrot Françoise 1
Guivel-Benhassine Florence 1
Bolland William-Henry 16
Planas Delphine 12
Hadjadj Jérôme 5
Handala Lynda 34
Planchais Cyril 7
Prot Matthieu 8
Simon-Lorière Etienne 8
André Emmanuel 910
Baele Guy 11
Cuypers Lize 9
Mouthon Luc 5
Mouquet Hugo 7
Buchrieser Julian 1
Sève Aymeric 12
Prazuck Thierry 12
Maes Piet 11
Terrier Benjamin 5
Hocqueloux Laurent 12
Schwartz Olivier 12∗∗
1 Virus and Immunity Unit, Institut Pasteur, Université Paris Cité, CNRS UMR3569, Paris, France
2 Vaccine Research Institute, Créteil, France
3 INSERM U1259, Université de Tours, Tours, France
4 CHRU de Tours, National Reference Center for HIV-Associated Laboratory, Tours, France
5 Department of Internal Medicine, National Reference Center for Rare Systemic Autoimmune Diseases, AP-HP, APHP.CUP, Hopital Cochin, Paris, France
6 Université Paris Cité, École doctorale BioSPC 562, Paris, France
7 Humoral Immunology Laboratory, Institut Pasteur, Université Paris Cité, INSERM U1222, Paris, France
8 G5 Evolutionary Genomics of RNA Viruses, Institut Pasteur, Université Paris Cité, Paris, France
9 University Hospitals Leuven, Department of Laboratory Medicine, National Reference Centre for Respiratory Pathogens, Leuven, Belgium
10 KU Leuven, Department of Microbiology, Immunology and Transplantation, Laboratory of Clinical Microbiology, Leuven, Belgium
11 KU Leuven, Department of Microbiology, Immunology and Transplantation, Laboratory of Clinical and Epidemiological Virology, Leuven, Belgium
12 CHR d’Orléans, Service de Maladies Infectieuses, Orléans, France
∗ Corresponding author
∗∗ Corresponding author
13 Lead contact
17 11 2022
17 11 2022
10085016 8 2022
10 10 2022
15 11 2022
© 2022 The Author(s)
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
The emergence of Omicron sublineages impacts the therapeutic efficacy of anti-severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) monoclonal antibodies (mAbs). Here, we evaluate neutralization and antibody-dependent cellular cytotoxicity (ADCC) activities of 6 therapeutic mAbs against Delta, BA.2, BA.4, and BA.5. The Omicron subvariants escape most antibodies but remain sensitive to bebtelovimab and cilgavimab. Consistent with their shared spike sequence, BA.4 and BA.5 display identical neutralization profiles. Sotrovimab is the most efficient at eliciting ADCC. We also analyze 121 sera from 40 immunocompromised individuals up to 6 months after infusion of Ronapreve (imdevimab + casirivimab) or Evusheld (cilgavimab + tixagevimab). Sera from Ronapreve-treated individuals do not neutralize Omicron subvariants. Evusheld-treated individuals neutralize BA.2 and BA.5, but titers are reduced. A longitudinal evaluation of sera from Evusheld-treated patients reveals a slow decay of mAb levels and neutralization, which is faster against BA.5. Our data shed light on antiviral activities of therapeutic mAbs and the duration of effectiveness of Evusheld pre-exposure prophylaxis.
Graphical abstract
Bruel et al. show that BA.4 and BA.5 Omicron sublineages escape most therapeutic antibodies, but cilgavimab and bebtelovimab remain fully active. Sotrovimab is the most efficient at eliciting antibody-dependent cellular cytotoxicity. Sera from individuals receiving Evusheld (cilgavimab/tixagevimab) pre-exposure prophylaxis neutralize Delta, BA.2, and BA.5 up to 6 months after injection.
Keywords
SARS-CoV-2
antibodies
Omicron
neutralization
ADCC
Published: November 17, 2022
==== Body
pmcIntroduction
Nine months after its emergence, the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) Omicron lineage outcompeted previous variants of concern (VOCs). Sublineages with improved transmissibility have replaced the initial Omicron BA.1 strain. As of September 2022, Omicron was composed of 5 main lineages, BA.1, BA.2, BA.3, BA.4, and BA.51 , 2, which further diversify into sublineages, such as BA.2.75, BA.2.75.2, BA.4.6, or BQ.1.1.3 The BA.1 strain has 34 mutations in its spike, which are associated with antibody escape,4 , 5 , 6 , 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 CD8 T cell evasion,16 and modified tropism.17 , 18 , 19 BA.2 harbors 30 mutations, 21 of which are not present in BA.1.2 The BA.4 and BA.5 sublineages share the same spike sequence, which differs from BA.2 by three mutations (including one reversion) in the receptor-binding domain (RBD) and one deletion in the N-terminal domain (NTD).1 The BA.5 sublineage was dominant in many countries in September 2022.20
Several monoclonal antibodies (mAbs) targeting the spike protein of SARS-CoV-2 are used in therapeutic, pre-exposure prophylaxis (PrEP) and post-exposure prophylaxis (PEP) settings.21 Therapeutic administration of mAbs is highly effective, reaching 85% efficacy in preventing coronavirus disease 19 (COVID-19)-related hospitalization or death.22 , 23 , 24 Antibody-based prophylaxis also achieves high levels of protection. Ronapreve (imdevimab + casirivimab) and Evusheld (cilgavimab + tixagevimab) cocktails provide 81% and 77% protection against symptomatic infection, respectively.25 , 26 These successes are mitigated by viral evolution. Omicron variants display considerable escape from mAbs.4 , 5 , 7 , 8 , 9 , 10 , 13 , 15 , 27 , 28 , 29 The use of Ronapreve (imdevimab + casirivimab) and Sotrovimab was discouraged after BA.1 and BA.2 emergence.30 It is recommended to inject a double dose of Evusheld instead, as serum neutralization is decreased against BA.1 and BA.2 in individuals receiving these antibodies as PrEP.29 , 30 , 31 , 32 Bebtelovimab is similarly effective against ancestral strains and Omicron BA.1 and BA.2,33 but its access is so far restricted to the United States.34 Thus, a continuous evaluation of mAb efficacy against new variants is needed to optimize their utilization.
As other Omicron sublineages, BA.4 and BA.5 escape most neutralizing mAbs.35 , 36 , 37 , 38 , 39 , 40 , 41 , 42 The neutralization profile of BA.4 and BA.5 is similar to that of BA.2, with only cilgavimab and bebtelovimab being efficient against these strains with high potency.35 , 36 , 37 , 38 , 39 Animal models revealed that some mAbs do not only rely on neutralization for therapeutic efficacy.43 , 44 Antibodies can trigger effector mechanisms through their fragment crystallizable (Fc) region. These Fc-effector functions mediate killing of infected cells through activation of antibody-dependent cellular cytotoxicity (ADCC) by natural killer (NK) cells and antibody-dependent complement-mediated lysis (ADCML) by complement, or clearance of viral particles, through macrophage-mediated antibody-dependent cellular phagocytosis (ADCP).45 Interaction between the Fc region and cognate Fc receptors (FcRs) may also promote inflammation and antibody-dependent enhancement (ADE) of infection.46 Hence, some therapeutic mAbs were mutated in the Fc region to abrogate FcR recognition and eliminate a putative risk of ADE.21 This is the case for cilgavimab and tixagevimab, which bear a triple mutation (TM) motif (L234F, L235E, and P331S). Fc engineering may also modulate neonatal Fc receptor (FcRn) affinity and extend antibody half-life.46 Such modifications were introduced in sotrovimab (M428L and N434S, called LS) and cilgavimab and tixagevimab (M252Y, S254Y, and T256E, called YTE). In contrast, imdevimab, casirivimab, and bebtelovimab have an unmutated Fc domain.21 Overall, the therapeutic activity of antibodies is the sum of neutralization potency, Fc-effector functions, and bio-disponibility.
Here, we evaluated the neutralization and ADCC activity of 6 therapeutic mAbs against BA.4 and BA.5 isolates. To consider variations in pharmacokinetic, dosage, or drug interactions, we also analyzed serum neutralization from 40 immunocompromised individuals up to 6 months post-infusion of Ronapreve (imdevimab + casirivimab) and Evusheld (cilgavimab + tixagevimab).
Results
In vitro neutralization of BA.4 and BA.5
We first investigated the sensitivity of two isolates of BA.4 and BA.5 to neutralization by mAbs. We selected 6 antibodies that are either used in patients (cilgavimab, tixagevimab, and bebtelovimab) or that were withdrawn because of Omicron escape (sotrovimab, casirivimab, and imdevimab). We used the commercial formulation, except for eebtlovimab, which is not available in France. We also tested Ronapreve (imdevimab + casirivimab) and Evusheld (cilgavimab + tixagevimab) cocktails. As controls, we included Delta and BA.2 strains.27 , 29 We used our S-Fuse neutralization assay,26 , 28 , 47 , 48 based on syncytia formation, to quantify infection via a green fluorescent protein (GFP) split system. As control, we included isolates of Delta and Omicron BA.2. Figure S1 summarizes the mutational landscape of VOCs included in the study compared with the ancestral Wuhan strain.
The IC50 of 4 out of the 6 mAbs (sotrovimab, tixagevimab, casirivimab, and imdevimab) were higher for BA.4 and BA.5 than for Delta (Figure 1A; Table 1 ). Tixagevimab and casirivimab lacked neutralization in the range of concentrations tested (Figure 1A; Table 1). Sotrovimab and imdevimab remained active but lost potency. Compared with Delta, sotrovimab was 15- and 17-fold less potent against BA.4 and BA.5, respectively. The increase in IC50s was higher for imdevimab: 110- and 86-fold against BA.4 and BA.5, respectively. Imdevimab remained more potent than sotrovimab against both strains (IC50 of 265 and 996 ng/mL for BA.4 and 208 and 1088 ng/mL for BA.5, respectively) (Figure 1A; Table 1). Importantly, cilgavimab and bebtelovimab displayed no or only minimal changes compared with Delta and remained highly potent against BA.4 and BA.5. When compared with BA.2, BA.4 and BA.5 display slightly improved neutralization by imdevimab (4.2- and 5.3-fold) and sotrovimab (9- and 8.3-fold) (Figure 1A; Table 1). We also analyzed the combination of cilgavimab and tixagevimab (Evusheld by Astrazeneca) and casirivimab and imdevimab (Ronapreve by Regeneron). Both displayed a drop in potency compared with Delta, which was less marked for Evusheld (BA.4: 10.4-fold and BA.5: 9-fold) than Ronapreve (BA.4: 330-fold and BA.5: 350-fold) (Figure 1A; Table 1).Figure 1 Neutralization and antibody-dependent cellular cytotoxicity of Omicron BA.4 and BA.5 by therapeutic mAbs
(A) Neutralization curves of mAbs using the S-Fuse system. Dose-response analysis of the neutralization by the indicated antibodies and by Evusheld, a combination of cilgavimab and tixagevimab, and Ronapreve, a combination of casirivimab and imdevimab. Data are mean ± SD of 2 independent experiments. The IC50 values for each antibody are presented in Table 1. The dashed line indicates the limit of detection.
(B) mAbs binding at the surface of Raji cells stably expressing the indicated spikes. Raji cells transduced with a control empty vector not coding for any spike (Empty). Depicted are EC50, calculated with a curve fitting the percentage of mAb-positive cells measured by flow cytometry against antibody concentration in limiting dilutions. Data are mean of 2 independent experiments. The EC50 values for each antibody are also presented in Table 1.
(C) Activation of the CD16 pathway as a surrogate of the capacity of each mAb to elicit antibody-dependent cellular cytotoxicity (ADCC). The area under curve of a dose-response analysis of CD16 activation by each mAb against each SARS-CoV-2 variant is depicted. Data are mean ± SD of 2 independent experiments.
Table 1 Neutralization, binding, and ADCC of therapeutic mAbs against Delta, BA.2 BA.4, and BA.5
Neutralization, IC50 (ng/ML) Binding, EC50 (ng/mL) ADCC, AUC
Delta BA.2 BA.4 BA.5 Delta BA.2 BA.4/5 Delta BA.2 BA.4/5
Sotrovimab 64.18 >9,000 996 1,088 72.3 1,429 848.2 11.1 5.7 6.4
Cilgavimab 7.9 6.1 6.5 11 22 39 49.3 0.1 0.2 0.2
Tixagevimab 1.6 >9,000 >9,000 >9,000 7.8 4,226 >9,000 0.2 0.2 0.1
Evusheld (cilgavimab/tixagevimab) 2.5 24.7 26.1 22.6 13.5 72.9 92.3 0.2 0.4 0.1
Bebtelovimab 3.8 4.5 1.3 2 12.2 20.2 17.4 1.6 1.6 2.1
Casirivimab 1.3 >9,000 >9,000 >9,000 11.1 >9,000 >9,000 0.9 0.3 0.2
Imdevimab 2.4 1,120 265 208 12.8 610 391 5.4 3.2 3.9
Ronapreve (imdevimab/casirivimab) 2 1,985 660 700 12.8 1,106 739 5.2 2.6 3.2
Overall, our data reveal a large escape of therapeutic mAbs by Omicron BA.4 and BA.5. BA.2., BA.4, and BA.5 have a similar profile of neutralization by these mAbs. Cilagavimab and bebtelovimab remain fully active against these variants.
Antibody binding to BA.4/5 spike and induction of ADCC
Next, we evaluated the capacity of these mAbs to bind to the BA.4 and BA.5 spike (referred to as BA.4/5 spike) and trigger ADCC. We assessed antibody binding by flow cytometry using Raji cells stably expressing the BA4/5 spike. As controls, we included Delta, BA.2 spikes, and cells transduced with an empty vector (Empty). To confirm spike expression and compare the various cells lines, we stained the cells with bebtlovimab, which neutralized Delta, BA.2, BA.4, and BA.5 with a similar potency, and analyzed them by flow cytometry (Figures S2A and S2B). All cell lines showed high levels of spike expression with a similar median fluorescence intensity (MFI) across variants (MFIs of 97,853, 71,735, and 68,635 for Delta, BA.2, and BA.4/5, respectively) (Figure S2B). No binding was observed with Raji-Empty cells (MFI of 784) (Figure S2B).
We analyzed the binding of the therapeutic mAbs and their combinations against these cells. We performed limiting dilution tests to calculate exact effective concentrations 50% (EC50) (Figures 1B and S2C; Table 1). All antibodies bound the Delta spike, with EC50 <100 ng/mL. This was expected given their neutralizing potency against this strain. Binding profiles were generally similar between BA.2 and BA.4/5, with both spikes displaying a high level of escape compared with Delta (Figure 1B). Bebtelovimab and cilgavimab displayed similar binding levels. Tixagevimab and casirivimab did not recognize the BA.4/5 spike, even at a high concentration (10 μg/mL). Sotrovimab and imdevimab recognized the BA.4/5 spike, with a loss of potency compared with delta (11.7- and 30.5-fold, respectively) (Figures 1B and S2C; Table 1).
Then, we investigated the capacity of these mAbs to trigger ADCC. We used a surrogate assay that measures the activation of the CD16 pathway. We previously demonstrated that this assay correlates to killing of infected cells by primary NK cells.47 We tested the antibodies by limiting dilution. We measured the area under the curve (AUC) to depict the ADCC capacity of the mAbs against each viral spike (Figures 1C and S2D; Table 1). As expected, none of the mAbs elicited CD16 activation against the Raji-Empty cells. CD16 activation was detectable against Raji-Delta, -BA.2, and -BA.4/5 cells. Sotrovimab was the most efficient mAb regardless of the viral strain, albeit the AUC was reduced against BA.2 and BA.4/5 compared with Delta (Figures 1C and S2D; Table 1). Cilgavimab and tixagevimab alone or in the Evusheld cocktail did not activate ADCC, in line with the mutations in their Fc domain that decrease binding to FcR (Figures 1C and S2D; Table 1). Bebtelovimab induced similar levels of ADCC activation against all strains, yet at low levels (Figures 1C and S2D; Table 1). Imdevimab and casirivimab, alone or in the Ronapreve cocktail, displayed intermediate levels of activation (Figures 1C and S2D; Table 1).
Next, we investigated the association between neutralization, binding, and ADCC capacity. Neutralization is positively correlated to binding (r = 0.97; p < 0.0001) but not to ADCC (r = 0.0053; p = 0.98) (Figure S3).
Overall, our results indicate that BA.4/5 avoid antibody recognition and ADCC activation by most of therapeutic mAbs tested. Sotrovimab is the most efficient ADCC inducer, and cilgavimab and tixagevimab lack ADCC activity.
Serum neutralization of BA.5 in individuals receiving mAbs
Next, we investigated antibody levels and neutralization in sera from 40 immunocompromised individuals who received, by intra-muscular injection, either 300 (n = 29) or 600 mg (n = 11) Evusheld as PrEP. Patients were sampled prior to and at a median of 26 (range 10–40) or 37 (range 14–48) days after the single or double dose, respectively. Among the 29 individuals who received 300 mg Evusheld, 17 previously received Ronapreve as PrEP. The last injection of Ronapreve occurred at a median of 35 days (range 29–49) before the first Evusheld injection. Two out of the 11 individuals who received 600 mg Evusheld also received Ronapreve. The injections were spaced by >160 days, which is ∼5 times above the half-life of Ronapreve.48 Participants were included in the study in two places, the Centre Hospitalier Regional (CHR) in Orléans (France; n = 8) or the Hôpital Cochin in Paris (France; n = 32). Most of the patients were female (n = 28), were diagnosed with anti-neutrophil cytoplasmic antibody (ANCA)-associated vasculitis (n = 26) and treated with rituximab as immunosuppressive therapy (n = 31). A complete description of the patients’ characteristics is provided in Table 2 . The 8 individuals recruited at the CHR Orléans were longitudinally sampled every month as part of an ongoing prospective cohort.Table 2 Characteristics of patients
Orléans cohort Cochin group Total (%)
Patient characteristics
N 8 32 40 (100)
Gender, female (♀) 6 22 28 (70)
Gender, male (♂) 2 10 12 (30)
Obesity 3 1 4 (10)
Diseases
Rheumatoid arthritis 5 2 7 (17.5)
Kidney graft 2 0 2 (5)
Myelodysplasia 1 0 1 (2.5)
ANCA-associtaed vasculitis 0 26 26 (65)
Polychondritis 0 1 1 (2.5)
Lupus 0 1 1 (2.5)
Systemic sclerosis 0 1 1 (2.5)
Cryoglobuminemic vasculitis 0 1 1 (2.5)
Medications
Rituximab (anti-CD20) 5 26 31 (77.5)
Infliximab (anti-tumor necrosis factor [TNF]) 0 1 1 (2.5)
Mepolizumab (anti-interleukin-5 [IL-5]) 0 1 1 (2.5)
Prednisone 4 10 14 (35)
Mycofenolate mofetil 2 1 3 (7.5)
Methotrexate 0 3 3 (7.5)
5-Azacytidine 1 0 1 (2.5)
Tacrolimus 1 0 1 (2.5)
Cyclosporin 1 0 1 (2.5)
Cyclophosphamide 0 1 1 (2.5)
Vaccines doses
2 0 1 1 (2.5)
3 5 25 30 (75)
4 3 6 9 (22.5)
Previous COVID-19 0 8 8 (20)
PrEP
Ronapreve 3 17 20 (50)
Evusheld 300 mg 8 21 29 (72.5)
Evusheld 600 mg 0 11 11 (27.5)
We analyzed the 40 individuals before and after infusion of Evusheld. We categorized the patients into 5 groups: naive of any antibody administration (n = 11; before treatment); Ronapreve (n = 18; who received 1200 mg Ronapreve); Ronapreve + Evusheld (n = 18; who received 1,200 mg Ronapreve followed by 300 mg Evulsheld); Evusheld (n = 11; 300 mg); and Evusheld×2 (n = 11; 600 mg in a single injection). All individuals received at least two doses of vaccine but failed to mount an antibody response >260 binding antibody unit (BAU)/mL, making them eligible for Evusheld according to French recommendations. Eight out the 40 individuals included in the cohort had COVID-19 but remained >260 BAU/mL. Three patients had COVID-19 after Evusheld administration. These breakthrough infections were previously described.29
We first measured the levels of anti-spike immunoglobulin Gs (IgGs) in BAU (BAU/mL) using the S-Flow assay.49 Compared with the naive group, sera containing mAbs display a sharp increase in anti-spike (S) IgG (median of 38 versus 3,449, 3,591, 1,323, and 2,623 BAU/mL for Ronapreve, Ronapreve + Evusheld, Evusheld, and Evusheld×2, respectively) (Figure 2A). We then investigated serum neutralization against Delta, BA.2, and BA.5 with the S-Fuse assay. We tested sera in limiting dilutions to calculate titers as effective dilution 50% (ED50). We did not include BA.4 as its profile of neutralization is identical to BA.5. Untreated individuals (naive group) did not neutralize the three strains, except for one patient who slightly neutralized Delta. Infusion of mAbs dramatically increased Delta neutralization, with increases from 552- to 2,484-fold compared with the naive group. Individuals who received Ronapreve neutralized BA.2 and BA.5 at low levels (non-significant compared with naive individuals; two-sided Kruskall-Wallis test with Dunn’s multiple comparison correction). The groups that included Evusheld neutralized BA.2 and BA.5 at levels significantly higher than the naive and Ronapreve groups (Figure 2B). Of note, neutralization titers were higher in the Evusheld×2 group with all variants tested (16,585 versus 23,772, 992 versus 1,908, and 511 versus 539 against Delta, BA.2, and BA.5, respectively), without reaching statistical significance. In this group, the two individuals who received Ronapreve >160 days prior to Evusheld did not harbor significantly higher levels of neutralization (Mann-Whitney test; p = 0.1455, p = 0.1455, and p = 0.2182, for Delta, BA.2, and BA.5, respectively) (Figure S4A). Neutralization titers tended to be lower against BA.5 than BA.2, but this difference was significant only for individuals who received 300 mg Evusheld (1,549 versus 489; p = 0.0228) (Figure S4B).Figure 2 Antibody levels and neutralization of Delta, BA.2, and BA.5 in sera of immunocompromised individuals receiving mAbs
(A) Anti-S IgGs were measured using the flow cytometry-based S-Flow assay in sera of individuals before PrEP (naive; n = 11), treated with Ronapreve (n = 18), treated with 300 (n = 11) or 600 mg (n = 11) Evusheld, or treated with both Ronapreve and 300 mg Evusheld (n = 18). Indicated are the binding antibody units (BAUs) per mL (BAU/mL) of anti-S IgGs. Two-sided Kruskall-Wallis test with Dunn’s multiple comparison correction. Each dot is an individual. Red bars indicate medians.
(B) Serum neutralization of Delta and Omicron BA.2 and BA.5 in the same individuals as in (A). Indicated are effective dilution 50% (ED50; titers) as calculated with the S-Fuse assay. Two-sided Kruskall-Wallis test with Dunn’s multiple comparison correction. Each dot is an individual. Red bars indicate median. The dashed line indicates the limit of detection.
(C) Longitudinal measurement of anti-S levels in 5 immunocompromised individuals who initiated an Evusheld PrEP with no history of Ronapreve. All individuals and sampling points are depicted (black lines and dots). The red lines indicate medians. Indicated are the BAU/mL of anti-S IgGs. The dashed line indicates the limit of detection.
(D) Sero-neutralization of Delta and Omicron BA.2 and BA.5 in the same individuals as in (C). Indicated are ED50 (titers) as calculated with the S-Fuse assay. Two-sided Kruskall-Wallis test with Dunn’s multiple comparison correction. Each dot is an individual. Red bars indicate medians. The dashed line indicates the limit of detection.
Overall, these data show that the serum neutralization activity of individuals receiving Ronapreve and Evusheld as PrEP is decreased against BA.2 and BA.5. The diminution is less marked with Evusheld- than Ronapreve-treated individuals.
Kinetics of serum neutralization up to 6 months after infusion of Evusheld
Longitudinal sampling was performed for 8 out of the 40 individuals (for patients’ characteristics, see Table 2). The serum samples were available up to 186 days post-administration of Evusheld (300 mg). We investigated anti-S IgG levels and neutralization (Figures 2C, 2D, and S5). We first analyzed 5 patients who were naive at the time of Evusheld injection (the 3 others were previously under Ronapreve PrEP) (Figures 2C and 2D). Antibody levels peaked at 28 (range 28–30) days after Evusheld administration, with a median of 1,400 (range 646–4,014) BAU/mL. Anti-S IgG then slowly decreased, reaching 500 (range 452–548) BAU/mL at 176 (range 175–177) days post-administration (Figure 2C). Neutralization of Delta, BA.2, and BA.5 mirrored anti-S levels, showing a sharp increase upon administration and a steady decrease until month 6 (Figure 2D). Neutralization of Delta remains consistently higher than that of BA.2 and BA.5, in line with our other observations (Figures 1A, 2B, and S4B). After almost 6 months of follow up, the five patients who received Evusheld harbored detectable levels of neutralization against the tested strains. The neutralization levels against the two Omicron subvariants were low at 6 months (ED50 of 1,503, 202, and 59, for Delta, BA.2, and BA.5, respectively) (Figure 2D). We also analyzed the 3 individuals who received Ronapreve prior to Evusheld (Figure S5). Together, their profiles were similar to those who only received Evusheld. However, they consistently harbored higher levels of anti-S and neutralization titers in the first 2 months, suggesting a disappearance of Ronapreve, but a maintenance of Evusheld, as expected given the longer half-life of Evusheld than Ronapreve. We also analyzed ADCC activity in these sera by evaluating their capacity to activate the CD16 pathway. Some individuals displayed slight activation against cells expressing the delta S, but generally ADCC activation was undetectable (Figures S6A and S6B). This is consistent with the mutated Fc of cilgavimab and tixagevimab.
Overall, these results show that a single administration of Evusheld allowed serum neutralization of BA.5 for 6 months, with reduced titers compared with Delta.
Discussion
We show here that BA.4 and BA.5 escape neutralization by most therapeutic mAbs, in line with previous reports.35 , 36 , 37 , 38 , 39 , 40 , 41 , 42 Some antibodies remain effective. Bebtelovimab is the most potent, followed by cilgavimab. Tixagevimab and casirivimab lost any neutralizing activity, and imdevimab and sotrovimab were poorly active against BA.4/BA.5. We observed a slightly higher neutralization of BA.5 by sotrovimab compared with BA.2. Similar findings were reported,38 , 39 , 42 but others have found decreased or identical neutralization of BA.5 compared with BA.235 , 36 , 37 , 40 , 41. There is also some discrepancy in the literature regarding cilgavimab. We observed a similar neutralization against BA.2 and BA.5 for this antibody, whereas some reports showed a slight decrease against BA.5 compared with BA.236 , 37 , 39. Further investigations are needed to determine why some mAbs more susceptible to experimental variations. It may be due to the use of different target cells (which vary in their ACE2 levels), viral isolates, or pseudotypes.50 Overall, the data presented here and the literature indicate that BA.2 and BA.5 have a very close spectrum of neutralization by clinically available mAbs.
In addition to our in vitro evaluation of mAbs neutralization, we analyzed the sera of immunocompromised individuals receiving Ronapreve or Evusheld as PrEP. In line with our in vitro observation, Ronapreve-treated individuals barely neutralized Omicron sublineages. Evusheld-treated individuals had detectable neutralization against BA.2 and BA.5, albeit decreased compared with Delta. We observed a trend for a higher neutralization of BA.2 than BA.5 in individuals receiving Evusheld. Longitudinal evaluation in 8 patients showed a slightly faster decay of antibody responses against BA.5. This difference between BA.2 and BA.5 may be explained by the loss of BA.5 binding and neutralization by tixagevimab. This decrease may be negligible when the Evusheld antibodies are tested alone but is more visible in the serum. How serum components might affect Evusheld potency against BA.5 deserve further investigations. Cilgavimab may also be slightly less potent against BA.4/5 than BA.2, as reported by others.36 , 37 , 39 The difference of BA.2 and BA.5 serum neutralization in Evusheld-treated individuals and the observation of a faster antibody decay stress the need for a booster dose of mAbs after 6 months, as is currently recommended. It may be of great interest to evaluate the impact of an earlier booster dose of Evusheld to compensate for Omicron escape.
How vaccination and antibody-based PrEP may be combined is an interesting question. Infusion of the HIV-1 broadly neutralizing antibody 3BNC117 increases humoral immunity and cross-neutralization in patients infected with HIV-1.51 The mechanisms underlying this improvement of autologous immune response remain ill defined but likely include the formation of immune complexes and their processing by antigen-presenting cells.52 Besides this “vaccinal effect,” mAbs may mask epitopes and alter immune responses, as suggested by an in-depth analysis of the B cell compartment in a cohort of individuals receiving COVID-19 vaccination after infusion of two anti-S mAbs.53 Thus, the interplay between mAbs and the immune system is complex, with outcomes that remain difficult to predict. Whether these antibody feedbacks loops may be manipulated to improve immunotherapies and prophylaxis of immunocompromised individuals deserves further investigation.
There was large inter-individual variability in neutralization and antibody levels after mAb administration. Recent work demonstrated an impact of the body mass index (BMI) on antibody levels after Evusheld injection, with a high BMI associated with low titers.31 This is consistent with the unique recommended dosage of Evusheld (initially 300 and then 600 mg). In our study, we observed a non-significant trend for higher titers in individuals receiving 600 mg and no association with BMI. This lack of significance is likely due to the small number of individuals tested and to additional factors accounting for the inter-individual’s variability. An investigation of a larger cohort previously demonstrated a significant increase in antibody levels in individuals received 600 compared with 300 mg.32 It will be interesting to evaluate whether adapting the dose to BMI may homogenize the response to Evusheld PrEP and improve its efficacy.
We also tested the binding and ADCC capacity of these mAbs. Binding correlated to neutralization but not to ADCC. The most potent antibody to activate ADCC against Omicron sublineages was sotrovimab, even if its neutralization IC50 was relatively high compared with other antibodies. This is likely the consequence of a distinct binding mode of sotrovimab, with an orientation and/or a positioning of its Fc that may facilitate the interaction with FcRs.54 This ADCC activity may help understand why sotrovimab remains clinically active against BA.2 despite its very limited neutralization.55 Similarly, it has been reported that non-neutralizing antibodies capable of mediating Fc-effector functions display some efficacy in animal models.56 It may be worth examining whether a combination of sotrovimab and Evusheld or bebtelovimab may improve therapeutic efficacy of the mAbs.
In conclusion, we provide here an in-depth evaluation of the efficacy of therapeutic mAb and serum from mAb-treated patients against Omicron sublineages. The BA.5 variant remains sensitive to Evusheld, but the decay of the serum neutralizing activity in treated individuals is accelerated, compared with previously circulating variants.
Limitations of the study
Our study has limitations. First, our sample size is small, precluding the analysis of patient characteristics associated with high serum neutralization titers. Whether gender, age, ongoing medication, or underlying conditions modulate bio-disponibility of mAbs remain open questions. Second, we did not have access to mucosal samples. Systemic levels of antibodies are known to be key to prevent severe COVID-19, whereas mucosal mAb levels may correlate with protection from infection. Third, we restricted our investigation of Fc-effector functions to ADCC. Thus, it will be worth determining how ADCP and ADCML contribute to the antiviral activity of therapeutic anti-S mAbs. Our study was also limited to BA.4 and BA.5, and we did not analyze the sensitivity of other Omicron subvariants, such as BA.2.75, BA.2.75.2, BA.4.6, or BQ1.13 , 57 , 58 , 59. We did not have access to the medical formulation of bebtelovimab.34 Fc-effector functions are influenced by the method of antibody preparation, the isotype, Fc glycosylation, and mutations. Our observation that bebtelovimab is a poor ADCC inducer deserves to be confirmed using the medical formulation.
STAR★Methods
Key resources table
REAGENT or RESOURCE SOURCE IDENTIFIER
Antibodies
Bamlanivimab Kind gift of Dr Thierry Prazuck (CHR d’Orléans, France) N/A
Casirivimab Kind gift of Dr Thierry Prazuck (CHR d’Orléans, France) N/A
Etesevimab Kind gift of Dr Thierry Prazuck (CHR d’Orléans, France) N/A
Imdevimab Kind gift of Dr Thierry Prazuck (CHR d’Orléans, France) N/A
Cilgavimab Kind gift of Dr Thierry Prazuck (CHR d’Orléans, France) N/A
Tixagevimab Kind gift of Dr Thierry Prazuck (CHR d’Orléans, France) N/A
Sotrovimab Kind gift of Dr Thierry Prazuck (CHR d’Orléans, France) N/A
Bebtelovimab Kind gift of Dr Thierry Prazuck (CHR d’Orléans, France) N/A
anti-IgG AlexaFluor647 Jackson ImmunoResearch Cat#A-21445
Bacterial and virus strains
D614G(hCoV-19/France/GE1973/2020) National Reference Centre for Respiratory Viruses (Institut Pasteur, Paris, France) EPI_ISL_41463
Delta Laboratory of Virology of Hopital Européen Georges Pompidou (Assistance Publique – Hopitaux de Paris) EPI_ISL_2029113
Omicron BA.2 NRC UZ/KU Leuven, Belgium EPI_ISL_10654979
Omicron BA.4 NRC UZ/KU Leuven, Belgium EPI_ISL_15728568
Omicron BA.5 CHU de Tours, France EPI_ISL_13660702
Chemicals, peptides, and recombinant proteins
Hoechst 33342 Invitrogen Cat#H3570
Paraformaldehyde 4% Alfa Aesar Cat#J19943.K2
Critical commercial assays
ADCC Reporter Bioassay Promega Cat#G7010
Bright-Glo Luciferase Assay System Promega Cat#E2620
Deposited data
D614G(hCoV-19/France/GE1973/2020) GISAID EPI_ISL_41463
Delta GISAID EPI_ISL_2029113
Omicron BA.2 GISAID EPI_ISL_10654979
Omicron BA.4 GISAID EPI_ISL_15728568
Omicron BA.5 GISAID EPI_ISL_13660702
Experimental models: Cell lines
293T ATCC Cat#CRL-3216
U2OS cells ATCC Cat#HTB-96
Raji ATCC Cat#CCL-86
Software and algorithms
Harmony High-Content Imaging and Analysis Software PerkinElmer Cat#HH17000012
Excel 365 Microsoft https://www.microsoft.com/en-ca/microsoft-365/excel
Prism 8 Graphpad https://www.graphpad.com/
FlowJo v10 Tree Star https://www.flowjo.com/
Resource availability
Lead contact
Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Timothée Bruel ([email protected]).
Materials availability
All unique/stable reagents generated in this study are available from the lead contact with a completed Materials Transfer Agreement.
Experimental model and subject details
Human subjects
Individuals under Evusheld PreP were recruited in the French cities of Orléans and Paris (CHR d’Orléans and Hôpital Cochin). The Neutralizing Power of Anti-SARS-CoV-2 Serum Antibodies (PNAS) cohort is an ongoing prospective, monocentric, longitudinal, observational cohort clinical study aiming to describe the kinetics of neutralizing antibodies after SARS-CoV-2 infection or vaccination (ClinicalTrials.gov identifier: NCT05315583). The cohort takes place in Orélans, France and enrolled immunocompromised individuals receiving Evusheld PreP. This study was approved by the Est II (Besançon) ethical committee. At enrollment, written informed consent was collected, and participants completed a questionnaire that covered sociodemographic characteristics, clinical information and data related to anti-SARS-CoV-2 vaccination. Blood sampling was performed on the day of Evusheld infusion and after 3 days, 15 days and then every months. None of the patients self-reported a COVID-19 during the study period. The ‘Cochin’ cohort is a prospective, monocentric, longitudinal, observational clinical study (NCT04870411) enrolling immunocompromised individuals with rheumatic diseases, aiming at describing immunological responses to COVID-19 vaccine in patients with autoimmune and inflammatory diseases treated with immunosuppressants and/or biologics. Ethics approval was obtained from the Comité de Protection des Personnes Nord-Ouest II. Leftover sera from usual care were used from these individuals in the setting of the local biological samples collection (RAPIDEM). A written informed consent was collected for all participants. None of the study participants received compensation.
Viral strains
All strain were isolated from a nasopharyngeal swab using Vero E6 cells (ADCC: CRL-1586™) tested negative for mycoplasma. The Delta and Omicron BA.2 strains were previously described.27 , 29 BA.4 and BA.5 strains were isolated from Belgian and French patients, respectively. BA.4 was isolated and sequenced by the NRC UZ/KU Leuven (Belgium). BA.5 was isolated from a 67-year-old female patient. On May 15, she experienced mild and unspecific symptoms, she tested positive for COVID-19 using a lateral flow assay. Due to pre-existing conditions (polymyalgia rheumatica), she presented at the hospital on 17/05, where a nasal swab was collected. A PCR testing (kit: Eurobioplex Fast-SVD-EBX-047 from Eurobio scientific) identified BA.5, which was confirmed by sequencing using the NEBNext ARTIC SARS-COV-2 Companion Kit for Oxford Nanopore (New England BioLabs) with Varskip Short v2 and BA.2 Spike-in supplemental primers (https://github.com/nebiolabs/VarSkip/commit/3fd0283adb878fe24e16b161c4e5c1c4364cd4c0) as per the manufacturer’s instruction. Her COVID-19 symptoms remain mild (arthralgia and cough). Both patients provided informed consent for the use of the biological materials. The sequences of the isolates were deposited on GISAID immediately after their generation, with the following Delta ID: EPI_ISL_2029113; Omicron BA.2 GISAID ID: EPI_ISL_10654979; Omicron BA.4 GISAID ID: EPI_ISL_15728568; Omicron BA.5: EPI_ISL_13660702. viral stocks were titrated in limiting dilution on Vero E6 cells and on S-Fuse cells.
mAbs
Bamlanivimab, Casirivimab, Etesevimab, Imdevimab, Cilgavimab, Tixagevimab and Sotrovimab were provided by CHR Orleans. Bebtelovimab was produced as previously described.29
Cell lines
Raji cells (ATCC CCL-86) were grown in complete RPMI medium (10% Fetal Calf Serum (FCS), 1% Penicillin/Streptomycin (PS)). 293T cells (ATCC CRL-3216) and U2OS cells (ATCCa HTB-96) were grown in complete DMEM medium (10% FCS, 1% PS). U2OS stably expressing ACE2 and the GFPsplit system (GFP1-10 and GFP11; S-Fuse cells) were previously described.60 Blasticidin (10 mg/mL) and puromycin (1 mg/mL) were used to select for ACE2 and GFPsplit transgenes expression, respectively. Raji cells stably expressing the SARS-CoV-2 Spike protein of Delta, BA.2 and BA.4/5 (GenBank: QHD43416.1, UJP23605.1 and UPN16705.1) were generated by lentiviral transduction and selection with puromycin (1 mg/mL). Absence of mycoplasma contamination was confirmed in all cell lines with the Mycoalert Mycoplasma Detection Kit (Lonza). All cell lines were cultured at 37°C and 5% CO2.
Method details
Anti-spike antibody binding and serology
Circulating levels of anti-S antibodies were measured with the S-Flow assay. This assay uses 293T cells stably expressing the spike protein (293T spike cells) and 293T control cells as control to detect anti-spike antibodies by flow cytometry.49 In brief, the cells were incubated at 4 °C for 30 minutes with sera (1:300 dilution) in PBS containing 0.5% BSA and 2 mM EDTA. Cells were then washed with PBS and stained with an anti-human IgG Fc Alexa Fluor 647 antibody (109-605-170, Jackson Immuno Research). After 30 minutes at 4 °C, cells were washed with PBS and fixed for 10 minutes using 4% PFA. A standard curve with serial dilutions of a human anti-spike monoclonal antibody (mAb48) was acquired in each assay to standardize the results as a binding Unit (BU). Data were acquired on an Attune NxT instrument using Attune NxT software version 3.2.2 (Life Technologies) and analyzed with FlowJo version 10.7.1 software (see Extended Data Figure 4 for gating strategy). The sensitivity is 99.2% with a 95% confidence interval of 97.69–99.78%, and the specificity is 100% (98.5–100%)40. To determine BAU/mL (Binding Antibody Units per mL), we analyzed a series of vaccinated (n = 144), convalescent (n = 59) samples and World Health Organization international reference sera (20/136 and 20/130) on S-Flow and on two commercially available ELISAs (Abbott 147 and Beckmann 56). Using this dataset, we performed a Passing–Pablok regression, which shows that the relationship between BU the S-Flow (see above) and BAU/mL is linear, allowing calculation of BAU/mL using S-Flow data.61 The binding mAbs to Delta, BA.2 and BA.4/5 spikes was assessed using Raji cells stably expressing these spikes. Stainings were performed at the indicated concentration of mAbs and following the S-Flow protocol, except that antibodies were biotinylated and revealed with a streptavidin conjugated to AlexaFluor647 (Life Technologies; dilution 1:400).
S-Fuse neutralization assay
U2OS-ACE2 GFP1-10 or GFP11 cells, also termed S-Fuse cells, become GFP + when they are productively infected by SARS-CoV-2. Cells tested negative for mycoplasma. Cells were mixed (ratio 1:1) and plated at 8 × 103 per well in a μClear 96-well plate (Greiner Bio-One). The indicated SARS-CoV-2 strains were incubated with serially diluted mAb or sera for 15 minutes at room temperature and added to S-Fuse cells. The sera were heat-inactivated for 30 minutes at 56 °C before use. Eighteen hours later, cells were fixed with 2% paraformaldehyde (PFA), washed and stained with Hoechst (dilution 1:1,000, Invitrogen). Images were acquired with an Opera Phenix high-content confocal microscope (PerkinElmer). The GFP area and the number of nuclei were quantified using Harmony software version 4.9 (PerkinElmer). The percentage of neutralization was calculated using the number of syncytia as value with the following formula: 100 × (1 − (value with serum −value in ‘non-infected’)/(value in ‘no serum’ − value in ‘non-infected’)). Neutralizing activity of each serum was expressed as the ED50. ED50 values (in μg ml−1 for mAbs and in dilution values for sera) were calculated with a reconstructed curve using the percentage of the neutralization at the different concentrations. We previously reported correlations between neutralization titers obtained with the S-Fuse assay and both pseudovirus neutralization and microneutralization assays.62 , 63 Of note, we previously reported that the neutralization assay with the S-Fuse system is not affected by differences in fusogenicity between variants.27
Antibody-dependent cellular cytotoxicity reporter assay
ADCC was quantified using the ADCC Reporter Bioassay (Promega) as previously described.47 Briefly, 5 × 104 Raji stably expressing the indicated spikes were co-cultured with 5 × 104 Jurkat-CD16-NFAT-rLuc cells in presence or absence of mAbs at the indicated concentration. Luciferase was measured after 18 h of incubation using an EnSpire plate reader (PerkinElmer). ADCC was measured as the fold induction of Luciferase activity compared to the ‘‘no serum’’ condition. Sera were tested at a 1:100 dilution and normalized to the control condition to account for inter-individual variations of the background.
Quantification and statistical analysis
Statistical analysis
No statistical methods were used to predetermine sample size. The experiments were not randomized, and the investigators were not blinded. Flow cytometry data were analyzed with FlowJo version 10 software. Calculations were performed using Excel 365 (Microsoft). Figures were drawn on Prism 9 (GraphPad Software). Statistical analysis was conducted using GraphPad Prism 9. Statistical significance between different groups was calculated using Kruskall–Wallis test with Dunn’s multiple comparisons, Friedman tests with Dunn’s multiple comparison correction and Spearman non-parametric correlation test. All tests were two-sided.
Supplemental information
Document S1. Figures S1–S6
Document S2. Article plus supplemental information
Data and code availability
SARS-CoV-2 variants genomes have been deposited at GISAID and are publicly available as of the date of publication. Accession numbers are listed in the key resources table. This study did not generate any new codes. Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.
Acknowledgments
We thank the European Health Emergency Preparedness and Response Authority (HERA) for supporting the work being done at 10.13039/501100003762 Institut Pasteur and UK Leuven. We thank the patients who participated in this study. We thank members of the Virus and Immunity Unit for discussions and help. We thank Ludivine Grzelak for her help in creating an illustration of SARS-CoV-2 variant mutations. We thank N. Aulner and the UtechS Photonic BioImaging (UPBI) core facility (Institut Pasteur), a member of the France BioImaging network, for image acquisition and analysis. The Opera system was co-funded by 10.13039/501100003762 Institut Pasteur and the Région Ȋle-de-France (DIM1Health). We thank F. Peira, V. Legros, and L. Courtellemont for their help with the cohorts. UZ Leuven, as a national reference center for respiratory pathogens, is supported by Sciensano, which is gratefully acknowledged. We thank Hélène Péré and David Veyer for their help in sequencing viral strains and helpful discussions. Work in the O.S. lab is funded by 10.13039/501100003762 Institut Pasteur ; Urgence COVID-19 Fundraising Campaign of 10.13039/501100003762 Institut Pasteur ; 10.13039/501100002915 Fondation pour la Recherche Médicale (10.13039/501100002915 FRM ); 10.13039/501100003323 ANRS ; the Vaccine Research Institute (ANR-10-LABX-77); Labex IBEID (ANR-10-LABX-62-IBEID); ANR/FRM Flash Covid PROTEO-SARS-CoV-2; and ANR Coronamito and IDISCOVR. Work in the UPBI facility is funded by grant ANR-10-INSB-04-01 and the Région Ȋle-de-France program DIM1Health. D.P. is supported by the Vaccine Research Institute. P.M. acknowledges the support of a COVID-19 research grant from “10.13039/501100003130 Fonds Wetenschappelijk Onderzoek ” /10.13039/501100003130 Research Foundation Flanders (grant G0H4420N) and “Internal Funds 10.13039/501100004040 KU Leuven ” (grant 3M170314). E.S.-L. acknowledges funding from the INCEPTION program (Investissements d’Avenir grant ANR-16-CONV-0005). The funders of this study had no role in study design, data collection, analysis, and interpretation or writing of the article.
Author contributions
Experimental strategy and design, T.B. and O.S.; laboratory experiments, T.B., D.T., I.S., F.P., F.G.-B., W.-H.B., D.P., M.P., E.S.-L., and J.B.; cohort management and clinical research, Y.N., J.H., L.M., A.S., T.P., B.T., and L.H.; viral strains and antibodies, K.S., L.H., C.P., E.A., G.B., L.C., and H.M.; manuscript writing, T.B. and O.S.; manuscript editing, T.B., Y.N., W.-H.B., B.T., L.H., and O.S.
Declaration of interests
T.B., C.P., H.M., and O.S. have a pending patent application for an anti-RBD mAb not used in this study (PCT/FR2021/070522).
Inclusion and diversity
We support inclusive, diverse, and equitable conduct of research.
Supplemental information can be found online at https://doi.org/10.1016/j.xcrm.2022.100850.
==== Refs
References
1 Tegally H. Moir M. Everatt J. Giovanetti M. Scheepers C. Wilkinson E. Subramoney K. Makatini Z. Moyo S. Amoako D.G. Emergence of SARS-CoV-2 omicron lineages BA.4 and BA.5 in South Africa Nat. Med. 28 2022 1785 1790 10.1038/s41591-022-01911-2 35760080
2 Viana R. Moyo S. Amoako D.G. Tegally H. Scheepers C. Althaus C.L. Anyaneji U.J. Bester P.A. Boni M.F. Chand M. Rapid epidemic expansion of the SARS-CoV-2 Omicron variant in southern Africa Nature 603 2022 679 686 10.1038/s41586-022-04411-y 35042229
3 Cao Y. Jian F. Wang J. Yu Y. Song W. Yisimayi A. Wang J. An R. Zhang N. Wang Y. Imprinted SARS-CoV-2 humoral immunity induces converging Omicron RBD evolution Preprint at bioRxiv 2022 10.1101/2022.09.15.507787 ▪▪▪–▪▪▪
4 Iketani S. Liu L. Guo Y. Liu L. Chan J.F.-W. Huang Y. Wang M. Luo Y. Yu J. Chu H. Antibody evasion properties of SARS-CoV-2 Omicron sublineages Nature 604 2022 553 556 10.1038/s41586-022-04594-4 35240676
5 VanBlargan L.A. Errico J.M. Halfmann P.J. Zost S.J. Crowe J.E. Purcell L.A. Kawaoka Y. Corti D. Fremont D.H. Diamond M.S. An infectious SARS-CoV-2 B.1.1.529 Omicron virus escapes neutralization by therapeutic monoclonal antibodies Nat. Med. 28 2022 490 495 10.1038/s41591-021-01678-y 35046573
6 Carreño J.M. Alshammary H. Tcheou J. Singh G. Raskin A.J. Kawabata H. Sominsky L.A. Clark J.J. Adelsberg D.C. Bielak D.A. Activity of convalescent and vaccine serum against SARS-CoV-2 Omicron Nature 602 2022 682 688 10.1038/s41586-022-04399-5 35016197
7 Cameroni E. Bowen J.E. Rosen L.E. Saliba C. Zepeda S.K. Culap K. Pinto D. VanBlargan L.A. De Marco A. di Iulio J. Broadly neutralizing antibodies overcome SARS-CoV-2 Omicron antigenic shift Nature 602 2022 664 670 10.1038/s41586-021-04386-2 35016195
8 Takashita E. Kinoshita N. Yamayoshi S. Sakai-Tagawa Y. Fujisaki S. Ito M. Iwatsuki-Horimoto K. Chiba S. Halfmann P. Nagai H. Efficacy of antibodies and antiviral drugs against covid-19 omicron variant N. Engl. J. Med. 386 2022 995 998 10.1056/nejmc2119407 35081300
9 Touret F. Baronti C. Bouzidi H.S. de Lamballerie X. In vitro evaluation of therapeutic antibodies against a SARS-CoV-2 Omicron B.1.1.529 isolate Sci. Rep. 12 2022 4683 10.1038/s41598-022-08559-5 35304531
10 Tada T. Zhou H. Dcosta B.M. Samanovic M.I. Chivukula V. Herati R.S. Hubbard S.R. Mulligan M.J. Landau N.R. Increased resistance of SARS-CoV-2 Omicron variant to neutralization by vaccine-elicited and therapeutic antibodies EBioMedicine 78 2022 103944 10.1016/j.ebiom.2022.103944
11 Garcia-Beltran W.F. St Denis K.J. Hoelzemer A. Lam E.C. Nitido A.D. Sheehan M.L. Berrios C. Ofoman O. Chang C.C. Hauser B.M. mRNA-based COVID-19 vaccine boosters induce neutralizing immunity against SARS-CoV-2 Omicron variant Cell 185 2022 457 466.e4 10.1016/j.cell.2021.12.033 34995482
12 Pérez-Then E. Lucas C. Monteiro V.S. Miric M. Brache V. Cochon L. Vogels C.B.F. Malik A.A. De la Cruz E. Jorge A. Neutralizing antibodies against the SARS-CoV-2 Delta and Omicron variants following heterologous CoronaVac plus BNT162b2 booster vaccination Nat. Med. 28 2022 481 485 10.1038/s41591-022-01705-6 35051990
13 Cao Y. Wang J. Jian F. Xiao T. Song W. Yisimayi A. Huang W. Li Q. Wang P. An R. Omicron escapes the majority of existing SARS-CoV-2 neutralizing antibodies Nature 602 2022 657 663 10.1038/s41586-021-04385-3 35016194
14 Cele S. Jackson L. Khoury D.S. Khan K. Moyo-Gwete T. Tegally H. San J.E. Cromer D. Scheepers C. Amoako D.G. Omicron extensively but incompletely escapes Pfizer BNT162b2 neutralization Nature 602 2022 654 656 10.1038/s41586-021-04387-1 35016196
15 Liu L. Iketani S. Guo Y. Chan J.F.-W. Wang M. Liu L. Luo Y. Chu H. Huang Y. Nair M.S. Striking antibody evasion manifested by the Omicron variant of SARS-CoV-2 Nature 602 2022 676 681 10.1038/s41586-021-04388-0 35016198
16 Dolton G. Rius C. Hasan M.S. Wall A. Szomolay B. Behiry E. Whalley T. Southgate J. Fuller A. Morin T. COVID-19 Genomics UK COG-UK consortium Emergence of immune escape at dominant SARS-CoV-2 killer T-cell epitope Cell 185 2022 2936 2951.e19 10.1016/j.cell.2022.07.002 35931021
17 Willett B.J. Grove J. MacLean O.A. Wilkie C. De Lorenzo G. Furnon W. Cantoni D. Scott S. Logan N. Ashraf S. SARS-CoV-2 Omicron is an immune escape variant with an altered cell entry pathway Nat. Microbiol. 7 2022 1161 1179 10.1038/s41564-022-01143-7 35798890
18 Hui K.P.Y. Ho J.C.W. Cheung M.C. Ng K.C. Ching R.H.H. Lai K.L. Kam T.T. Gu H. Sit K.-Y. Hsin M.K.Y. SARS-CoV-2 Omicron variant replication in human bronchus and lung ex vivo Nature 603 2022 715 720 10.1038/s41586-022-04479-6 35104836
19 Meng B. Abdullahi A. Ferreira I.A.T.M. Goonawardane N. Saito A. Kimura I. Yamasoba D. Gerber P.P. Fatihi S. Rathore S. Altered TMPRSS2 usage by SARS-CoV-2 Omicron impacts infectivity and fusogenicity Nature 603 2022 706 714 10.1038/s41586-022-04474-x 35104837
20 ECDC Country Overview Report: Week 27 2022 2022 https://www.ecdc.europa.eu/en/covid-19/country-overviews
21 Gruell H. Vanshylla K. Weber T. Barnes C.O. Kreer C. Klein F. Antibody-mediated neutralization of SARS-CoV-2 Immunity 55 2022 925 944 10.1016/j.immuni.2022.05.005 35623355
22 Dougan M. Nirula A. Azizad M. Mocherla B. Gottlieb R.L. Chen P. Hebert C. Perry R. Boscia J. Heller B. Bamlanivimab plus Etesevimab in mild or moderate covid-19 N. Engl. J. Med. 385 2021 1382 1392 10.1056/nejmoa2102685 34260849
23 Gupta A. Gonzalez-Rojas Y. Juarez E. Crespo Casal M. Moya J. Falci D.R. Sarkis E. Solis J. Zheng H. Scott N. Early treatment for covid-19 with SARS-CoV-2 neutralizing antibody Sotrovimab N. Engl. J. Med. 385 2021 1941 1950 10.1056/nejmoa2107934 34706189
24 Weinreich D.M. Sivapalasingam S. Norton T. Ali S. Gao H. Bhore R. Xiao J. Hooper A.T. Hamilton J.D. Musser B.J. REGEN-COV antibody combination and outcomes in outpatients with covid-19 N. Engl. J. Med. 385 2021 e81 10.1056/nejmoa2108163 34587383
25 Levin M.J. Ustianowski A. De Wit S. Launay O. Avila M. Templeton A. Yuan Y. Seegobin S. Ellery A. Levinson D.J. Intramuscular AZD7442 (Tixagevimab–Cilgavimab) for prevention of covid-19 N. Engl. J. Med. 386 2022 2188 2200 10.1056/nejmoa2116620 35443106
26 O’Brien M.P. Forleo-Neto E. Musser B.J. Isa F. Chan K.-C. Sarkar N. Bar K.J. Barnabas R.V. Barouch D.H. Cohen M.S. Subcutaneous REGEN-COV antibody combination to prevent covid-19 N. Engl. J. Med. 385 2021 1184 1195 10.1056/nejmoa2109682 34347950
27 Planas D. Saunders N. Maes P. Guivel-Benhassine F. Planchais C. Buchrieser J. Bolland W.-H. Porrot F. Staropoli I. Lemoine F. Considerable escape of SARS-CoV-2 Omicron to antibody neutralization Nature 602 2022 671 675 10.1038/s41586-021-04389-z 35016199
28 Yu J. Collier A.R.Y. Rowe M. Mardas F. Ventura J.D. Wan H. Miller J. Powers O. Chung B. Siamatu M. Neutralization of the SARS-CoV-2 omicron BA.1 and BA.2 variants N. Engl. J. Med. 386 2022 1579 1580 10.1056/nejmc2201849 35294809
29 Bruel T. Hadjadj J. Maes P. Planas D. Seve A. Staropoli I. Guivel-Benhassine F. Porrot F. Bolland W.-H. Nguyen Y. Serum neutralization of SARS-CoV-2 Omicron sublineages BA.1 and BA.2 in patients receiving monoclonal antibodies Nat. Med. 28 2022 1297 1302 10.1038/s41591-022-01792-5 35322239
30 NIH NIH COVID-19 treatment guidelines, anti-SARS-CoV-2 antibody products, summary recommendations Anti-SARS-CoV-2 Antibody Products https://www.covid19treatmentguidelines.nih.gov/therapies/anti-sars-cov-2-antibody-products/summary-recommendations/ 2022
31 Benotmane I. Velay A. Gautier-Vargas G. Olagne J. Thaunat O. Fafi-Kremer S. Caillard S. Pre-exposure prophylaxis with 300 mg Evusheld elicits limited neutralizing activity against the Omicron variant Kidney Int. 102 2022 442 444 10.1016/j.kint.2022.05.008 35618094
32 Stuver R. Shah G.L. Korde N.S. Roeker L.E. Mato A.R. Batlevi C.L. Chung D.J. Doddi S. Falchi L. Gyurkocza B. Activity of AZD7442 (tixagevimab-cilgavimab) against Omicron SARS-CoV-2 in patients with hematologic malignancies Cancer Cell 40 2022 590 591 10.1016/j.ccell.2022.05.007 35598602
33 Westendorf K. Žentelis S. Wang L. Foster D. Vaillancourt P. Wiggin M. Lovett E. van der Lee R. Hendle J. Pustilnik A. LY-CoV1404 (bebtelovimab) potently neutralizes SARS-CoV-2 variants Cell Rep. 39 2022 110812 10.1016/j.celrep.2022.110812
34 Hentzien M. Autran B. Piroth L. Yazdanpanah Y. Calmy A. A monoclonal antibody stands out against omicron subvariants: a call to action for a wider access to bebtelovimab Lancet Infect. Dis. 22 2022 00495 00499 10.1016/s1473-3099
35 Tuekprakhon A. Nutalai R. Dijokaite-Guraliuc A. Zhou D. Ginn H.M. Selvaraj M. Liu C. Mentzer A.J. Supasa P. Duyvesteyn H.M.E. Antibody escape of SARS-CoV-2 Omicron BA.4 and BA.5 from vaccine and BA.1 serum Cell 185 2022 2422 2433.e13 10.1016/j.cell.2022.06.005 35772405
36 Wang Q. Guo Y. Iketani S. Nair M.S. Li Z. Mohri H. Wang M. Yu J. Bowen A.D. Chang J.Y. Shah J.G. Nguyen N. Chen Z. Meyers K. Yin M.T. Sobieszczyk M.E. Sheng Z. Huang Y. Liu L. Ho D.D. Antibody evasion by SARS-CoV-2 Omicron subvariants BA.608.1, BA.4, and BA.5 Nature 2022 603 608 10.1038/s41586-022-05053-w
37 Arora P. Nehlmeier I. Kempf A. Cossmann A. Schulz S.R. Dopfer-Jablonka A. Baier E. Tampe B. Moerer O. Dickel S. Augmented neutralisation resistance of emerging omicron subvariants BA.2.12.1, BA.4, and BA.5 Lancet Infect. Dis. 22 2022 1537 1538 10.1016/s1473-3099 36116462
38 Cao Y. Yisimayi A. Jian F. Song W. Xiao T. Wang L. Du S. Wang J. Li Q. Chen X. BA.2.12.1, BA.4 and BA.5 escape antibodies elicited by Omicron infection Nature 608 2022 593 602 10.1038/s41586-022-04980-y 35714668
39 Yamasoba D. Kosugi Y. Kimura I. Fujita S. Uriu K. Ito J. Sato K. Genotype to Phenotype Japan G2P-Japan Consortium Neutralisation sensitivity of SARS-CoV-2 omicron subvariants to therapeutic monoclonal antibodies Lancet Infect. Dis. 22 2022 942 943 10.1016/s1473-3099(22)00365-6 35690075
40 Takashita E. Yamayoshi S. Simon V. van Bakel H. Sordillo E.M. Pekosz A. Fukushi S. Suzuki T. Maeda K. Halfmann P. Efficacy of antibodies and antiviral drugs against omicron BA.2.12.1, BA.4, and BA.5 subvariants N. Engl. J. Med. 387 2022 468 470 10.1056/nejmc2207519 35857646
41 Touret F. Baronti C. Pastorino B. Villarroel P.M.S. Ninove L. Nougairède A. Lamballerie X. de In vitro activity of therapeutic antibodies against SARS-CoV-2 Omicron BA.1, BA.2 and BA.5 Sci Rep-uk 12 2022 12609 10.1038/s41598-022-16964-z
42 Aggarwal A. Akerman A. Milogiannakis V. Silva M.R. Walker G. Kidinger A. Angelovich T. Waring E. Amatayakul-Chantler S. Roth N. SARS-CoV-2 Omicron BA.5: evolving tropism and evasion of potent humoral responses and resistance to clinical immunotherapeutics relative to viral variants of concern Preprint at medRxiv 2022 10.1101/2022.07.07.22277128 ▪▪▪–▪▪▪
43 Yamin R. Jones A.T. Hoffmann H.-H. Schäfer A. Kao K.S. Francis R.L. Sheahan T.P. Baric R.S. Rice C.M. Ravetch J.V. Bournazos S. Fc-engineered antibody therapeutics with improved anti-SARS-CoV-2 efficacy Nature 599 2021 465 470 10.1038/s41586-021-04017-w 34547765
44 Winkler E.S. Gilchuk P. Yu J. Bailey A.L. Chen R.E. Chong Z. Zost S.J. Jang H. Huang Y. Allen J.D. Human neutralizing antibodies against SARS-CoV-2 require intact Fc effector functions for optimal therapeutic protection Cell 184 2021 1804 1820.e16 10.1016/j.cell.2021.02.026 33691139
45 Lu L.L. Suscovich T.J. Fortune S.M. Alter G. Beyond binding: antibody effector functions in infectious diseases Nat. Rev. Immunol. 18 2018 46 61 10.1038/nri.2017.106 29063907
46 Lee W.S. Wheatley A.K. Kent S.J. DeKosky B.J. Antibody-dependent enhancement and SARS-CoV-2 vaccines and therapies Nat. Microbiol. 5 2020 1185 1191 10.1038/s41564-020-00789-5 32908214
47 Dufloo J. Grzelak L. Staropoli I. Madec Y. Tondeur L. Anna F. Pelleau S. Wiedemann A. Planchais C. Buchrieser J. Asymptomatic and symptomatic SARS-CoV-2 infections elicit polyfunctional antibodies Cell Rep. Med. 2 2021 100275 10.1016/j.xcrm.2021.100275
48 Regeneron Fact sheet for health care providers emergency use authorization (EUA) of REGEN-Cov® (Casirivimab and Imdevimab). 2022 Regeneron https://www.regeneron.com/downloads/treatment-covid19-eua-fact-sheet-for-hcp.pdf
49 Grzelak L. Velay A. Madec Y. Gallais F. Staropoli I. Schmidt-Mutter C. Wendling M.-J. Meyer N. Planchais C. Rey D. Sex differences in the evolution of neutralizing antibodies to SARS-CoV-2 J. Infect. Dis. 224 2021 983 988 10.1093/infdis/jiab127 33693749
50 Lempp F.A. Soriaga L.B. Montiel-Ruiz M. Benigni F. Noack J. Park Y.-J. Bianchi S. Walls A.C. Bowen J.E. Zhou J. Lectins enhance SARS-CoV-2 infection and influence neutralizing antibodies Nature 598 2021 342 347 10.1038/s41586-021-03925-1 34464958
51 Schoofs T. Klein F. Braunschweig M. Kreider E.F. Feldmann A. Nogueira L. Oliveira T. Lorenzi J.C.C. Parrish E.H. Learn G.H. HIV-1 therapy with monoclonal antibody 3BNC117 elicits host immune responses against HIV-1 Science 352 2016 997 1001 10.1126/science.aaf0972 27199429
52 Pelegrin M. Naranjo-Gomez M. Piechaczyk M. Antiviral monoclonal antibodies: can they Be more than simple neutralizing agents? Trends Microbiol. 23 2015 653 665 10.1016/j.tim.2015.07.005 26433697
53 Schaefer-Babajew D. Wang Z. Muecksch F. Cho A. Raspe R. Johnson B. Canis M. DaSilva J. Ramos V. Turroja M. Antibody feedback regulation of memory B cell development in SARS-CoV-2 mRNA vaccination Preprint at medRxiv 2022 10.1101/2022.08.05.22278483
54 Pinto D. Park Y.-J. Beltramello M. Walls A.C. Tortorici M.A. Bianchi S. Jaconi S. Culap K. Zatta F. De Marco A. Cross-neutralization of SARS-CoV-2 by a human monoclonal SARS-CoV antibody Nature 583 2020 290 295 10.1038/s41586-020-2349-y 32422645
55 Martin-Blondel G. Marcelin A.-G. Soulié C. Kaisaridi S. Lusivika-Nzinga C. Dorival C. Nailler L. Boston A. Melenotte C. Cabié A. Sotrovimab to prevent severe COVID-19 in high-risk patients infected with Omicron BA.2 J. Infect. 85 2022 e104 e108 10.1016/j.jinf.2022.06.033 35803386
56 Beaudoin-Bussières G. Chen Y. Ullah I. Prévost J. Tolbert W.D. Symmes K. Ding S. Benlarbi M. Gong S.Y. Tauzin A. A Fc-enhanced NTD-binding non-neutralizing antibody delays virus spread and synergizes with a nAb to protect mice from lethal SARS-CoV-2 infection Cell Rep. 38 2022 110368 10.1016/j.celrep.2022.110368 35123652
57 Cao Y. Yu Y. Song W. Jian F. Yisimayi A. Yue C. Feng R. Wang P. Yu L. Zhang N. Neutralizing antibody evasion and receptor binding features of SARS-CoV-2 Omicron BA.2.75 Preprint at bioRxiv 2022 10.1101/2022.07.18.500332
58 Yamasoba D. Kimura I. Kosugi Y. Uriu K. Fujita S. Ito J. Sato K. Consortium, T.G. to P.J. (G2P-J) Neutralization sensitivity of Omicron BA.2.75 to therapeutic monoclonal antibodies Preprint at bioRxiv 2022 10.1101/2022.07.14.500041
59 Sheward D.J. Kim C. Fischbach J. Muschiol S. Ehling R.A. Björkström N.K. Hedestam G.B.K. Reddy S.T. Albert J. Peacock T.P. Evasion of neutralizing antibodies by Omicron sublineage BA.2.75 Preprint at bioRxiv 2022 10.1101/2022.07.19.500716
60 Buchrieser J. Dufloo J. Hubert M. Monel B. Planas D. Rajah M.M. Planchais C. Porrot F. Guivel-Benhassine F. Van der Werf S. Syncytia formation by SARS-CoV-2-infected cells Embo J 39 2020 10.15252/embj.2020106267 e2020106267
61 Hadjadj J. Planas D. Ouedrani A. Buffier S. Delage L. Nguyen Y. Bruel T. Stolzenberg M.-C. Staropoli I. Ermak N. Immunogenicity of BNT162b2 vaccine against the Alpha and Delta variants in immunocompromised patients with systemic inflammatory diseases Ann. Rheum. Dis. 81 2022 720 728 10.1136/annrheumdis-2021-221508 annrheumdis-2021-221508 35022159
62 Pelleau S. Woudenberg T. Rosado J. Donnadieu F. Garcia L. Obadia T. Gardais S. Elgharbawy Y. Velay A. Gonzalez M. Kinetics of the severe acute respiratory syndrome Coronavirus 2 antibody response and serological estimation of time since infection J. Infect. Dis. 224 2021 1489 1499 10.1093/infdis/jiab375 34282461
63 Sterlin D. Mathian A. Miyara M. Mohr A. Anna F. Claër L. Quentric P. Fadlallah J. Devilliers H. Ghillani P. IgA dominates the early neutralizing antibody response to SARS-CoV-2 Sci. Transl. Med. 13 2021 10.1126/scitranslmed.abd2223 eabd2223
| 36450283 | PMC9706550 | NO-CC CODE | 2022-12-01 23:19:46 | no | Cell Rep Med. 2022 Nov 17;:100850 | utf-8 | Cell Rep Med | 2,022 | 10.1016/j.xcrm.2022.100850 | oa_other |
==== Front
Methods
Methods
Methods (San Diego, Calif.)
1046-2023
1095-9130
Elsevier Inc.
S1046-2023(22)00245-6
10.1016/j.ymeth.2022.11.004
Article
A lightweight network for COVID-19 detection in X-ray images
Shi Yong a
Tang Anda b
Xiao Yang c
Niu Lingfeng a⁎
a Research Center on Fictitious Economy and Data Science, Chinese Academy of Sciences, Beijing, 100190 China
b School of Mathematical Sciences, University of Chinese Academy of Sciences, Beijing, 100190, China
c Faculty of Information Technology, Beijing University of Technology, Beijing, China
⁎ Corresponding author.
29 11 2022
29 11 2022
23 11 2022
© 2022 Elsevier Inc. All rights reserved.
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
The Novel Coronavirus 2019 (COVID-19) is a global pandemic which has a devastating impact. Due to its quick transmission, a prominent challenge in confronting this pandemic is the rapid diagnosis. Currently, the commonly-used diagnosis is the specific molecular tests aided with the medical imaging modalities such as chest X-ray (CXR). However, with the large demand, the diagnoses of CXR are time-consuming and laborious. Deep learning is promising for automatically diagnosing COVID-19 to ease the burden on medical systems. At present, the most applied neural networks are large, which hardly satisfy the rapid yet inexpensive requirements of COVID-19 detection. To reduce huge computation and memory demands, in this paper, we focus on implementing lightweight networks for COVID-19 detection in CXR. Concretely, we first augment data based on clinical visual features of CXR from expertise. Then, according to the fact that all the input data are CXR, we design a targeted four-layer network with either 11×11 or 3×3 kernels to recognize regional features and detail features. A pruning criterion based on the weights importance is also proposed to further prune the network. Experiments on a public COVID-19 dataset validate the effectiveness and efficiency of the proposed method.
Keyword
COVID-19 detection
neural network
network pruning
==== Body
pmc1 Introduction
The Novel Coronavirus 2019 (COVID-19) is a public health-threatening virus that may infect the upper respiratory tract and lungs, even progress to multi-organ dysfunction. It spreads rapidly to become a global pandemic that the numbers of cases and deaths have increased on a daily basis [1]. At present, there is no effective cure for the COVID-19 disease, so an essential step in the combat against this virus is to effectively screen infected patients. Therefore the infected patients can receive in-time care and treatment, as well as be isolated to mitigate the spread [2].
Currently, the commonly-used diagnosis for COVID-19 is the specific molecular tests on respiratory samples, such as throat swab [3], aided with the medical imaging modalities, such as chest X-ray (CXR) images [4], [5], [6]. However, compared with the specific molecular tests, because there are relatively few medical professionals and experienced radiologists, the diagnoses for CXR images are more time-consuming and laborious. As the spreading of COVID-19 is fast, there is a need for a quick yet accurate automated mechanism to assist medical personnel to diagnose COVID-19 in CXR images. For this situation, COVID-19 detection with the help of deep learning model attracts a lot of attention [7], [8], [9], [10].
Currently, available COVID-19 CXR images are usually limited. As a result, the samples may not be sufficient for training deep models. To solve these problems, Minaee et al. [9] use the pretrained models as a feature extractor and train a classifier on top of them. Basu et al. [2] train AlexNet, VGG-16 and ResNet-50 based on domain extension transfer learning. Ismael et al. [10] utilize transfer learning on pretrained deep networks to classify COVID-19 and normal CXR images. These works employ transfer learning by leveraging sophisticated deep networks and achieve good performances. However, the models are too complex to to be deployed in large-scale emergency detection of the medical field. Are there some easy-to-implement networks that have comparable performances in COVID-19 detection? To actualize the above target, in this work, we intend to design relative simpler networks for COVID-19 detection.
Another challenge is that deep models demand huge computation and memory resources. Until now, various networks have been proposed to make automated COVID-19 detection such as VGG [11], ResNet [12] or other modified networks with many layers [7], [8]. These networks are too large to be applied for real-time COVID-19 detection. Considering the efficiency of the networks, some studies propose utilizing light network design such as EfficientNet and deep SqueezeNet [13], [14], [15]. These works achieve good accuracy and efficiency but limited on using the developed network or modified version. These models generally used to tackle with complicated datasets from multi-scenarios. Considering the fact that, the data sources are only the CXR images in this task, it is not necessary to use these sophisticated networks. Therefore, at this paper, instead of adopting some off-the-shelf networks directly, we intend to design more targeted but simpler networks based on the principle of ‘model capacity-model complexity’ balance.
In this paper, we propose novel lightweight networks for automatically detecting the COVID-19 in CXR images. Specifically, to solve the problems caused by insufficient data, we first conduct data augmentation based on clinical visual features extraction of CXR from expertise. In order to enhance the tendency that the model extract clinical visual features, such as interstitial abnormalities, ground-glass opacity or lesions in a particular area, we augment the images by flipped, cropped, noise added and whiten versions. And then, considering the lightweight need, we design a four-layer network architecture based on the principle of ‘model capacity-model complexity’ balance. To better extract features, we design the network architecture pointed at the CXR such as kernel size design. We utilize 11×11 kernels to extract regional features such as lesions in a relative large area, and use 3×3 kernels to recognize detail features such as interstitial abnormalities.
To further reduce the network complexity, after each iteration of network training, we prune the network by a specific pruning criterion which is based on four principles, including threshold effect, value decay modification, importance separation and sign-preserving property. Finally, the lightweight network is obtained by conducting the training and the pruning alternately until the network convergence. To validate the performance of the proposed method, we conduct experiments on the public COVID-19 dataset. The experimental results show that our proposed lightweight networks can achieve more than 95% accuracy and more than 98% recall on COVID-19 detection. In addition, the lightweight networks reduce about 95% of the parameters compared with the designed network prototype. In brief, the main contributions of this paper are summarized as follows.1) We propose a lightweight network for automatically detecting the COVID-19 in CXR images.
2) To solve the problem caused by insufficient data, we conduct data augmentation based on clinical visual features of CXR from expertise.
3) To extract features in the CXR more effectively and efficiently, we design a more targeted but simpler network architecture .
4) To further compress the network, we give a pruning criterion by which the weights can be pruned and modified according to the weights importance.
The remainder of the paper is structured as follows. We review the literature concerning neural networks and lightweight networks employed for COVID-19 detection in the CXR images in Section 2. In Section 3, we introduce our methodology in detail. We carry out extensive experiments to evaluate the performance of our method in Section 4. Finally, Section 5 summarizes this paper.
2 Literature review
2.1 Neural networks for COVID-19 detection in CXR
CXR images have been proven an attractive option for the COVID-19 detection since its wide availability in diverse health care settings [16], [17], [11]. Motivated by the need for faster interpretation of the radiography images, a number of deep learning based methods have been proposed. Wang et al. [7] propose the COVID-Net tailored for the COVID-19 detection from CXR images. Kaur et al. [18] propose a metaheuristic deep COVID-19 screening model based on modified AlexNet for CXR. Ozturk et al. [8] implement the DarkNet with seventeen convolutional layer and different filters. Bukhari et al. [12] apply ResNet-50 network architectures to diagnose the cases of COVID-19 infections. Hemdan et al. [11] utilize seven different deep architectures to identify the COVID-19 infection. To solve the problem of inadequate COVID-19 data, some works employ transfer learning or pretrained model. Minaee et al. [9] train ResNet-18, ResNet-50, SqueezeNet, and DenseNet-121 via transfer learning to identify COVID-19 disease in the CXR. Basu et al. [2] employ domain extension transfer learning to train AlexNet, VGG-16 and ResNet-50 to predict COVID-19 from CXR. Apostolopoulos et al. [19] evaluate the performance of various networks with transfer learning and achieve remarkable results over the detection of various abnormalities in small medical image datasets. Ismael et al. [10] utilize transfer learning on pretrained deep networks to classify COVID-19 and normal CXR. More and more results have shown that deep learning based methods have promising performances in COVID-19 detection.
2.2 Lightweight networks for COVID-19 detection in CXR
Currently, most methods use deep neural networks for COVID-19 detection in CXR images. These networks have enormous amount of parameters so that they demand huge computation and storage. Considering the efficiency of the networks, lightweight networks are investigated to detect COVID-19 viral infection. Ucar et al. [13] propose a deep Bayes-SqueezeNet based COVID-19 diagnosis where the SqueezeNet comes forward for its light network design. They conduct hyperparameters fine-tuning and data augmentation to make the proposed network obtain an out-performed COVID-19 diagnosis accuracy. Luz et al. [14] propose a new family of models based on the EfficientNet architecture [20] for COVID-19 screening in CXR images. Chaudhary et al. [15] propose to utilize EfficientNet to surpass COVID-net. They focus on models EfficientNet-B0 to EfficientNet-B3 and find that EfficientNet-B1 generates best accuracy and efficiency. Rajaraman et al. [21] utilize a pruned deep learning model for COVID-19 detection via iteratively increasing pruning percentage and achieve nearly half the compression. To our knowledge, although some efficient networks achieve good performance for COVID-19 detection in the CXR images, the networks are developed networks, or modified on them. Considering the fact that in this task of COVID-19 detection, the data source is only the CXR images rather than complicated datasets from multi-scenarios, it is not necessary to use these sophisticated networks. Therefore, at this paper, instead of adopting some off-the-shelf networks directly, we intend to design more targeted and lightweight networks.
3 Methodology
3.1 Overall description of the proposed method
In this work, based on each given CXR image, we intend to make one of the three predictions such as (1) the patient is not infected; (2) the patient is infected with COVID-19; (3) the patient is infected with other types of viral pneumonia. Therefore, the model can not only help identifying whether the patient is infected, but also determine whether the patient is infected with COVID-19 or other types of viral pneumonia so that the clinicians can choose specific treatment strategies depending on the different cause of infection. For brevity, we denote these three categories with the label ‘Normal’, ‘COVID-19’ and ‘viral pneumonia’ respectively. We train lightweight networks to automatically identifying the three classes. The Fig. 1 depicts the overall workflow of our proposed method.Fig. 1 The framework of the proposed lightweight network for the COVID-19 classification in CXR images.
In the first step, we conduct data preprocessing for the images of the three classes. Specifically, we use data augmentation to enlarge the training dataset. And then, to make the data suitable as inputs for the neural network, we convert the images into proper size and shuffle them. And we convert labels into vectors to indicate the case of ‘Normal’, ‘COVID-19’ and ‘viral pneumonia’ for each image. In the second step, we design two network architecture prototypes based on the principle of ‘model capacity-model complexity’ balance. Concretely, we choose the proper kernel sizes to better extract patterns and features from images while we constrain the layer number of the network to keep efficiency. In the third step, we conduct network training and compression. In particular, we update the weights by optimizing the loss function of the network. And then we prune weights based on a proposed non-structural pruning criterion. We alternately conduct training and pruning until convergence. We choose the final model based on over-fitting minimization and the satisfying compression ratio. Specifically, if the validation accuracy results under different compression ratios do not vary much, we choose the network with the highest compression ratio. We also set a value to control the maximum epochs. If the epoch number reach this value, we will stop the training process. Finally, we get the lightweight network based on over-fitting minimization and the satisfying sparsity. After achieving the final lightweight network, we can detect COVID-19 by implement classification on the images. According to the ‘softmax’ values of the three categories outputted by the lightweight network, we predict the image as the category with the largest ‘softmax’ value.
3.2 Data augmentation
The sample complexity refers to the number of training data that a model needs in order to be successfully trained. The number of training data is related to the amount of information provided by the entire training sample set. The sample quality can strongly affect the generalization ability of neural networks. Currently, available COVID-19 CXR images are usually limited so that the sample complexity may not be sufficient for neural networks. To solve this problem, we utilize data augmentation [22] which applies transformations to the images. It enlarges the training dataset so that there will be adequate data inputs for our model. Specifically, the transformations are chosen to the images including (1) flip: we flip the original images horizontally to obtain the mirrored version images; (2) crop: we cut off 10% to 12.5% around the original images to retain the main part of the chest in the images; (3) noise: we add zero-mean Gaussian noise to images; (4) whitening: we eliminate the correlations between different pixels of the image, and keep the variance of each pixel of the image as one.
Compared with distinction between the features of infected and uninfected patients, it is more difficult to distinguish the features of between ’COVID-19’ and ’viral pneumonia’ because they both have lesions in the lungs. Former studies prove that bilateral infiltrates, interstitial abnormalities and ground-glass opacity are visible in the ‘COVID-19’ CXR images [1], [13], [7], [23]. In contrast, ‘viral pneumonia’ CXR images shows lesions along the direction of trachea and bronchus. These features are different. In order to enhance the tendency that the model extract these features, we make these features more prominent by removing the redundancy in the image as much as possible. In detail, we conduct crop on the raw images to retain the part of the lung. We also conduct zero-phase component analysis whitening [24] to process the images in this work. We first calculate the mean of each feature among the images in one class rather than the pixel means in one image. And then we calculate the standard deviation vector along each feature. We conduct mean normalization and standardization of images and calculate the singular value decomposition of the covariance matrix. Finally, we perform the whitening operation according to the equation in the work [24]. Some instances of data augmentation are shown in Fig. 1. It is noticed that we only implement augmentation on the training data except the test data.
3.3 Network architecture prototype
Most methods tend to use sophisticated networks that are generally used to tackle with complicated dataset from multi-scenarios. It is not necessary to employ these networks for the task where the dataset are all CXR. Therefore, we design a network architecture based on heuristic hypotheses and practices. In detail, considering the need of lightweight design and sample complexity, we do not choose too deep network in this work. The capacity of a four-layer network is enough for the initial prototype. To sufficiently learn the patterns of images, we use two convolutional layers as first two layer to fulfill feature extraction. And to conduct diversified feature transformation, we choose two fully connected layers with adequate connections as the third and fourth layer.
The diagnosis for the radiography examination is generally based on clinical visual indicators. The clinical visual indicators include bilateral infiltrates, interstitial abnormalities, ground-glass opacity and lesions in a particular area [1], [13], [7], [23]. We think that the lesions in lung show regional features in the CXR image. ‘COVID-19’ and ‘viral pneumonia’ may show lesions in different area of the lung. And the visual indicator such as ground-glass opacity show detail and local features in the CXR image. These detail features of CXR images in different classes may be distinct. Under this assumption, we propose to either use a larger convolution kernel size, such as 11×11, to distinguish COVID-19 by lesions in specific areas, or use a smaller convolution kernel size, such as 3×3, to make classification by detailed features. Besides, since the decisive features may appears in a specific area and this specific area cannot be determined without prior knowledge of professionals, in the feature extraction phase, we set the stride of convolution kernel step as one so that it will extract features of each area in the CXR images. According to these concept, we have designed two network architecture with different convolution kernel sizes, called ‘Conv4net-S’ and ‘Conv4net-L’ respectively. The details of the network architectures of ‘Conv4net-S’ and ‘Conv4net-L’ is given in Table 1 .Table 1 Details of the designed networks (‘conv-i’ denotes the convolution architecture in the i-th layer. ‘fc-i’ denotes the fully-connected architecture in the i-th layer. ‘bn’ denotes the batch-normalization. ‘mp’ denotes the max-pool with the hyper-parameter α=0.2. ‘ac’ denotes the activation function which is leaky-relu.).
network architecture size stride padding FLOP Para
Conv4net-L conv-1 [11, 11, 3, 64] [1, 1, 1, 1] ‘SAME’ 1,197,121,728 23,232
ac +bn
mp [1, 2, 2, 1] [1, 2, 2, 1] ‘SAME’ – –
conv-2 [7, 7, 64, 64] [1, 1, 1, 1] ‘SAME’ 2,608,349,184 200,704
ac+bn
mp [1, 2, 2, 1] [1, 2, 2, 1] ‘SAME’ – –
fc-3 [57×57×64, 1024] – – 212,926,464 212,926,464
dropout
ac
fc-4 [1024,3] – – 3,072 3,072
Conv4net-S conv-1 [3, 3, 3, 64] [1, 1, 1, 1] ‘SAME’ 89,042,112 1,728
ac +bn
mp [1, 2, 2, 1] [1, 2, 2, 1] ‘SAME’ – –
conv-2 [3, 3, 64, 64] [1, 1, 1, 1] ‘SAME’ 479,084,544 36,864
ac +bn
mp [1, 2, 2, 1] [1, 2, 2, 1] ‘SAME’ – –
fc-3 [57×57×64, 1024] – – 212,926,464 212,926,464
dropout+ac
fc-4 [1024,3] – – 3,072 3,072
3.4 Network training and compression
In this stage, the final network architecture will be generated by a proposed strategy where data, the initial network prototype along with some specific criteria act as a guide to prune and modify the weights of the network. The final lightweight networks based on ‘Conv4net-S’ and ‘Conv4net-L’ are denoted as ‘Light4net-S’ and ‘Light4net-L’ respectively.
We first trains the network based on data fidelity. After one iteration of weights update, we intend to zero out some weights by a non-structural pruning criterion. We design the criterion based on the following principles. First, we think that the smaller the weight value is, the less important this weight is. Therefore, we design a criterion to zero-out weights based on the weight magnitude where a weight will be set to zero if its value is below a given threshold. Second, we think that the preserved weights should be easier to be zeroed in next iteration. Thus we decrease the value of weights that are not zeroed so that they tend to be less than threshold in next iteration. Third, among the preserved weights, the important ones and the secondary ones should be further separated . We use modification when decreasing the weights so that the value reduction of the less important weights should be larger than that of the important weights. Last but not the least, we do not want the weights to change its positive and negative after reduction and correction. Specifically, we design the criterion under these principles in Eq. (1).(1) wt+1(l)=sign(wt(l))|wt(l)|-(μl)3/2k|wtl|,if|wt(l)|-μl⩾00,otherwise
where the subscript t denotes the t-th iteration, the superscript (l) denotes the l-th layer, wt(l) denotes each element Wg,i(l) of the weight matrix Wt(l),k>1 is the hyper-parameter to control the value reduction of the weight wt(l),μl>0 is the threshold hyper-parameter of the l-th weight-layer. In Eq. (1), if a weight |wt|<μl, this weight will be estimated to zero, which satisfies the first principle. If a weight |wt(l)|⩾μl, there is an operation as |wt(l)|-(μl)3/2k|wtl| which is used to shrink the weight wt(l). It satisfies the second principle. Besides, the term (μl)3/2k|wtl| can be seen as 1kμl|wtl|μl. It means that when |wt(l)|=μl, the value reduction of |wt(l)| is equal to μlk. When |wt(l)|>μl, the value reductions are smaller than μlk. Depending on the weights importance, the value reductions will vary. Since μl|wtl| will be decreasing with the increasing of the weight, the value reduction (μl)3/2k|wtl| will be smaller if the weight is more important. And if the weight is less important, μl|wtl|will be larger, so that the value reduction will be comparably larger. It controls different drops according to the weight magnitude which satisfies the third principle. In addition, since |wt(l)|-μl⩾0,|wt(l)|-(μl)3/2k|wtl|⩾0. Thus sign(wt(l)) keeps the positive and negative of the weights.
To indicate this function clearly, we visualize the curve of the function (1) in Fig. 2 . We plot the curves of the function with different μ-k pairs. And we plot the curve of y=x as a reference to display the value reduction effect of our proposed function. As seen in this figure, μ is the threshold which controls the compression level. With the value of μ increasing, more weights tend to be pruned. It means that the larger the μl value is, the more significant the compression effect in l-th layer is. k controls the value reduction of the weight. The larger the value of k is, the fewer value reduction of the weight produces.Fig. 2 The curve of the function (1).
4 Experiments
4.1 Datasets
We utilize a COVID-19 dataset of CXR images from publicly available data repository called COVID-19 radiography database [25]. This database comes from a group of researchers from Qatar University in Doha, Qatar and Dhaka University in Bangladesh, as well as collaborators from Pakistan and Malaysia, who cooperated with doctors to establish a database of CXR images for covid-19 positive cases, as well as normal and viral pneumonia images. All the images are in portable network graphics (PNG) file format and the resolution is 1024×1024 pixels, which can be easily converted to 224×224 or 227×227 pixels typically required by the popular convolutional neural networks. In this dataset, there are 219 COVID-19 positive images along with 1341 normal images and 1345 viral pneumonia images. Obviously, the phenomenon of data imbalance exists that the images of class ‘COVID-19’ is significantly fewer than that of the other two classes. In the research [26], the authors indicate that the data imbalance of the class distribution has a detrimental effect on the classification performance of the models. Learning with an unbalanced dataset could bias the prediction model towards the classes with more samples, leading to inferior classification models. To overcome this issue, we find extra CXR images of COVID-19 patients 1 to augment the raw dataset. After that, we randomly selected around 10% of the images from each category, with a total of 400 samples as the validation set. In the validation set, there are 141 ‘COVID-19’ samples, 134 ‘Normal’ samples and 125 ‘Viral Pneumonia’ samples.
4.2 Evaluation criterion
In this work, the disease detection performance of the model is quantitatively measured via several metrics. Accuracy(Acc) is the basic measurement to evaluate the classification performance. Besides the metric accuracy, we use some specific metrics like precision (Prec) , recall (Rec) , F1 score (F1) and Matthew Correlation Coefficient (MCC) which are the convincing evaluations in the medical field. Precision refers to the correct proportion of the positive prediction. Recall refers to the correct proportion of the predictions among the samples with positive ground truth. F1-score is the harmonic mean of the precision and the recall. MCC are a statistic from the confusion matrix to measure the quality of the model.(2) Acc=(TP+TN)/(TP+FP+TN+FN)
(3) Prec=TP/(TP+FP)
(4) Rec=TP/(TP+FN)
(5) F1=2×Prec×RecPrec+Rec
(6) MCC=(TP×TN)-(FP×FN)(TP+FP)×(TP+FN)×(TN+FP)×(TN+FN)
where TP refers to the number of True Positive samples of a category. FN refers to the number of False Negative samples of a category. TN refers to the number of True Negative samples of a category. FP refers to the number of False Positive samples of a category.
In addition, we utilize the percentage of floating-point operations compared with the non-compressed network (FLOP) and the percentage of parameters used in the networks (Para) to evaluate the compression effect. FLOP indicates the computation cost of networks to some extent. Parameter used reflects the memory-consuming saving, which is an important indicator of compression effect. Generally speaking, the smaller the value of the FLOP and the parameter used are, the more efficient the model is.
4.3 Experimental results
To verify the effectiveness of the data augmentation, we perform our lightweight network ‘Light4net-S’ and ‘Light4net-L’ on the raw dataset and augmented dataset respectively. Besides, to demonstrate the advantages of our kernel size assumption, we compare our networks to a network with 5×5 kernels. We denote this network as ‘Light4net-5×5’. Except for the convolution size, other settings of ‘Light4net-5×5’ are consistent with our networks. We list the experimental results in Table 2 . The best results in terms of Acc, FLOP and Para are highlighted in the bold form in the Table 2. From the table, it can be observed that, (1) compared with the disease detection performance on the raw dataset, the performance on the augmented dataset are promoted in terms of the accuracy, precision, recall, F1 score and MCC. It verifies the effectiveness of the data augmentation. In addition, the compression ratio of each network is approximately doubled under the data augmentation. The reason may be that the data augmentation reduces the pressure of feature extraction and transformation. (2) On considering in the kernel size design, the disease detection performances of the networks with three kernel designs are comparable on the raw dataset. However, in the case of data augmentation, compared with the network with 5×5 kernels, our proposed networks both outperform with the margin of about 4% in terms of disease detection performance. The possible reason is that when data features are difficult to extract, the convolution size has little impact on the model performance. After the data is augmented, the model features are easier to extract. Thus the network needs a proper kernel size to extract features. It verifies the effectiveness of our kernel size design.Table 2 Comparison results of our proposed method.
data method(performance) class Prec(%) Rec(%) F1(%) MCC(%)
Light4net-L COVID-19 100 84.40 91.54 88.20
(Acc 92.0% FLOP 72.29 % Para 11.23 %) Normal 96.24 95.52 95.88 93.82
Viral Pneumonia 81.76 96.8 88.64 83.51
average 92.67 92.24 92.02 88.51
Light4net-S COVID-19 100 83.69 91.12 87.67
(Acc 91.75% FLOP 61.29 % Para 11.17 %) Normal 96.95 94.78 95.85 93.81
Viral Pneumonia 80.79 97.6 88.41 83.24
average 92.58 92.02 91.79 88.24
Light4net-5×5 COVID-19 100 80.85 89.41 85.57
(Acc 90.75 % FLOP 73.00 % Para 11.20 %) Normal 96.92 94.03 95.45 93.24
Viral Pneumonia 78.85 98.4 87.54 82.11
average 91.92 91.09 90.80 86.97
augmented Light4net-L COVID-19 97.22 99.30 98.25 97.29
(Acc 96.25% FLOP 64.80% Para 5.28%) Normal 97.64 92.54 95.02 92.68
Viral Pneumonia 93.80 96.8 95.28 93.10
average 96.22 96.21 96.20 94.36
Light4net-S COVID-19 98.58 98.58 98.58 97.81
(Acc 97.5% FLOP 58.63% Para 5.23%) Normal 97.04 97.76 97.40 96.08
Viral Pneumonia 96.77 96.0 96.39 94.75
average 97.46 97.45 97.45 96.21
Light4net-5×5 COVID-19 92.05 98.58 95.21 92.59
(Acc 91.75% FLOP 64.20 % Para 5.24 %) Normal 92.13 87.31 89.66 84.72
Viral Pneumonia 90.98 88.8 89.88 85.37
average 91.72 91.56 91.58 87.56
Then we focus on the precision (Prec) , recall (Rec) , F1 score (F1) and Matthew Correlation Coefficient (MCC) of methods in Table 2. Since our task is a multi-classification, we interpret the precision , recall , F1 score and MCC results of each class. In order to better display these indicators, we choose macro-average to calculate the four and list the average results in the table. To better analyze our model, We further plot the confusion matrix of our method in Fig. 3 . Overall, our method has the satisfactory classification performance in terms of the four measurement. The accuracy results of our networks are both up to 96%. The results of F1 and MCC exceeds 97%, which proves that our method obtains a stable diagnosis model. It also can be found that ‘Light4net-L’ has the best COVID-19 recall rate up to 99.3%, it means this model miss few cases of the ‘COVID-19’. This diagnosis results is critical to COVID-19 detection so that the medical systems can take rapid measures of isolation of infected patients to slow down the virus propagation. Although the ‘Light4net-L’ recall rate of the class ‘Normal’ is the comparative low, it has little bad influence in a COVID-19 detection task. This will lead to an tightening forecast for the ‘COVID-19’ warning. Since this phenomenon means that when our model classifies the ‘Normal’ category, it has a certain probability to classify it into the diseased.Fig. 3 Confusion matrices of the Conv4net-L and Conv4net-S in testing phase(‘C’ denotes the class ‘COVID-19’,‘N’ denotes the class ‘NORMAL’, ‘V’ denotes the class ‘Viral Pneumonia’).
And then we compare the performance of our lightweight network with networks using other training methods. As the performances of ‘Light4net-S’ and ‘Light4net-L’ are comparable in the Table 2, we choose the prototype network ‘Conv4net-L’ of the ‘Light4net-L’ as the basic network architecture for comparison in the following parts. We compare ‘Light4net-L’ with prototype network ‘Conv4net-L’ using normal training, the classic regularization training method ℓ2 and classic lightweight network training method ℓ1 respectively. We list the experimental results in Table 3 . The best results in terms of Acc, FLOP and Para are highlighted in the bold form in the Table 3. From the table, it can be observed that our proposed method achieves the best results in terms of Acc, FLOP and Para. Our method achieve about 20 × the compression ratio with accuracy improvement. (1) On considering in the model generalization, the accuracy of our method can outperforms the prototype network using normal training. In addition, among the method ℓ2 and the method ℓ1, our lightweight network has the best accuracy. It shows that our method has the improvement effect on the network generalization ability. (2) On considering in the efficiency of methods, compared with the prototype and the non-sparsity inducing ℓ2, our method reduce the computation amount by about 40% and reduce the parameter usage by about 95%. Compared with lightweight training method ℓ1, our network outperforms the compressed network with the margin 27.02% and 31.09% in terms of FLOP and parameter usage respectively. These results demonstrate the advantages of our proposed method.Table 3 evaluation metrics results of our proposed method on Conv4net-L.
method(performance) class Prec(%) Rec(%) F1(%) MCC(%)
Conv4net-L COVID-19 95.17 97.87 96.50 94.58
(Acc 95.50% FLOP 100 % Para 100 %) Normal 96.88 92.54 94.66 92.11
Viral Pneumonia 94.49 96.0 95.24 93.05
average 95.52 95.47 95.47 93.25
ℓ2 COVID-19 96.48 97.16 96.82 95.08
(Acc 96.0% FLOP 100 % Para 100 %) Normal 97.66 93.28 95.42 93.25
Viral Pneumonia 93.85 97.6 95.69 93.71
average 95.99 96.02 95.98 94.01
ℓ1 COVID-19 95.77 96.45 96.11 93.99
(Acc 94.50% FLOP 89.28% Para 36.36 %) Normal 96.85 91.79 94.25 91.55
Viral Pneumonia 90.84 95.2 92.97 89.72
average 94.49 94.48 94.44 91.75
Light4net-L COVID-19 97.22 99.30 98.25 97.29
(Acc 96.25% FLOP 64.80% Para 5.28%) Normal 97.64 92.54 95.02 92.68
Viral Pneumonia 93.80 96.8 95.28 93.10
average 96.22 96.21 96.20 94.36
To further analyze the network training process of our method, we report the variation of validation accuracy, training loss, FLOP and Para at the training stage in Fig. 4 . The figure plot the results per one three hundred iterations. From Fig. 4, we can see that, (1)As training proceeds, the training loss are decreasing with fluctuations. Its first cliff drop occurs about the 10 thousand iterations. With the sharp decline of the loss value, the validation accuracy has improved significantly at this moment. And then the training loss achieves a large reduction at around 13-18 thousand iterations. Consequently, the training loss converges with the validation accuracy has slight increase, which represent the model saturation. (2)FLOP and Para keep decreasing. This phenomenon validates the effect of the compression ability of our method. In general, the downtrend of FLOP decreases with the downtrend of Para. The reason for these phenomena is that the reduction of computation is based on the sparsity of weights. The computational cost becomes less and less because of the continuing reduction of weights.Fig. 4 Training loss, validation accuracy, FlOP and parameter used during training.
In addition, we report the FLOP and Para in each layer to test the effect of our method. In general, the convolution layer has a small number of parameters but a large amount of computation due to the mechanism of parameter sharing. The fully-connected layer does not have a very large amount of computation, but has a very large number of parameters and a large parameter redundancy. By tuning the μl of each layer, we can achieve various FLOP and Para level according to the requirements for the computation or storage. It also needs to consider the network performance when tuning the μl. Since the high compression ratio in the convolution layers would influence feature extraction ability and expressive power, it needs to enforce a certain compression ratio to between different layers. Former studies states that features in primary layers may be more important than the features transformed in the fully-connected layers [27]. Therefore, we mainly zero-out the weights in the third layer to achieve high compression ratio and keep the important features in this work.
We report the amount and percentage of FLOP and Para of each layer in ‘Light4net-L’ and its prototype respectively in Table 4 , where conv-1 and conv-2 represent the first and the second layer, fc-3 and fc-4 represent the third and the fourth layer. From the table, it can be observed that (1) For the prototype network, the convolution layers possess a small number of parameters accounting for 1.0% of the whole, while it produce the most of computation accounting for 95% of the whole. (2) The reduction of computation is mainly derived from the second layer accounting for 65% of the whole reduction. The reduction of parameter amount mainly comes from the third layer accounting for 99% of the whole reduction. The results coincides with the above statement.Table 4 FLOP and Para of each layer in Light4net-L
Network Layer Total
conv-1 conv-2 fc-3 fc-4
FLOP prototype 1.2×109 2.6×109 2.1×108 3.1×103 4.0 ×109
(29.79%) (64.91%) (5.30%) (<0.01%) (100%)
Light4net-L 9.0×108 1.7×109 1.1×107 2.7×103 2.6×109
(34.72%) (64.85%) (0.42%) (<0.01%) (100%)
FLOP Reduction 2.9×108 9.2×108 2.0×108 4.0×102 1.4×109
Para prototype 2.3×104 2.0×105 2.1×108 3.1×103 2.1×108
(0.01 %) (0.09 %) (99.89%) (<0.01%) (100%)
Light4net-L 1.7×104 1.3×105 1.1 ×107 2.7×103 1.1 ×107
(0.16 %) (1.16%) (98.66 %) (0.2%) (100%)
Para Reduction 5.7 ×103 7.1×104 2.0×108 4.0×102 2.0×108
5 Conclusion
A rapid diagnosis method of COVID-19 is a prominent factor to control the global pandemic situations. With the increasing demand for COVID-19 screenings, deep learning has emerged as a promising supplement to automatically diagnose COVID-19. Some limitations of the resource-demand deep networks reveal a need for lightweight networks to make a quick and accurate diagnosis of COVID-19. In this paper, we propose lightweight networks for automatically detecting the COVID-19 in CXR images. To solve the problem caused by insufficient data, we first conduct data augmentation according to the CXR clinical visual features from expertise. Then, in order to extract features in the CXR more effectively and efficiently, we design a more targeted but simpler network architecture. finally, to further compress the network, we give a pruning criterion with pruning effect, value decay effect, importance separation effect, and sign-preserving effect on weights.
To validate the performance of the proposed method, we conduct experiments on the public COVID-19 dataset. The experimental results show that our proposed lightweight networks can achieve more than 95% validation accuracy and more than 98% recall on COVID-19 detection. In addition, the lightweight networks reduce about 95% of the parameters compared with the the designed network prototype. Thus the lightweight networks have competitive classification performance with a few parameter usage. And the F1 and MCC results prove that our method obtains a stable diagnosis model. Besides, the high recall rate means that our models miss few cases of the ‘COVID-19’. This diagnosis results is critical to COVID-19 detection so that the medical systems can take rapid measures of isolation of infected patients to slow down the virus propagation.
For future work, we take ethical considerations of the neural network based method for COVID-19 detection in X-ray images. As a learning based method, the COVID-19 detection effect of the neural network depends on the quantity and quality of X-ray data. However, since these training X-ray images are from patients, the gap between technical feasibility and what is allowed from privacy legislation is growing. To integrate the medical data from medical institutions worldwide without violating privacy rights and interests, in the future, federated learning can be considered to build a decentralized model, so as to share the model and learn parameters without sharing data.
Data availability
Data will be made available on request.
Acknowledgements
This paper has been supported by the National Science Foundation of China (No.12271503, No.11991021,No.72231010, No.71932008).
1 https://aistudio.baidu.com/aistudio/datasetdetail/84451
==== Refs
References
1 Singhal T. A review of coronavirus disease-2019 (covid-19) The indian journal of pediatrics 87 4 2020 281 286 32166607
2 Basu S. Mitra S. Saha N. Deep learning for screening covid-19 using chest x-ray images, in, IEEE Symposium Series on Computational Intelligence IEEE 2020 2020 2521 2527
3 Jin Y.-H. Cai L. Cheng Z.-S. Cheng H. Deng T. Fan Y.-P. Fang C. Huang D. Huang L.-Q. Huang Q. A rapid advice guideline for the diagnosis and treatment of 2019 novel coronavirus (2019-ncov) infected pneumonia (standard version) Military Medical Research 7 1 2020 1 23 31928528
4 M.-Y. Ng, E.Y. Lee, J. Yang, F. Yang, X. Li, H. Wang, M.M.-s. Lui, C.S.-Y. Lo, B. Leung, P.-L. Khong, et al., Imaging profile of the covid-19 infection: radiologic findings and literature review, Radiology: Cardiothoracic Imaging 2 (1).
5 Rubin G.D. Ryerson C.J. Haramati L.B. Sverzellati N. Kanne J.P. Raoof S. Schluger N.W. Volpi A. Yim J.-J. Martin I.B. The role of chest imaging in patient management during the covid-19 pandemic: a multinational consensus statement from the fleischner society Chest 158 1 2020 106 116 32275978
6 Chen N. Zhou M. Dong X. Qu J. Gong F. Han Y. Qiu Y. Wang J. Liu Y. Wei Y. Epidemiological and clinical characteristics of 99 cases of 2019 novel coronavirus pneumonia in wuhan, china: a descriptive study The lancet 395 10223 2020 507 513
7 Wang L. Lin Z.Q. Wong A. Covid-net: A tailored deep convolutional neural network design for detection of covid-19 cases from chest x-ray images Scientific Reports 10 1 2020 1 12 31913322
8 Ozturk T. Talo M. Yildirim E.A. Baloglu U.B. Yildirim O. Acharya U.R. Automated detection of covid-19 cases using deep neural networks with x-ray images Computers in biology and medicine 121 2020 103792
9 Minaee S. Kafieh R. Sonka M. Yazdani S. Soufi G.J. Deep-covid: Predicting covid-19 from chest x-ray images using deep transfer learning Medical image analysis 65 2020 101794
10 Ismael A.M. Şengür A. Deep learning approaches for covid-19 detection based on chest x-ray images Expert Systems with Applications 164 2021 114054
11 E.E.-D. Hemdan, M.A. Shouman, M.E. Karar, Covidx-net: A framework of deep learning classifiers to diagnose covid-19 in x-ray images, arXiv preprint arXiv:2003.11055.
12 S.U.K. Bukhari, S.S.K. Bukhari, A. Syed, S.S.H. Shah, The diagnostic evaluation of convolutional neural network (cnn) for the assessment of chest x-ray of patients infected with covid-19, MedRxiv.
13 Ucar F. Korkmaz D. Covidiagnosis-net: Deep bayes-squeezenet based diagnosis of the coronavirus disease 2019 (covid-19) from x-ray images Medical Hypotheses 140 2020 109761
14 Luz E. Silva P. Silva R. Silva L. Guimarães J. Miozzo G. Moreira G. Menotti D. Towards an effective and efficient deep learning model for covid-19 patterns detection in x-ray images Research on Biomedical Engineering 38 1 2022 149 162
15 Chaudhary Y. Mehta M. Sharma R. Gupta D. Khanna A. Rodrigues J.J. Efficient-covidnet: deep learning based covid-19 detection from chest x-ray images, in, IEEE international conference on e-health networking, application & services IEEE 2021 2020 1 6
16 X. Li, C. Li, D. Zhu, Covid-mobilexpert: On-device covid-19 screening using snapshots of chest x-ray, arXiv preprint arXiv:2004.03042.
17 Li X. Li C. Zhu D. Covid-mobilexpert: On-device covid-19 patient triage and follow-up using chest x-rays, in, IEEE international conference on bioinformatics and biomedicine IEEE 2020 2020 1063 1067
18 Kaur M. Kumar V. Yadav V. Singh D. Kumar N. Das N.N. Metaheuristic-based deep covid-19 screening model from chest x-ray images, Journal of Healthcare Engineering 2021
19 Apostolopoulos I.D. Mpesiana T.A. Covid-19: automatic detection from x-ray images utilizing transfer learning with convolutional neural networks Physical and engineering sciences in medicine 43 2 2020 635 640 32524445
20 M. Tan, Q. Le, Efficientnet: Rethinking model scaling for convolutional neural networks, in: International conference on machine learning, PMLR, 2019, pp. 6105–6114.
21 Rajaraman S. Siegelman J. Alderson P.O. Folio L.S. Folio L.R. Antani S.K. Iteratively pruned deep learning ensembles for covid-19 detection in chest x-rays IEEE Access 8 2020 115041 115050 32742893
22 Goodfellow I. Bengio Y. Courville A. Deep learning MIT press 2016
23 Huang C. Wang Y. Li X. Ren L. Zhao J. Hu Y. Zhang L. Fan G. Xu J. Gu X. Clinical features of patients infected with 2019 novel coronavirus in wuhan, china The lancet 395 10223 2020 497 506
24 Pal K.K. Sudeep K. Preprocessing for image classification by convolutional neural networks in: 2016 IEEE International Conference on Recent Trends in Electronics 2016 Information & Communication Technology IEEE 1778 1781
25 R.S. of North America, Covid-19 radiography database, https://www.kaggle.com/tawsifurrahman/covid19-radiography-database, https://drive.google.com/file/d/1xt7g5LkZuX09e1a8rK9sRXIrGFN6rjzl/view?usp=sharing (2020).
26 Buda M. Maki A. Mazurowski M.A. A systematic study of the class imbalance problem in convolutional neural networks Neural networks 106 2018 249 259 30092410
27 J. Yoon, S.J. Hwang, Combined group and exclusive sparsity for deep neural networks, in: Proceedings of the 34th International Conference on Machine Learning-Volume 70, JMLR. org, 2017, pp. 3958–3966.
| 36460228 | PMC9706991 | NO-CC CODE | 2022-12-13 23:17:36 | no | Methods. 2023 Jan 29; 209:29-37 | utf-8 | Methods | 2,022 | 10.1016/j.ymeth.2022.11.004 | oa_other |
==== Front
Med Drug Discov
Med Drug Discov
Medicine in Drug Discovery
2590-0986
The Author(s). Published by Elsevier B.V.
S2590-0986(22)00029-X
10.1016/j.medidd.2022.100148
100148
Article
Repurposing of drugs for combined treatment of COVID-19 cytokine storm using machine learning
Gantla Maanaskumar R. a
Tsigelny Igor F. bcd⁎
Kouznetsova Valentina L. bc
a MAP Program, UC San Diego, Calif, USA
b San Diego Supercomputer Center, UC San Diego, Calif, USA
c BiAna, La Jolla, Calif, USA
d Dept. of Neurosciences, UC San Diego, Calif, USA
⁎ Corresponding author at: San Diego Supercomputer Center, UC San Diego, Calif, USA.
29 11 2022
2 2023
29 11 2022
17 100148100148
15 9 2022
26 10 2022
1 11 2022
© 2022 The Author(s)
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Severe acute respiratory syndrome coronavirus 2 (SARS‑CoV‑2) induced cytokine storm is the major cause of COVID-19 related deaths. Patients have been treated with drugs that work by inhibiting a specific protein partly responsible for the cytokines production. This approach provided very limited success, since there are multiple proteins involved in the complex cell signaling disease mechanisms. We targeted five proteins: Angiotensin II receptor type 1 (AT1R), A disintegrin and metalloprotease 17 (ADAM17), Nuclear Factor‑Kappa B (NF‑κB), Janus kinase 1 (JAK1) and Signal Transducer and Activator of Transcription 3 (STAT3), which are involved in the SARS‑CoV‑2 induced cytokine storm pathway. We developed machine-learning (ML) models for these five proteins, using known active inhibitors. After developing the model for each of these proteins, FDA-approved drugs were screened to find novel therapeutics for COVID‑19. We identified twenty drugs that are active for four proteins with predicted scores greater than 0.8 and eight drugs active for all five proteins with predicted scores over 0.85. Mitomycin C is the most active drug across all five proteins with an average prediction score of 0.886. For further validation of these results, we used the PyRx software to conduct protein–ligand docking experiments and calculated the binding affinity. The docking results support findings by the ML model. This research study predicted that several drugs can target multiple proteins simultaneously in cytokine storm-related pathway. These may be useful drugs to treat patients because these therapies can fight cytokine storm caused by the virus at multiple points of inhibition, leading to synergistically effective treatments.
Keywords
COVID-19
SARS-CoV-2
Docking
Machine learning
Multi-targeted drug discovery
Screening of FDA-approved drugs
Abbreviations
1D 2D 3D, one- two- three-dimensional
ADAM17, A disintegrin and metalloprotease 17
ARDS, acute respiratory distress syndrome
AT1R, Angiotensin II receptor type 1
AUROC, Area under receiver operator characteristic curve
COVID-19, coronavirus disease 2019
CRS, cytokine release syndrome
CXCL10, CXC-chemokine ligand 10
FDA, Food and Drug Administration
G-CSF, granulocyte colony stimulating factor
IC50, half maximal inhibitory concentration
ICU, intensive care unit
IL, interleukin
JAK1, Janus kinase 1
MCP1, monocyte chemoattractant protein-1
MIP1α, macrophage inflammatory protein 1
ML, machine learning
NF-κB, Nuclear Factor-Kappa B
PaDEL, Pharmaeutical data exploration laboratory
PDB, Protein Data Bank
ROC, receiver operator characteristic curve
SMILES, Simplified Molecular-Input Line-Entry System
STAT3, signal transducer and activator of transcription 3
TNFα, tumor necrosis factor α
WEKA, Waikato Environment for Knowledge Analysis
==== Body
pmc1 Introduction
The COVID-19 pandemic caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) resulted in millions of infected patients and deaths worldwide [1], [2]. Patients frequently encountered complications with significant mortality, particularly by acute respiratory distress syndrome (ARDS) with a broad spectrum of issues such as multiple-organ failure, and blood clots [3], [4]. There has been tremendous amount of research going on towards discovering therapeutics for the COVID-19, and few drugs have been approved by FDA such as remdesivir, Paxlovid and molnupiravir, and all of them mainly target viral proteins [5], [6].
Mounting research data reveals that the severity of COVID-19 is mainly associated with an increased level of inflammatory mediators including cytokines and chemokines such as interleukin IL-2, IL-7, IL-8, IL-9, IL-10, IL-17, tumor necrosis factor alpha (TNFα), monocyte chemoattractant protein-1 (MCP1), macrophage inflammatory protein 1 alpha (MIP1α), granulocyte colony stimulating factor (G-CSF), CXC-chemokine ligand 10 (CXCL10), C-reactive protein, ferritin, and d-dimers in blood upon SARS-CoV-2 infection [7], [8], [9], [10], [11], [12], [13]. More specifically, patients in intensive care unit (ICU) showed higher levels of plasma inflammatory cytokines compared to non-ICU patients [14], and therefore fatal COVID-19 is characterized as a cytokine release syndrome (CRS) that is caused by a cytokine storm. Thus, targeting proteins responsible for cytokine storm serves as a possible mechanism of treatment for severe COVID-19 patients [15], [16], [17].
The SARS-CoV-2 induced cytokine storm pathway [18] shows that there are multiple proteins involved in the disease signaling mechanisms. Cytokines are cell signaling, small protein molecules that aid cell to cell communication in immune responses and stimulate the movement of cells towards sites of inflammation, infection, and trauma [19]. Cytokine Storm is essentially an unregulated immune response characterized by an excessive release of multiple pro-inflammatory cytokines [20], [21].
It has been identified that proteins such as Angiotensin II receptor type 1 (AT1R), A disintegrin and metalloprotease 17 (ADAM17), Nuclear Factor-Kappa B (NF-κB), Janus kinase 1 (JAK1) and Signal transducer and activator of transcription 3 (STAT3) are implicated in the production of proinflammatory cytokines and are considered as a promising COVID-19 therapeutic targets [15]. Therefore, discovering a drug that can interfere with function of either all of the proteins or most of them synergistically would become an effective therapeutic. Based on literature search, as of now there are no such therapeutics exist. Discovery of novel effective drugs and therapies for COVID-19 is critical for tackling the disease. However, discovery and development of effective therapies can be costly and time-consuming. For this reason, it would be ideal to repurpose already existing FDA-approved drugs given the proven safety, if they can also interfere effectively with proteins responsible for cytokine storm.
In this pathway, we have investigated five proteins: AT1R, ADAM17, NF-κB1, JAK1, STAT3. AT1R signaling axis activates ADAM17, which results in the production of cytokines TNFα and IL-6. The IL-6 amplifier plays a critical role in chronic inflammatory diseases. Activation of NF-κB, JAK1 and STAT3 triggers the IL-6 amplifier, which causes the cytokine storm and leads to the ARDS and multiple-organ failure. Targeting these five proteins would prevent cytokine storm to yield the best potent COVID-19 drug.
Conventional methods of drug discovery are very expensive, complex processes that takes several years to bring drugs to the clinic. We used machine learning to expedite the drug discovery process by screening FDA drugs, so that the treatment for COVID-19 is available sooner.
Recently, machine learning (ML) has emerged as an important computational technique and has been applied to various tasks in drug discovery, such as molecular property prediction and drug–target interaction prediction. Given the great advantage of this computational tool in terms of the cost and time, in this project we have used ML classification model with a random forest algorithm in WEKA software [22] for repurposing of some FDA-approved drugs for use as COVID-19 therapeutics. These predictions can then be confirmed through structure-based virtual screening, specifically using docking simulators PyRX. The docking provides the binding energy for each conformer and helps validate the accuracy of prediction.
2 Materials and methods
All research was completed in silico. The programs, tools, and websites used: PubChem, ZINC database subsection covering FDA-approved drugs, Protein Data Bank (PDB), Pharmaceutical Data Exploration Laboratory (PaDEL)-Descriptor, Waikato Environment for Knowledge Analysis (WEKA), PyRx, Discovery Studio Visualizer. A flowchart of methods is presented in Fig. 1 .Fig. 1 Flowchart of the methods.
2.1 Data collection
Data for known active inhibitors and a control set of random compounds obtained from PubChem are listed in Table 1 . Data for FDA-approved drugs obtained from the ZINC database. Activity values and SMILES (Simplified Molecular-Input Line-Entry System) [23] files for compounds tested with the proteins AT1R, ADAM17, NF-κB, JAK1, STAT3 were retrieved from PubChem. To limit the tested compounds to the strongest inhibitors, compounds with top 100 IC50 values were chosen for training the model. One-thousand-six-hundred-fifteen FDA-approved drugs and their SMILES were retrieved from ZINC database.Table 1 Known inhibitors obtained from PubChem.
Protein Number of Known Inhibitors IC50 values range (μM)
AT1R 1192 0.00005–19.98
ADAM17 1813 0.000026–44.0
NF-κB 348 0.003–49.6
JAK1 4596 0.0000013–39.81
STAT3 588 0.0084–48.0
The chemical structures are obtained in SMILES format. These files are 1D ASCII strings that represent 3D molecular structure. An example of top ten inhibitors for AT1R are shown in Table 2 .Table 2 The top ten inhibitors for AT1R.
Compound IC50(nM)
BDBM50049199 0.05
Saralasin 0.06
2Botbmip 0.08
CHEMBL42775 0.01
CHEMBL158809 0.01
CHEMBL298417 0.01
BDBM50283219 0.01
BDBM50283237 0.01
BDBM50283194 0.01
BDBM50283245 0.01
2.2 Molecular descriptor calculation
PaDEL-Descriptor software [24] is used to calculate the molecular descriptors for the compounds. These descriptors are the characteristics of the compound that are used for training of the ML model. For example, number of aromatic rings, number of pi bonds, molecular weight, atom count, etc. The software currently calculates 1875 descriptors (1444 1D and 2D descriptors and 431 3D descriptors) and 12 types of fingerprints. For our model building we have used 1444 1D and 2D descriptors.
2.3 InfoGain filtering in WEKA to select top 100 descriptors
To narrow down the calculated 1D and 2D descriptors from 1444 to only the most significant ones, we utilized attribute selection from WEKA [25], an open-source ML software. The descriptors were ranked by the Information Gain Attribute Evaluation (InfoGain) function, an unsupervised machine-learning algorithm, that measures how important each descriptor is in determining whether a given molecule is an inhibitor or not. InfoGain measures how each feature contributes to decreasing the overall computational entropy. Only the most significant descriptors were selected to be used by the ML model to reduce noise.
2.4 Building a machine-learning model
Machine-learning model for each protein was built using WEKA [22]. WEKA provides both standard and extensive ML functionality, integrated within classification, regression, clustering and other pattern recognition capabilities. Data for the model is prepared by taking top 100 descriptors of top 100 inhibitors for each of the proteins and 100 control set of random molecules.
First, we submitted the prepared inhibitor file containing selected and random compounds with their molecular descriptors into WEKA. Then used the Random Forest algorithm 10-folds cross validation to build the model. Also, we used Random Forest algorithm with an 80/20% training–testing split to evaluate the performance of the model. Such the training–testing split ensures that there is no overfitting as 20% of the data. It was not used to build the model but used for testing. Then we analyzed the model accuracy and elucidated the ROC curves. Saved model was used in the next step to screen FDA- approved drugs. The Receiver Operating Characteristic (ROC) curves were calculated to measure the effectiveness of the model. ROC curve summarizes the prediction performance of a classification model at all classification thresholds. Fig. 2 a–2e present the ROC graphs for machine-learning models of proteins AT1R, ADAM17, NF-κB, JAK1, and STAT3. Model accuracy is 91.5–99.0% range and Area Under the Receiver-Operating Characteristic curve (AUROC) is 0.97–1.00. That values confirm the accuracy of the models. Receiver-Operating Characteristic (ROC) curves for five proteins are shown in Fig. 2.Fig. 2 Accuracies and AUROC of the predictions of inhibitors for five proteins related to cytokine storm in COVID-19: (a) AT1R Accuracy 98.5% and AUROC 99%; (b) ADAM17 Accuracy 98.5% and AUROC 99%; (c) NF-κB Accuracy 96%, AUROC 99%; (d) JAK1 Accuracy 98.5%, AUROC 100%; (e) STAT3 Accuracy 91.5%, AUROC 97.8%.
2.5 Screening of FDA-approved drugs using the model
FDA approved drugs are downloaded from ZINC database [26]. Using PaDEL-Descriptor software, the molecular descriptors were calculated for all the downloaded 1665 FDA-approved drugs. Out of 1444 descriptors, the same 100 descriptors were selected as the training set of the corresponding protein inhibitors. These were then screened with the ML model built using WEKA. The output was analyzed for the prediction scores and the predicted drugs were ranked based on the ML predicted score. The predicted drugs were ranked for each protein and averaged the score among all five proteins listed in Table S1 (Supplementary information).
2.6 Docking of predicted FDA-approved drugs to selected proteins
To confirm the activity and binding to the protein, docking of predicted FDA-approves drugs was performed using PyRx tool [27] with Discovery Studio software [28] to visualize the results. For docking the selected compounds, the crystal structure of the protein was downloaded from PDB [29], [30] for each of the five proteins. PDB IDs for selected proteins are AT1R−4ZUD, ADAM17–2FV5, NF-κB–1SVC, JAK1–4EI4, and STAT3–6NUQ. A binding active site is defined for each protein based on the reported ligand interactions with protein.
To validate the specificity of the docked compounds, docking of random compounds was also conducted. A random number generator without repetition was used to obtain 100 random compound IDs and to select entries from the PubChem database that correspond to the random numbers obtained.
Each of the five proteins’ 3D structure with a known ligand was downloaded from the PDB database. Each predicted FDA-approved drug’s 3D structure was downloaded from PubChem. From each downloaded protein–ligand complex, the ligand was removed in Discovery Studio and remained protein was loaded into PyRx. The active sites of each protein were defined as a box that encompasses residues of the binding site. Then we ran the AutoDock Wizard for the top 12 predicted compounds, 12 best-activity known compounds, and 12 random compounds with each protein. For each compound nine conformers were generated and docked. In total there were generated 324 conformers, which were docked to each protein. The docked protein–ligand complexes were analyzed to elucidate the interactions of compounds with amino-acid residues. Binding Free Energy values are listed in Table 8.
3 Results
3.1 Machine-learning prediction results
The results of the ML models’ predictions were evaluated using confusion matrices and their derivatives: the accuracy (ACC), precision (PREC), Matthews correlation coefficient (MCC), true-positive rate (TPR) or recall (REC), false-positive rate (FPR), as well as the area under the receiver operating characteristic (ROC) curve (AUROC), and the area under the precision–recall curve (PRC area).
The weighted averages for each of these metrics are listed in Table 3 . The ROC curve compares the sensitivity and specificity across a range of values. Thus, the vertical axis is the TPR, that is, the sensitivity or recall; and the horizontal axis is the FPR or (1 − specificity). The FPR is the probability of falsely classifying a positive class. The model’s low FPR of 0.015 to 0.040 demonstrates a low probability of wrongly classifying an inactive compound to active one. The TPR (sensitivity) is the probability of correctly classifying a positive class. The model’s high TPR of 0.915 to 0.985 indicates a high probability of correctly classifying an active compound. The large average AUROC value 0.978 to 1.0 indicates that the classification is accurate. Another way to evaluate the performance of the proposed method is the PRC area, which shows precision values for the corresponding sensitivity (recall, i.e., TPR) values. The model’s large PRC area value of 0.979 to 1.0 again shows the good performance of our method for all the five proteins.Table 3 Performance of the developed ML models for the five proteins related to cytokine storm in COVID–19.
Protein ACC TPR FPR PREC MCC AUROC PRC Area
AT1R 98.5 % 0.985 0.015 0.985 0.970 0.999 0.999
ADAM17 98.5 % 0.985 0.015 0.985 0.970 0.999 0.999
NF--κB 96.0 % 0.960 0.040 0.961 0.921 0.993 0.993
JAK1 98.5 % 0.985 0.015 0.985 0.970 0.999 0.999
STAT3 91.5 % 0.915 0.085 0.915 0.830 0.978 0.979
Note: ACC, accuracy; TPR, true-positive rate; FPR, false-positive rate; PREC, precision; MCC, Matthews correlation coefficient; AUROC, area under the receiver-operating characteristic curve; PRC area, area under the precision–recall curve.
The predicted drugs were ranked for each protein and averaged the score among all five proteins listed in Table S1 (Supplementary information). Total 45 compounds found to be active inhibiting all five proteins AT1R, ADAM17, NF-κB, STAT3, JAK1. Forty-five active compounds with the greater than 0.6 average predictive score is shown in Table S1.
Top eight FDA-approved drugs predicted active for all five proteins AT1R, ADAM17, NF-κB, STAT3, and JAK1 are shown in Table 4 .Table 4 Eight compounds active for five proteins with greater than 0.85 average predictive score.
Name of Predicted Inhibitor Prediction score Average score ML Rank
AT1R ADAM17 NF-κB JAK1 STAT3
Mitomycin C 0.99 0.76 0.90 0.78 1.00 0.886 1
Valrubicin 0.97 0.73 1.00 0.71 1.00 0.882 2–3
Pomalidomide 1.00 0.76 0.92 0.73 1.00 0.882 2–3
Fludarabine 1.00 0.73 0.94 0.73 1.00 0.880 4
Clarithromycin 0.99 0.70 0.95 0.72 1.00 0.872 5–6
Trabectedin 0.99 0.75 0.91 0.71 1.00 0.872 5–6
Capreomycin 1.00 0.70 0.89 0.75 1.00 0.868 7
Sonidegib 0.91 0.72 0.94 0.71 0.98 0.852 8
Mitomycin C (Table 4) is top ranked FDA drug across all five proteins with average prediction score of 0.886.
Top 20 FDA-approved drugs predicted active for four proteins (AT1R, ADAM17, NF-κB, STAT3) showed an average score of greater than 0.83 are shown in Table 5 .Table 5 Twenty compounds active for four proteins with greater than 0.83 average predictive score.
Name of Predicted Inhibitor Prediction score Average score ML Rank
AT1R ADAM17 NF-κB STAT3
Abacavir 0.93 0.73 1.00 1.00 0.9600 1–2
Raltegravir 0.73 0.74 0.98 0.89 0.9600 1–2
Saxagliptin 0.99 0.88 0.97 1.00 0.9250 3–4
Valrubicin 0.97 0.73 1.00 1.00 0.9250 3–4
Pomalidomide 1.00 0.76 0.92 1.00 0.9200 5
Nimodipine 0.97 0.72 0.95 1.00 0.9175 6–7
Fludarabine 1.00 0.73 0.94 1.00 0.9175 6–7
Suvorexant 0.95 0.72 1.00 1.00 0.9150 8
Boceprevir* 0.72 0.72 0.99 0.97 0.9125 9–11
Mitomycin C 0.99 0.76 0.90 1.00 0.9125 9–11
Trabectedin 0.99 0.75 0.91 1.00 0.9125 9–11
Saquinavir 0.99 0.88 0.97 1.00 0.9100 12–13
Clarithromycin** 0.99 0.70 0.95 1.00 0.9100 12–13
Capreomycin 1.00 0.70 0.89 1.00 0.8975 14–16
Balsalazide 0.93 0.71 0.96 0.94 0.8875 14–16
Sonidegib 0.91 0.72 0.94 0.98 0.8875 14–16
Minocycline 0.73 0.74 0.98 0.89 0.8850 17
Eribulin 0.92 0.71 0.86 1.00 0.8500 18
Isradipine 0.98 0.72 0.94 1.00 0.8350 19–20
Cangrelor 0.96 0.70 0.95 1.00 0.8350 19–20
*Boceprevir [31] showed in-vitro activity for COVID-19.
**Clarithromycin [32], [33] is in clinical trials to treat COVID-19.
Thirteen drugs (Table 5) showed average scores of greater than 0.9. Abacavir and raltegravir (Table 5) showed top average scores of 0.96. Chemical structures of top active compounds are shown in Fig. 3 .Fig. 3 Chemical structures of the top predicted active for treatment of cytokine storm FDA-approved drugs: (a) mitomycin C; (b) abacavir; (c) raltegravir.
Current use of predicted active drugs is shown in Table S2 (Supplementary information).
3.2 Docking results
Docking of the top three predicted FDA-approved drugs mitomycin C, abacavir and raltegravir with the proteins' binding sites are shown in Fig. 4 and Fig. 5 .Fig. 4 Docking of mitomycin C with proteins: (a) AT1R; (b) ADAM17; (c) NF-κB; (d) JAK1; (e) STAT3.
Fig. 5 Docking of abacavir and raltegravir with proteins: (a) AT1R–abacavir; (b) ADAM17–abacavir; (c) NF-κB–abacavir; (d) JAK1–abacavir; (e) STAT3–abacavir; (f) AT1R–raltegravir; (g) ADAM17–raltegravir; (h) NF‑κB–raltegravir; (i) JAK1–raltegravir; (j) STAT3–raltegravir.
Mitomycin C docked with all five proteins showed binding affinity ranging from −6.0 to −7.5 kcal/mol. (Table 6 ).Table 6 Binding free energy for mitomycin C.
Protein Binding free energy (kcal/mol)
AT1R −7.5
ADAM17 −7.0
NF-κB −6.0
JAK1 −7.2
STAT3 −6.5
Abacavir docked with four proteins showed binding free energy ranging from −6.1 to −8.6 kcal/mol (Table 7 ). Raltegravir docked with four proteins showed binding energy ranging from −7.4 to −9.6 kcal/mol (Table 7) which is considered as reasonable binding affinity.Table 7 Binding free energy for abacavir and raltegravir.
Protein Binding Free energy with abacavir (kcal/mol) Binding Free energy with raltegravir (kcal/mol)
AT1R −7.8 −9.4
ADAM17 −8.6 −9.6
NF-κB −6.1 −7.4
STAT3 −7.2 −8.3
Binding affinities calculated using PyRx docking for the top 12 compounds. Table 8 lists the top predicted active compounds docked for all five proteins.Table 8 Binding free energies of the top 12 ML-predicted FDA-approved drugs docked using PyRx software.
Compound Binding free energy to
AT1R ADAM17 NF-κB JAK1 STAT3
Mitomycin C −9.2 −7.9 −6.1 −7.4 −6.7
Pomalidomide −8.5 −7.1 −5.9 −8.3 −7.2
Fludarabine −8.1 −9.1 −6.2 −8.1 −7.5
Sonidegib −10.1 −10.1 −7.3 −9.1 −8.5
Abacavir −9.6 −8.5 −7.8 −10.1 −7.9
Raltegravir −7.8 −7.3 −6.1 −7.7 −7.2
Saxagliptin −7.3 −8.6 −5.3 −6.9 −5.9
Nimodipine −7.9 −6.2 −6.4 −8.4 −7.5
Suvorexant −9.4 −8.9 −7.4 −9.7 −8.3
Boceprevir −8.6 −9.6 −6.8 −7.4 −7.8
Balsalazide −7.5 −7.1 −6.0 −7.2 −6.5
Minocycline −6.5 −7.0 −5.6 −7.2 −6.1
The box plots (Fig. 6 ) demonstrated that the predicted inhibitors had the better binding affinities than known and random compounds. They were constructed using binding affinities obtained from docking for predicted compounds, known inhibitors and control random compounds and are shown in Fig. 6. Docking free energies are listed in Table S3 (Supplementary information).Fig. 6 Free energies of docking interactions—docking scores—of predicted, known, and random compounds: (a) AT1R; (b) ADAM17; (c) NF-κB; (d) JAK1; and (e) STAT3.
The box plots (Fig. 6) demonstrates that the predicted inhibitors had an average binding affinity of −6.40 kcal/mol (NF-κB) to −8.37 kcal/mol for AT1R (Fig. 6a), which was better than that of known inhibitors, which had an average of −6.0 kcal/mol for STAT3 (Fig. 6e) to −7.63 kcal/mol for AT1R (Fig. 6). The control group of random molecules had an average binding affinity of −4.78 kcal/mol for NF-κB (Fig. 6c) to −7.0 kcal/mol for ADAM17 (Fig. 6b). This confirms that the predicted inhibitors performed statistically better than the control group.
In this study, five crucial proteins—AT1R, ADAM17, NF-κB, JaK1, STAT3—playing the important roles in cytokine production pathway are targeted to predict the best potential drug for treatment of COVID-19 cytokine storm. Number of known inhibitors with reported IC50 values for each protein obtained from PubChem were 1192, 1813, 348, 4596, and 588 respectively. Machine-learning models developed exhibited an accuracy ranging from 91.5 to 99.0%, with their AUROC values ranging from 0.98 to 1.0, which is considered as excellent predictive performance of the models.
The box plots (Fig. 6) show that the predicted active compounds have better binding energies than the already known inhibitors and control set of random compounds. One can see that the predicted inhibitors had an average binding affinity of −6.40 kcal/mol (NF-κB) to −8.37 kcal/mol (AT1R), which was better than that of known inhibitors, which had an average of −6.0 kcal/mol (STAT3) to −7.63 kcal/mol (AT1R). The control group of random molecules had an average binding affinity of −4.78 kcal/mol (NF-κB) to −7.0 kcal/mol (ADAM17). This confirms the predicted inhibitors performed statistically better than the control group.
From the docking results, binding free energy values ranging from −6.0 to −9.6 kcal/mole targeted for five and four proteins confirms that the predicted compounds bind at the active site of the proteins. The amino acids that showed interaction in the docking experiments for the top three drugs mitomycin C, abacavir and raltegravir are listed in Table 9 .Table 9 Summary of binding residues involved for each protein with top predicted drugs. Bold are residues that are involved in binding to more than one compound.
Protein binding residues Mitomycin C Abacavir Raltegravir
AT1R Tyr35, Trp84, Thr88, Arg167, Ile288 Tyr35, Trp84, Tyr87, Tyr92, Ile288, Val108 Trp84, Val108, Arg167, Lys199, Ile288
ADAM17 Gly354, Ser355, Ser360, Gly362, Thr461, Ser457 Leu348, Glu398, Val402, Glu406, His405, Leu401, Try436, Ile438, Ala439, Val440 Gly346, Thr347, Leu348, Gly349, Val402, Glu406, His405, His415, Ile438, Ala439
NF-κB Ser243, Ser249, Asp250, Asp274, Lys275, Lys244 Arg54, Lys52, Ala73, Lys252, Leu251, Glu341 Lys52, Gln53, Arg54, Ala73, Glu341, Thr342
JAK1 Asp1003, Gly884, Asp1042 Gly887, Leu910, Leu922, Leu1024, Gly884, His885, Asp1042 Asp1039, His918, Asp921, Lys1026, Lys908, Leu1024, Arg1002, His885
STAT3 Lys383, Ser381, Leu436, His437 Gln247, Cys251, Ile252, Pro256, Glu324, Arg325, Asp334, Pro336 Asp371, Ser381, Leu430, Leu438, Lys488, Val490
From Table 9 one can see that AT1R’s binding residues—Tyr35, Trp84, Val108, Arg167, and Ile288—are common for mitomycin C, abacavir and raltegravir; ADAM17′s binding residues—Glu406, His405, Ile438, and Ala439—are common for abacavir and raltegravir; NF-κB’s binding residues—Lys52, Ala73, and Glu341—are for abacavir and raltegravir; JAK1′s binding residues—Asp1042, Gly884, and His885—are common for abacavir and raltegravir; STAT3′s binding residue—Ser381—is for abacavir and raltegravir binding to STAT3.
These results support the idea that the compounds are binding at the active site of protein and not at the non-bonding sites, thus proving that these compounds could act specifically on the protein makes them some of the most promising candidates to treat COVID-19.
4 Discussion
The main goal of this study was to predict the drugs that can target as many as possible cytokine-related genes. We know that there is a number of such genes that are activated in response on SARS-CoV-2 viral proteins. Here we have a problem that is eternal for medicine. We cannot prescribe more drugs than a set that would be tolerated by the organism. Here we selected five genes that have to be targets of inhibiting agent to prevent the cytokine storm or just decrease the immune response to viral agents. Our model predicted drugs that can simultaneously target at least four of such cytokine-related genes. Based on model's average prediction scores over 0.6, 45 FDA-approved drugs were predicted active, with 20 drugs having predicted scores greater than 0.8 for 4 proteins (AT1R, ADAM17, NF-κB, STAT3). Eight FDA-approved drugs had predicted scores over 0.85 for all five proteins involved in the mechanism of the cytokine storm (AT1R, ADAM17, NF-κB, JAK1, STAT3). Mitomycin C is the most active drug across all five proteins with an average prediction score of 0.886. Abacavir and raltegravir are the top active compounds for four proteins with average scores of 0.96.
We predicted several drugs that can target simultaneously several proteins in cytokine storm related pathway. These may be useful drugs to treat patients because these therapies can fight cytokine storm caused by the virus at multiple points of inhibiting, leading to synergistically effective treatments.
Mitomycin C ranked top with highest average scores for all five proteins suggesting that it possesses all required chemical functional groups with desired spatial arrangements, so that it can interact and bind well with all proteins. Mitomycin C is approved drug for cancer but has several side effects such as bone marrow suppression and septicemia because of leukopenia and needs to be further evaluated in clinical experiments. We need to note that despite of clinical usage in multiple cancers, mitomycin C has been reported with several side-effects and listed as potent DNA crosslinker. Furthermore, mitomycin C is a probable human carcinogen, classified as weight-of-evidence Group B2 under the EPA Guidelines for Carcinogen Risk. Abacavir and raltegravir with excellent predictive scores for four proteins (except JAK1), suggesting that changes in chemical structure of the drugs made the difference in biological activity.
Two of predicted drugs—boceprevir [31] and clarithromycin [32], [33] are already tested for COVID-19 treatment.
5 Conclusions
Our hypothesis that it is possible to develop a valid predicting machine-learning model to select the drugs that would target multiprotein pathways for treatment of the COVID-19 cytokine storm is confirmed and gave the model accuracy ranging from 91.5 to 99%, with AUROC ranging from 0.978 to 1.0, considered as excellent predictive performance of the models.
This study not only provides drug candidates that could treat COVID-19, but it also demonstrates the application of predictive models for multitarget drug discovery approach with machine learning.
Future steps for this project would be to confirm the inhibitory activity of the predicted drugs against the target proteins in animal models. We need to note that there are differences between the binders (identified by computational screening) and the therapeutic active drugs for specific diseases. The formers have to be further tested in pharmacology experiments to know their actual action modes (agonism, antagonism, inhibitory effect or so on). These drugs definitely need to be tested in pharmacological experiments to establish the mechanism of action before a clinical use. The methods of this study could be extended to predictive models for discovering therapeutics for other disease areas, such as chronic inflammatory diseases.
CRediT authorship contribution statement
Maanaskumar R. Gantla: Methodology, Software, Formal analysis, Data curation, Writing – original draft. Igor F. Tsigelny: Methodology, Project administration. Valentina L. Kouznetsova: Conceptualization, Methodology, Validation, Writing – review & editing, Supervision.
==== Refs
References
1 World Health Organization (WHO). Coronavirus Disease 2019 (COVID-19) Situation Report: Weekly epidemiological Update on COVID-19. https://www.who.int/emergencies/diseases/novel-coronavirus-2019/situation-reports.
2 Johns Hopkins University & Medicine: Coronavirus Resource Center https://coronavirus.jhu.edu.
3 Hsu CY, Lai CC, Yeh YP, Chuane CC, Chen HH. Progression from Pneumonia to ARDS as a Predictor for Fatal COVID-19. J Infect Public Health 2021; 14: 504-507. https:/doi.org/10.1016/j.jiph.2020.12.026.
4 Aslan A. Aslan C. Zolbanin N.M. Jafari R. Acute respiratory distress syndrome in COVID-19: possible mechanisms and therapeutic management Pneumonia 13 2021 1 15 10.1186/s41479-021-00092-9 33487176
5 Robinsona P.C. Liew D.F.L. Tannera H.L. Graingerf J.R. Dwekg R.A. Reislerh R.B. COVID-19 therapeutics: Challenges and directions for the future PNAS 119 2022 1 10 10.1073/pnas.2119893119
6 Rando HM, Wellhausen N, Ghosh S, Lee AJ, Dattoli AA, Hu F, et al. Greene CS. Identification and Development of Therapeutics for COVID-19. mSystems 2021; 6: 1–52. https://journals.asm.org/doi/10.1128/mSystems.00233-21.
7 Tay M.Z. Poh C.M. Rénia L. MacAry P.A. Ng L.F.P. The trinity of COVID-19: Immunity, inflammation and intervention Nat Rev Immunol 20 6 2020 363 374 10.1038/s41577-020-0311-8 32346093
8 Chen G. Wu D. Guo W. Cao Y. Huang D. Wang H. Clinical and immunological features of severe and moderate coronavirus disease 2019 J Clin Invest 130 5 2020 2620 2629 10.1172/JCI137244 32217835
9 Chen N. Zhou M. Dong X. Qu J. Gong F. Han Y. Epidemiological and clinical characteristics of 99 cases of 2019 novel coronavirus pneumonia in Wuhan, China: A descriptive study Lancet 395 10223 2020 507 513 10.1016/S0140-6736(20)30211-7 32007143
10 Kim J.Y. Ko J.H. Kim Y. Kim Y.J. Kim J.M. Chung Y.S. Viral load kinetics of SARS–CoV–2 infection in first two patients in Korea J Korean Med Sci 35 7 2020 e86 32080991
11 Liu Y, Yang Y, Zhang C, Huang F, Wang F, Yuan J, et al. Clinical and biochemical indexes from 2019–nCoV infected patients linked to viral loads and lung injury. Sci China Life Sci 2020; 63(3): 364–374. https://doi.org/0.1007/s11427-020-1643-8.
12 Pan Y. Zhang D. Yang P. Poon L.L.M. Wang Q. Viral load of SARS-CoV-2 in clinical samples Lancet Infect Dis 20 4 2020 411 412 10.1016/S1473-3099(20)30113-4 32105638
13 Phan L.T. Nguyen T.V. Luong Q.C. Nguyen T.V. Nguyen H.T. Le H.Q. Importation and human-to-human transmission of a novel coronavirus in Vietnam N Engl J Med 382 9 2020 872 874 10.1056/NEJMc2001272 31991079
14 Huang C. Wang Y. Li X. Ren L. Zhao J. Hu Y. Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China Lancet 395 10223 2020 497 506 10.1016/S0140-6736(20)30183-5 31986264
15 Hirano T. Murakami M. COVID–19: A new virus, but a familiar receptor and cytokine release syndrome Immunity 52 5 2020 731 733 10.1016/j.immuni.2020.04.003 32325025
16 Mahmudpour M. Roozbeh J. Keshavarz M. Farrokhi S. Nabipour I. COVID–19 cytokine storm: The anger of inflammation Cytokine 133 2020 155151 10.1016/j.cyto.2020.155151
17 McGonagle D. Sharif K. O’Regan A. Bridgewood C. The role of cytokines including interleukin–6 in COVID–19 induced pneumonia and macrophage activation syndrome-like disease Autoimmun Rev 19 6 2020 102537 10.1016/j.autrev.2020.102537
18 Hojyo S. Uchida M. Tanaka K. Hasebe R. Tanaka Y. Murakami M. How COVID–19 induces cytokine storm with high mortality Inflamm Regen 40 2020 37 10.1186/s41232-020-00146-3 33014208
19 Zhang J.M. An J. Cytokines, inflammation and pain Int Anesthesiol Clin 45 2 2007 27 37 10.1097/AIA.0b013e318034194e 17426506
20 Ragab D. Eldin H.S. Taeimah M. Khattab R. Salem R. The COVID–19 cytokine storm; What we know so far Front Immunol 11 2020 1446 10.3389/fimmu.2020.01446 32612617
21 Tisoncik J.R. Korth M.J. Simmons C.P. Farrar J. Martin T.R. Katze M.G. Into the eye of the cytokine storm Microbiol Molecular Biol Rev 76 2012 16 32 10.1128/MMBR.05015-11
22 Frank E, Hall MA, Whitten IH. The WEKA Workbench. Online Appendix for Data Mining: Practical Machine Learning Tools and Techniques, Morgan Kaufmann, Fourth Edition, 2016; https://www.cs.waikato.ac.nz/ml/weka/book.html.
23 NCI/CADD Group. Online SMILES Translator and Structure File Generator. National Institutes of Health, U.S. Department of Health and Human Services, National Cancer Institute. 21 Apr. 2020; https://cactus.nci.nih.gov/translate (Last accessed 25 May 2022).
24 Yap C.W. PaDEL-descriptor: An open-source software to calculate molecular descriptors and fingerprints J Comput Chem 32 7 2010 1466 1474 10.1002/jcc.21707 21425294
25 WEKA 3.8.5 University of Waikato; software download; https://www.cs.waikato.ac.nz/∼ml/weka/.
26 (a) Irwin JJ, Soichet BK. ZINC--a free database of commercially available compounds for virtual screening. J Chem Inf Model. 2005; 45(1): 177–182. https://doi.org/ 10.1021/ci049714+. (b) Sterling T, Irwin JJ. ZINC 15--Ligand discovery for everyone. J Chem Inf Model 2015; 55(11): 2324–2337. https://doi.org/10.1021/acs.jcim.5b00559. (c) https://zinc.docking.org/ (Last accessed 25 May 2022).
27 (a) Dallakyan S, Olson AJ. Small-molecule library screening by docking with PyRx. Methods Mol Biol 2015; 1263: 243–250. https://doi.org/10.1007/978-1-4939-2269-7_19; (b) Python Prescription: Virtual Screening Tool. https://pyrx.sourceforge.io/ (Last accessed 23 May 2022).
28 Dassault Systemes, Free Download: BIOVIA Discovery Studio Visualizer. https://discover.3ds.com/discovery-studio-visualizer-download (Last accessed 27 May 2022).
29 (a) Berman HM, Westbrook J, Feng Z, Gilliland G, Bhat TN, Weissig H, et al. The Protein Data Bank. Nucleic Acids Res 2000; 28(1): 235–242. https://doi.org/10.1093/nar/28.1.235; (b) RCSB PDB: Protein Data Bank. http://www.rcsb.org/ (Last accessed 23 May 2022).
30 Mukund V. Behera S.K. Alam A. Nagaraju G.P. Molecular docking analysis of nuclear factor–κB and genistein interaction in the context of breast cancer Bioinformation 15 1 2019 11 17 10.6026/97320630015011 31359993
31 Fu L. Ye F. Feng Y. Yu F. Wang Q. Wu Y. Both Boceprevir and GC376 efficaciously inhibit SARS–CoV–2 by targeting its main protease Nat Commun 11 1 2020 4417 10.1038/s41467-020-18233-x 32887884
32 Tsiakos K. Tsakiris A. Tsibris G. Voutsinas P.M. Panagopoulos P. Kosmidou M. Early start of oral clarithromycin is associated with better outcome in COVID–19 of moderate severity: The ACHIEVE Open-Label Single-Arm Trial Infect Dis Ther 10 4 2021 2333 2351 10.1007/s40121-021-00505-8 34363189
33 Yamamoto K. Hosogaya N. Sakamoto N. Yoshida H. Ishii H. Yatera K. Efficacy of clarithromycin in patients with mild COVID 19 pneumonia not receiving oxygen administration: Protocol for an exploratory, multicentre, open-label, randomised controlled trial (CAME COVID 19 study) BMJ Open 11 9 2021 e053325
| 36466363 | PMC9706997 | NO-CC CODE | 2022-12-07 23:16:30 | no | Med Drug Discov. 2023 Feb 29; 17:100148 | utf-8 | Med Drug Discov | 2,022 | 10.1016/j.medidd.2022.100148 | oa_other |
==== Front
J Hepatol
J Hepatol
Journal of Hepatology
0168-8278
1600-0641
European Association for the Study of the Liver. Published by Elsevier B.V.
S0168-8278(22)03312-8
10.1016/j.jhep.2022.11.018
Letter to the Editor
Comment on: Impact of the COVID-19 pandemic on liver disease-related mortality rates in the United States
Zhang Han M.D.
Peng Yan M.D.
Tang Xiaowei M.D., Ph.D ∗
Department of Gastroenterology, the Affiliated Hospital of Southwest Medical University, Luzhou, China
∗ Corresponding author. Department of Gastroenterology, Affiliated Hospital of Southwest Medical University, Street Taiping No.25, Region Jiangyang, Luzhou, 646099, Sichuan Province, China. Tel.: +8608303165200, Fax: +86083061641541, .
29 11 2022
29 11 2022
7 11 2022
18 11 2022
© 2022 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved.
2022
European Association for the Study of the Liver
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
==== Body
pmcTo the editor:
We read with great interest the article by Gao et al. 1, who used a national death dataset to determine the impact of the COVID-19 pandemic on people with liver disease in the USA. Finally, they found that age-standardised mortality rates for alcohol-associated liver disease (ALD) and non-alcoholic fatty liver disease (NAFLD) increased at an alarming rate during the COVID-19 pandemic. Hospitalization for ALD is also on the rise during the COVID-19 pandemic due to increased alcohol consumption. We conducted a relevant analysis in support of this view to confirm it (Fig. 1 ).Figure 1
Previous studies reported an increase in alcohol abuse in 2020 due to increased economic insecurity, unemployment, and psychological distress in the COVID-19 pandemic 2. In the first year of the pandemic, alcohol sales in the United States increased significantly, from $7.1 billion in 2019 to $9.5 billion in 2020 2. With the increase in alcohol sales, recent studies have also identified an increase in hospitalization for ALD [3], [4], [5]. Therefore, we conducted a meta-analysis and aimed to assess the admission growth rate of ALD. Using the pre-designed search strategy, we identified the relevant studies that compared the admission rate of ALD between before and during the COVID-19 pandemic in Embase, Cochrane, and PubMed databases from inception to October 2022. The primary outcome was the pooled admission growth rate of ALD. Finally, a total of three studies were included [3], [4], [5]. Overall, the pooled admission growth rate of ALD was 43.6% (95% CI, 27.3% to 61.4%; I2 = 94.139%). We observed a significant increase in the number of ALD admissions during the COVID-19 pandemic compared with previous years.
The COVID-19 pandemic is having a significant impact on the health status of people with ALD. Increased alcohol consumption has the potential to exacerbate the condition in patients with liver impairment, leading to increased admission rates for ALD. Alcohol consumption is a major contributor to acute-on-chronic liver failure, a common syndrome in patients with underlying cirrhosis characterized by acute decompensation of the liver cirrhosis, organ failure, and high short-term mortality. So the increased alcohol consumption during the COVID-19 pandemic is bound to bring increased rate of hospitalization, decompensation of cirrhosis, and mortality.
Conflict of interest statement
Han Zhang, Yan Peng, and Xiaowei Tang declare that they have no conflict of interest.
Author contributions
Study conception and design: Xiaowei Tang; acquisition of data: Han Zhang; drafting of manuscript: Han Zhang; revision of manuscript: Yan Peng; and final approval of manuscript: Xiaowei Tang.
Financial support
This study is independent research funded by no grants.
==== Refs
References
1 Gao X. Lv F. He X. Zhao Y. Liu Y. Impact of the COVID-19 pandemic on liver disease-related mortality rates in the United States J Hepatol 2022 S0168-8278(22)02994-4
2 Deutsch-Link S. Curtis B. Singal A.K. Covid-19 and alcohol associated liver disease Dig Liver Dis 54 11 2022 1459 1468 35933291
3 Sohal A. Khalid S. Green V. Gulati A. Roytman M. The Pandemic Within the Pandemic: Unprecedented Rise in Alcohol-related Hepatitis During the COVID-19 Pandemic J Clin Gastroenterol 56 3 2022 e171 e175 34653062
4 Gonzalez H.C. Zhou Y. Nimri F.M. Rupp L.B. Trudeau S. Gordon S.C. Alcohol-related hepatitis admissions increased 50% in the first months of the COVID-19 pandemic in the USA Liver Int 42 4 2022 762 764 35094494
5 Shaheen A.A. Kong K. Ma C. Doktorchik C. Coffin C.S. Swain M.G. Impact of the COVID-19 Pandemic on Hospitalizations for Alcoholic Hepatitis or Cirrhosis in Alberta, Canada Clin Gastroenterol Hepatol 20 5 2022 e1170 e1179 34715379
| 36460166 | PMC9707017 | NO-CC CODE | 2022-12-01 23:20:19 | no | J Hepatol. 2022 Nov 29; doi: 10.1016/j.jhep.2022.11.018 | utf-8 | J Hepatol | 2,022 | 10.1016/j.jhep.2022.11.018 | oa_other |
==== Front
J Clin Virol
J Clin Virol
Journal of Clinical Virology
1386-6532
1873-5967
Elsevier B.V.
S1386-6532(22)00279-7
10.1016/j.jcv.2022.105347
105347
Article
Mixed outcomes following a third SARS-CoV-2 vaccine dose in previously unresponsive people with HIV
Hassold Nolan a
Brichler Ségolène b
Gater Yamina b
Leclerc Delphine a
Gordien Emmanuel b
Bouchaud Olivier a
Carbonnelle Etienne bc
Mechai Frédéric ac
Cordel Hugues a
Delagreverie Héloïse bc⁎
a Department of Infectious and Tropical Diseases, Hôpital Avicenne AP-HP, 125 rue de Stalingrad, 93000, Bobigny, France
b Department of Clinical Microbiology, Hôpital Avicenne AP-HP, 125 rue de Stalingrad, 93000, Bobigny, France
c Faculté de Médecine, INSERM U1137 IAME, Université Sorbonne Paris Nord and Université Paris Cité, 16 rue Henri Huchard, 75018, Paris, France
⁎ Corresponding author at: Laboratoire de Microbiologie Clinique, Hôpital Avicenne, 125 rue de Stalingrad, 93000, Bobigny, France.
29 11 2022
1 2023
29 11 2022
158 105347105347
4 8 2022
15 11 2022
© 2022 Elsevier B.V. All rights reserved.
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Keywords
HIV-1
CD4+ T-cell
COVID-19
Anti-SARS-COV-2 antibodies
SARS-COV-2 vaccine
==== Body
pmcDear Editor,
In the continuing SARS-CoV-2 pandemics, protecting immunocompromised populations is a top priority. Vaccination is key to worldwide healthcare strategies; the standard course involves two baseline injections followed with a first booster dose at six months. However, effective immunization may require additional doses and vaccination may even fail to elicit immunity in fragile populations [1,2]. People living with HIV (PLWH) are not enlisted as “at-risk” despite well-known HIV-related immune alterations and increased COVID-19 mortality [3]. Large vaccine trials did include some PLWH with normal CD4+ T-cell counts on antiretroviral therapy and vaccine efficiency was reportedly unchanged in these participants. However, studies by our group and others described impaired seroconversion rates and low anti-SARS-CoV-2-spike IgG titers following primary two-dose vaccination in PLWH with decreased CD4+ T-cell counts [4,5].
We report inconsistent antibody responses to a booster third vaccine dose in non- and low-responder PLWH (anti-SARS-CoV-2-Spike IgG titer <260 BAU/mL [6] after two injections, n = 38) from a previous study in adult PLWH [4]. Over the December 2021 – April 2022 period, 28 of 38 participants agreed to receive a third dose (booster) of vaccine (BNT162b2 or mRNA-1273). Four contracted COVID-19 with mild symptoms before providing serum samples, and one asymptomatic participant with post-infection anti-SARS-CoV-2-Capsid (N) antibodies was excluded. Twenty-three participants with post-boost anti-SARS-CoV-2-Spike (S) antibody titers and no history of COVID-19 were included: 70% were male, the median age was 56 years (interquartile range (IQR) 53–67) and the median CD4+ T-cell count was globally improved on antiretroviral therapy from 386 cells/µl at primary vaccination to 556 cells/µl at the time of the study (Table 1 ). Thirteen of 23 (56%) were initially vaccinated with CD4+ T-cell counts less than 500 cells/µl; they were 7 (30%) at booster vaccination.Table 1 Participant characteristics.
Table 1 Primary 2-dose vaccination Booster third dose p
n 23 -
Male: n (%) 16 (70%) -
Comorbidities (hypertension, diabetes, respiratory disease, BMI>30kg/m2): n(%) 10 (43.5%) -
Age at booster dose: yr [IQR] 56.0 [52.9–66.8] -
Time since HIV diagnosis, at booster dose: yr [IQR] 25.4 [10.5–30.8] -
Vaccine (n): BNT162b2/mRNA-1273/ChAdOx1-nCoV19 15/3/5 18/5/0 -
On a prescription of antiretroviral therapy when vaccinated: n (%) 22 (96%) 23 (100%) -
CD4+ T-cells /µL: n [IQR] 386 [217–585] 556 [286–726] 0.01
Months since last immunization at anti-Spike IgG quantification: n [IQR] (Min / Max) 2.7 [1.87–3.43] (0.33 / 4.9) 2.4 [1.63–3.0] (0.67 / 4.5) 0.38
Anti-Spike IgG antibodies: median BAU/mL [IQR] 89.0 [47.4-169] 387.3 [242–1816] <0.0001
Anti-Spike antibodies <260 BAU/mL: n (%) 23 (100) 6 (26.1) <0.0001
n: number, BMI: body mass index, IQR: interquartile range, yr: years, BAU: normalized binding antibody units. Distributions were compared with the Mann-Whitney, Wilcoxon and Chi2 tests.
Booster shots were received 6 months after primary immunization (IQR: 5.5–7.0). The overall anti-SARS-CoV-2-Spike IgG titer at 2.4 months (IQR: 1.7–3) after the booster was 387 BAU/mL (IQR: 251–1749). In the 7 participants with CD4+T-cell counts less than 500/µl it was 356 (IQR: 37–1817). Antibody titers were generally improved by the boost (Fig. 1 ). However, 6 of 23 participants (26%) still did not achieve the putative protection threshold of 260 BAU/mL [6] after the third shot. They were not older than responders (median age 55.8 years, p = 0.72) and 5 of 6 had increased CD4+T-cells counts since dose 2. The two participants with 5 and 40 BAU/mL (80-year-old with 322 CD4+T-cells/µl and 58-year-old with 15 CD4+T-cells/µl, respectively) contracted the Omicron variant and died of COVID-19 despite ICU care.Fig. 1 Anti-SARS-COV-2-Spike IgG titers following booster vaccination in low-responder people living with HIV.
Specific anti-SARS-CoV-2-Spike IgG antibodies were quantified in 23 participants in normalized binding antibody units (BAU) (Architect SARS-COV-2 IgG II Quant, Abbott, USA) after vaccine dose 2 and after vaccine dose 3. Open circles: CD4+ T-cell count < 500/µL; black dots: CD4+ T-cell count > 500/µL. Cross symbols: COVID-19 deaths. Dotted line: putative protection threshold of 260 BAU/mL[6]. The 23 included samples tested negative for anti-SARS-CoV-2-N protein antibodies (Architect SARS-COV-2 IgG Assay, Abbott).
Fig 1
In this group of primary non- and low-responders to the standard 2-dose immunization, a booster dose increased anti-SARS-CoV-2-Spike antibody titers in most participants. This result emphasizes the importance of booster doses, including for patients previously unresponsive and who might lose interest in receiving extra shots. Among 7 cases of breakthrough SARS-COV-2 infections in the Delta/Omicron wave after three doses of vaccine, five exhibited no or mild symptoms. However, two persistent non-responders, with low CD4 T-cell counts, died of COVID-19.
These observations highlight the inconsistency of vaccine-induced SARS-CoV-2 immunity in PLWH, particularly with decreased CD4+ T-cell counts (participants with CD4+T-cells <500/µl: 356 (IQR 37-1817) BAU/mL anti-Spike IgG). Such outcomes in fragile patients support post-immunization serological testing and additional vaccine challenges and/or immunotherapy in unresponsive PLWH, along with efforts towards HIV care programs in regions with high HIV prevalence and susceptibility to prolonged and severe SARS-CoV-2 infections.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
The authors are staff of Paris Hospitals (Assistance Publique-Hôpitaux de Paris) and did not receive additional specific funding from any agency for this work. The authors have no competing interests to declare.
The study was registered by the local institutional ethics committee under number CLEA-2021-215.
==== Refs
References
1 Naylor KL Kim SJ Smith G McArthur E Kwong JC Dixon SN Effectiveness of first, second, and third COVID-19 vaccine doses in solid organ transplant recipients: a population-based cohort study from Canada Am. J. Transplant 2022 10.1111/ajt.17095 Published Online First: 16 May
2 Herishanu Y Rahav G Levi S Braester A Itchaki G Bairey O Efficacy of a third BNT162b2 mRNA COVID-19 vaccine dose in patients with CLL who failed standard 2-dose vaccination Blood 139 2022 678 685 34861036
3 Ssentongo P Heilbrunn ES Ssentongo AE Advani S Chinchilli VM Nunez JJ Epidemiology and outcomes of COVID-19 in HIV-infected individuals: a systematic review and meta-analysis Sci. Rep. 11 2021 6283 33737527
4 Hassold N Brichler S Ouedraogo E Leclerc D Carroue S Gater Y Impaired antibody response to COVID-19 vaccination in advanced HIV infection AIDS 36 2022 F1 F5 35013085
5 Liu Y Han J Li X Chen D Zhao X Qiu Y COVID-19 vaccination in people living with HIV (PLWH) in China: a cross sectional study of vaccine hesitancy, safety, and immunogenicity Vaccines (Basel) 9 2021 1458 34960204
6 Feng S Phillips DJ White T Sayal H Aley PK Bibi S Correlates of protection against symptomatic and asymptomatic SARS-CoV-2 infection Nat. Med. 2021 10.1038/s41591-021-01540-1 Published Online First: 29 September
| 36476807 | PMC9707020 | NO-CC CODE | 2022-12-05 23:15:19 | no | J Clin Virol. 2023 Jan 29; 158:105347 | utf-8 | J Clin Virol | 2,022 | 10.1016/j.jcv.2022.105347 | oa_other |
==== Front
Genes Dis
Genes Dis
Genes & Diseases
2352-4820
2352-3042
Chongqing Medical University. Production and hosting by Elsevier B.V.
S2352-3042(22)00295-1
10.1016/j.gendis.2022.10.023
Full Length Article
Weakened humoral and cellular immune response to the inactivated COVID-19 vaccines in Chinese individuals with obesity/overweight
Zhu Qian a1
Zhang Yingzhi a1
Kang Juan a1
Chen Zhiwei a
Peng Mingli a
Chen Min a
Zhang Gaoli a
Xiang Dejuan a
Xiao Shuang a
Li Hu a
Mei Ying b
Yang Jie b
Qi Xiaoya b
Cai Dachuan a∗∗
Ren Hong a∗
a Department of Infectious Diseases, Key Laboratory of Molecular Biology for Infectious Diseases (Ministry of Education), Institute for Viral Hepatitis, The Second Affiliated Hospital, Chongqing Medical University, Chongqing 400061, China
b Health Management Center, The Second Affiliated Hospital of Chongqing Medical University, Chongqing 400061, China
∗ Corresponding author. No. 288, Tianwen Avenue, Chayuan, Nan'an District, Chongqing 401336, China.
∗∗ Corresponding author.
1 These authors have contributed equally to this work.
29 11 2022
29 11 2022
19 6 2022
24 8 2022
23 10 2022
© 2022 Chongqing Medical University. Production and hosting by Elsevier B.V.
2022
Chongqing Medical University
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Inactivated COVID-19 vaccines have been widely used to vaccinate the Chinese population. However, limited literature exists to explore the effect of obesity on the humoral and cellular immune response to these vaccines. In this study, 132 high BMI (Body mass index) (obesity and overweight, BMI ≥ 24 kg/m2) and 82 normal BMI (BMI < 24 kg/m2) participants were enrolled. Adverse events (AEs), Spike receptor-binding domain IgG antibody (anti-RBD-IgG), neutralizing antibodies (NAbs), and specific B-cell and T-cell responses were evaluated 21–105 days after full-course inactivated COVID-19 vaccination. The overall incidence of AEs was similar in individuals with and without obesity/overweight. No serious vaccine-related AEs occurred. Individuals with obesity/overweight had a reduced seropositivity rate of NAbs compared to those with normal BMI. Anti-RBD-IgG and NAbs titers in the high BMI group were significantly lower than those in the normal BMI group. The frequencies of RBD-specific memory B cells (MBCs) and the numbers of spike-specific TNF-α+ spot-forming cells (SFCs) in individuals with obesity/overweight were reduced compared with those noted in individuals without obesity/overweight. A similar trend of weakened humoral responses was also observed in individuals with central obesity. Our study results suggested that inactivated COVID-19 vaccines were safe and well tolerated but induced poor humoral and cellular immune responses in Chinese individuals with obesity/overweight.
Keywords
COVID-19
Immune response
Inactivated vaccine
Obesity
Overweight
==== Body
pmcIntroduction
The pandemic of coronavirus disease 2019 (COVID-19) caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is causing severe health problems and economic burdens worldwide. Obesity has been reported as a risk factor for a more severe course and outcomes of COVID-19, such as increased rates of hospitalization, ICU admission, and mortality.1 , 2 This evidence suggests that individuals with obesity should be prioritized for COVID-19 vaccination. According to the emerging clinical data, three FDA-approved SARS-CoV-2 vaccines (Pfizer-BioNTech, Moderna, and Johnson & Johnson) are widely used in the obese population in the USA and some European countries.3
Inactivated COVID-19 vaccines (BBIBP-CorV, CoronaVac) have been widely used to vaccinate the Chinese population in China mainland. However, limited literature exists to explore the effect of obesity on the safety and immune response of these vaccines. A study evaluated the effectiveness of these vaccines and found that obesity might be a risk factor for decreased antibody titers.4 To the best of our knowledge, past studies have demonstrated associations between obesity and impaired immune responses to influenza, hepatitis B, tetanus, and rabies vaccines.5, 6, 7, 8 Therefore, related data about inactivated COVID-19 vaccines for Chinese people with obesity are needed and encouraged, especially targeted studies on different obesity subtypes or different patterns of body fat distribution.3
The purpose of this study was to evaluate the safety profile and provide new insights into inactivated COVID-19 vaccine-induced humoral and cellular immune responses in Chinese individuals with obesity/overweight.
Material and methods
Human subjects and clinical samples
This study was approved by the Ethics Committee of the Second Affiliated Hospital of Chongqing Medical University and conformed with the ethical guidelines of the Declaration of Helsinki. In this study, 132 individuals with obesity/overweight and 82 individuals with normal BMI were enrolled between August 16, 2021 and October 14, 2021. According to the Chinese BMI classification criteria recommended in the expert consensus on body weight management among patients with overweight or obesity (2021),9 obesity is defined as BMI ≥ 28 kg/m2, and overweight is defined as 24 kg/m2 ≤ BMI < 28 kg/m2. Normal BMI was defined as BMI < 24 kg/m2. Central obesity is defined as a waist circumference ≥ 90 cm for men or > 85 cm for women. All enrolled human subjects met the following inclusion criteria: (i) age above 18 years; (ii) completed full-course vaccination with BBIBP-CorV/CoronaVac for more than 21 days. The exclusion criteria included (i) individuals with a history of SARS-CoV-2 infection or a history of contact with a confirmed or suspected COVID-19 patient; (ii) coinfection with hepatitis B virus, hepatitis C virus, and HIV; (iii) a history of malignant tumor, renal failure, and other immune diseases; (iv) undergoing any immunosuppressant treatment; and (v) pregnancy. All human subjects were recruited from the health management center of the Second Affiliated Hospital of Chongqing Medical University. All human subjects provided signed informed consent before enrollment. This study has been registered at ClinicalTrials.gov (NCT05043272), and the follow-up is ongoing.
Data collection
Demographic information and clinical data of participants were obtained by electronic medical records. Vaccine type, vaccination time, and adverse events within 7 days or 30 days were all recorded by questionnaires. We defined 21–45 days after the second dose as 1 month, 46–75 days as 2 months, and 76–105 days as 3 months in the study. Adverse events included local adverse events (pain, swelling, itch, induration, and redness) and systemic adverse events (fatigue, dizziness, somnolence, cough, fever, diarrhea, laryngeal pain, muscle pain, rhinorrhea, nausea, chest distress, chest pain, abdominal pain, chills, constipation, elevated blood pressure, headache, inappetence, palpitation, pruritus, and rash). All participants' peripheral blood samples were collected 21–105 days after full-course COVID-19 vaccination for laboratory tests. The study design was summarized in Fig. 1 .Figure 1 The flowchart of the study design.
Fig. 1
Evaluation of spike protein receptor-binding domain IgG antibody
Indirect ELISA was used to detect IgG-binding antibodies against the RBD antigen of SARS-CoV-2 (Sino Biological, Beijing, China) (Supplementary Methods). The detection limit for the anti-RBD-IgG antibody test was 1:50 for the inactivated COVID-19 vaccine. Seropositivity was defined as anti-RBD-IgG antibody titers greater than 1:50. Undetectable antibody titers in plasma were assigned values of 1:25 for calculation.
Evaluation of neutralizing antibody (RBD-ACE2 blocking antibody)
The neutralizing antibody was assayed by competitive ELISA (Sino Biological, Beijing, China) (Supplementary Methods). The detection limit for the neutralizing antibody test was 1:5 for the inactivated COVID-19 vaccine. Neutralizing antibody titers ≥1:5 were considered to indicate seropositivity. Undetectable antibody titers in plasma were assigned values of 1:2.5 for calculation.
Detection of SARS-CoV-2-specific B cells and regulatory T cells by flow cytometry
According to the manufacturer's instructions, peripheral blood mononuclear cells (PBMCs) were isolated from heparinized whole blood and stained for 30 min at 4 °C using an antigen probe and conjugated antibodies (Supplementary Methods). Samples were evaluated by flow cytometry (Beckman Coulter, CytoFLEX) and analyzed by FlowJo (Treestar, 10.0.7r2).
Detection of spike-specific IFN-γ and TNF-α by FluoroSpot assay
Enumeration of cells secreting IFN-γ and tumor necrosis factor-α (TNF-α) against SARS-CoV-2 spike protein was conducted using a FluoroSpot kit (FSP-0109-2/FSP-0109-10, Mabtech, Sweden) according to the manufacturer's instructions (Supplementary methods). Plates were analyzed using an automated FluoroSpot reader (Elispot Reader, AID, Germany).
Statistical analysis
The Chi-square test and Fisher's exact test were used for categorical variables. The Mann–Whitney U test and Kruskal–Wallis test were used to compare two groups and multiple groups for continuous variables, respectively. All results of multiple comparisons were corrected using Bonferroni's correction. Geometric mean titers (GMTs) and their corresponding 95% confidential intervals (CIs) were calculated based on the standard normal distribution of the log-transformed antibody titers. A two-sided P value < 0.05 was considered statistically significant. SPSS (IBM, 22.0.0) was used for statistical analysis. GraphPad Prism (GraphPad Software Inc., 8.0.0) was used for plotting.
Results
Characteristics of all participants
A total of 132 individuals with high BMI (obesity or overweight) and 82 individuals with normal BMI were enrolled in this study. Participants ranged in age from 18 to 75 years. As shown in Table S1, 56.1% and 42.7% of males were in the high BMI group and normal BMI group, respectively. The age and sex distribution in the two groups were comparable (age, P = 0.603; sex, P = 0.057) (Table S1). No significant difference in the interval time (1, 2, and 3 months) from the second dose vaccination to sample collection was noted between the two groups (P = 0.482) (Table S1).
Adverse events of inactivated COVID-19 vaccines in all participants
The inactivated vaccines were well tolerated, and none of the participants had serious adverse events (grade 3 and 4 adverse events). Nineteen (13.9%) of 132 participants in the high BMI group and 12 (14.6%) of 82 participants in the normal BMI group had at least one adverse event within 7 days of receiving inactivated vaccines with no significant difference noted between groups (Table S2). All adverse events were mild (grade 1 or 2 adverse events). In participants with high BMI, the most common local adverse events were injection-site pain (7.3%) and injection-site swelling (3.6%), whereas the most common systemic adverse events were fatigue (1.4%) and dizziness (1.4%). When the observation was prolonged to 30 days, no new adverse events occurred (Table S2).
Specific antibody responses to inactivated COVID-19 vaccines in individuals with obesity/overweight
Overall, the seropositivity rate of anti-RBD-IgG was high in both groups, with 96.7% in the high BMI group and 98.8% in the normal BMI group (Fig. 2 A, left panel). However, GMTs of anti-RBD-IgG were significantly lower in the high BMI group compared with the normal BMI group [209.7 (95%CI 179.0–245.6) vs. 373.8 (303.6–460.3), P < 0.001] (Fig. 2A, right panel). As a critical antibody against SARS-CoV-2, NAbs were also analyzed in this study. The seropositivity rate of NAbs showed a significant difference between the two groups (55.3% in individuals with high BMI and 73.2% in individuals with normal BMI, P = 0.010) (Fig. 2B, left panel). Moreover, NAb titers in the high BMI groups were also remarkably lower than those in the normal BMI group [4.5 (95%CI 4.0–5.1) vs. 6.5 (5.2–8.0), P = 0.005] (Fig. 2B, right panel). As expected, the two types of antibodies were correlated (r 2 = 0.604, P < 0.001) (Fig. S1).Figure 2 Antibody responses to inactivated COVID-19 vaccines in obesity and overweight individuals. (A, B) The seropositivity rate and antibody titers of anti-RBD-IgG (A) and NAbs (B) in individuals with high BMI and normal BMI. (C, D) The seropositivity rate and antibody titers of anti-RBD-IgG (C) and NAbs (D) in individuals with obesity, overweight and normal BMI. The error bars in antibody titers indicate the 95%CI of the GMTs. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001. High BMI, BMI ≥ 24 kg/m2; normal BMI, BMI < 24 kg/m2. CI, confidential interval; anti-RBD-IgG, spike receptor-binding domain IgG antibody; GMTs, geometric mean titers; NAbs, neutralizing antibodies.
Fig. 2
Next, we further compared the antibody responses in individuals with obesity, overweight or normal BMI, which was consistent with the overall trend. The seropositivity rates of anti-RBD-IgG in the obesity, overweight and normal BMI groups were 97.50%, 96.7%, and 98.8%, respectively (Fig. 2C, left panel). The obesity and overweight groups had lower anti-RBD-IgG titers compared with the normal BMI group [200.0 (151.9–263.3) vs. 214.0 (175.9–260.4) vs. 373.8 (303.6–460.3), P = 0.001], but no difference was observed between the subgroups (Fig. 2C, right panel). The seropositivity rates of NAbs in the three groups were 57.5%, 54.4%, and 73.2%, respectively (Fig. 2D, left panel). The NAbs titers in individuals with obesity and overweight were also lower than those in individuals with normal BMI [4.2 (3.5–5.0) vs. 4.7 (4.0–5.4) vs. 6.5 (5.2–8.0), P = 0.050] but similar between subgroups (Fig. 2D, right panel).
Finally, we analyzed antibody responses to inactivated vaccines in female and male participants and in older and young participants. In the subgroup analysis based on sex, females tended to have higher anti-RBD-IgG and NAb titers than males in the normal BMI group (Fig. S2). In the subgroup analysis based on age, elderly individuals (≥50 years) had lower antibody titers and seropositivity rates of NAbs than young individuals (<50 years) in the normal BMI group (Fig. S3).
In summary, the antibody responses to inactivated vaccines in individuals with obesity/overweight were poorer than those in individuals with normal BMI.
Specific antibody responses to inactivated COVID-19 vaccines in individuals with obesity/overweight over time
To observe the effect of antibody responses over time, we established subgroup analyses for 1 month, 2 months, and 3 months. Although no significant difference was observed at 2 months (P = 0.354), anti-RBD-IgG titers in the high BMI group at other time points were significantly reduced (P = 0.002 at 1 month, P = 0.002 at 3 months) (Fig. 3 A, right panel). It is worth noting that the NAb seropositivity rate in the high BMI group decreased faster than that in the normal BMI group, and the seropositivity rate at 3 months was significantly lower (26.7% vs. 65.2%, P = 0.002) (Fig. 3B, left panel). NAb titers also showed a similar trend, but a significant difference was only observed at 3 months (P = 0.003) (Fig. 3B, right panel). No significant differences in antibody titer or seropositivity rate were observed between the obesity group and the overweight group (Fig. 3C, D).Figure 3 Antibody responses to inactivated COVID-19 vaccines in obesity and overweight over time. (A, B) The seropositivity rate and GMTs of anti-RBD-IgG (A) and NAbs (B) in high BMI and normal BMI individuals at 1 month, 2 months, and 3 months. (C, D) The seropositivity rate and GMTs of anti-RBD-IgG (C) and NAbs (D) in individuals with obesity, overweight, and normal BMI at 1 month, 2 months, and 3 months. The error bars in antibody titers indicate the 95%CI of the GMTs. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001. High BMI, BMI ≥ 24 kg/m2; normal BMI, BMI < 24 kg/m2. CI, confidential interval; anti-RBD-IgG, spike receptor-binding domain IgG antibody; GMTs, geometric mean titers; NAbs, neutralizing antibodies.
Fig. 3
In general, the antibody responses to inactivated COVID-19 vaccines in individuals with obesity/overweight showed a downward trend over time.
Specific memory B-cell responses to inactivated COVID-19 vaccines in individuals with obesity/overweight
To further evaluate the humoral immune responses induced by inactivated vaccines in the obesity and overweight groups, we analyzed B cells from the three groups by flow cytometry. Notably, the frequency of RBD+ CD27+ memory B cells (MBCs) within RBD+ B cells in the high BMI group was lower (35.7% vs. 44.1%, P = 0.015) (Fig. 4 A, left panel). In the specific MBC subsets, we found that the frequency of activated MBCs (actMBCs) in the high BMI group was significantly lower (14.7% vs. 17.7%, P = 0.022). Resting MBCs (rMBCs) showed the same decreasing trend (20.8% vs. 22.3%, P = 0.053). However, the high BMI group had a higher frequency of intermediate MBCs (intMBCs) than the normal BMI group (42.6% vs. 39.2%, P = 0.074) (Fig. 4A, right panel). Furthermore, the frequencies of RBD-specific CD38+ CD27+ MBCs and CD38+ actMBCs were significantly reduced in the high BMI group (CD38+ CD27+ MBCs, 2.21% vs. 3.24%, P = 0.001; CD38+ actMBCs, 4.86% vs. 6.16%, P = 0.007) (Fig. 4B). The gating strategy and representative results are shown in Figure S4.Figure 4 RBD-specific memory B-cell responses and T-cell responses to inactivated COVID-19 vaccines in obesity and overweight. (A) The frequencies of RBD+ CD27+ MBCs (left panel), and the percentage of rMBCs, intMBCs, actMBCs, atyMBCs in total RBD+ MBCs (right panel) in high BMI and normal BMI individuals. (B) The frequencies of RBD-specific CD38+ CD27+ MBCs (left panel) and CD38+ actMBCs (right panel) in high BMI and normal BMI individuals. (C) The frequencies of RBD+ CD27+ MBCs (left panel), and the percentage of rMBCs, intMBCs, actMBCs, atyMBCs in total RBD+ MBCs (right panel) in obesity, overweight, and normal BMI individuals. (D) The frequencies of RBD-specific CD38+ CD27+ MBCs (left panel) and CD38+ actMBCs (right panel) in obesity, overweight, and normal BMI individuals. The error bars represent median (IQR). (E) The number of spike-specific IFN-γ+ SFCs and TNF-α+ SFCs per 106 PBMC in high BMI and normal BMI individuals. (F) The number of spike-specific IFN-γ+ SFCs and TNF-α+ SFCs per 106 PBMC in obesity, overweight, and normal BMI groups. The error bars represent mean (SEM). (G) The frequency of Tregs within CD4+ T cells in high BMI and normal BMI individuals. (H) The frequency of Tregs within CD4+ T cells in obesity, overweight, and normal BMI groups. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001. actMBCs, activated MBCs; atyMBCs, atypical MBCs; intMBCs, intermediate MBCs; MBCs, memory B cells; IQR, interquartile range; PBMC, peripheral blood mononuclear cell; rMBCs, resting MBCs; SFC, spot-forming cell; SEM, standard error of mean; Treg, regulatory T cell.
Fig. 4
Although there was no significant difference, BMI subgroup analysis also showed a trend of lower frequencies of RBD-specific CD27+ MBCs, actMBCs, and rMBCs but higher frequencies of intMBCs in the obesity and overweight groups (Fig. 4C). In addition, the frequencies of CD38+ CD27+ MBCs and CD38+ actMBCs in the obesity and overweight groups decreased significantly (Fig. 4D).
The analysis of subgroups by time points also showed a similar trend of lower frequencies of RBD-specific CD27+ MBCs, actMBCs, rMBCs, and CD38+ CD27+ MBCs, and CD38+ actMBCs in individuals with obesity/overweight; however, the significant difference was observed only at 2 months and 3 months (Fig. S5).
In summary, the durable humoral immune response to inactivated COVID-19 vaccines in individuals with obesity/overweight was weakened.
T-cell responses to inactivated COVID-19 vaccines in individuals with obesity/overweight
We next evaluated the specific cellular immune response to inactivated COVID-19 vaccines in obese/overweight participants. Fluorospot assay results showed that the number of spike-specific TNF-α+ spot-forming cells (SFCs) decreased significantly in the high BMI group. Spike-specific IFN-γ+ SFCs showed the same trend, but no significant difference was observed (Fig. 4E). The BMI subgroup analysis also demonstrated lower TNF-α+ SFCs in the obesity and overweight groups (Fig. 4F).
We also measured regulatory T cells (Tregs) by flow cytometry, and the results showed that the frequency of Tregs within CD4+ T cells was significantly greater in individuals with a high BMI (Fig. 4G). Similar trends were displayed in the obesity and overweight subgroups (Fig. 4H). The gating strategy and representative results are shown in Figure S6.
Specific humoral immune response to inactivated COVID-19 vaccines in individuals with central obesity
To observe the humoral immune response to inactivated COVID-19 vaccines in individuals with different patterns of body fat distribution, we also analyzed the antibody responses and RBD-specific MBCs in individuals with central obesity. The seropositivity rates of anti-RBD-IgG in individuals with central obesity and noncentral obesity were 96.4% and 98.7%, respectively (Fig. 5 A, left panel). During the observation period, the anti-RBD-IgG titer in the central obesity group was obviously lower than that in the noncentral obesity group [192.6 (151.8–244.2) vs. 309.1 (246.7–387.3), P = 0.008] (Fig. 5A, right panel). The seropositivity rates of NAbs in individuals with central obesity and noncentral obesity were 60.0% and 66.7%, respectively (Fig. 5B, left panel). The titers of NAbs also showed the same trend [4.5 (3.8–5.3) vs. 5.5 (4.6–6.6)] (Fig. 5B, right panel). No differences in the frequencies of RBD-specific CD27+ MBCs and subsets (rMBC, actMBC, and intMBC) were noted between the central obesity and noncentral obesity groups (Fig. S7A). However, participants with central obesity had lower frequencies of RBD-specific CD38+ actMBCs (5.43% vs. 3.95%, P = 0.003) than those without central obesity (Fig. S7B).Figure 5 Antibody responses to inactivated COVID-19 vaccines in central obesity and noncentral obesity. (A, B) The seropositivity rate and antibody titers of anti-RBD-IgG (A) and NAbs (B) in central obesity and noncentral obesity individuals. (C, D) The seropositivity rate and antibody titers of anti-RBD-IgG (C) and NAbs (D) in central obesity and noncentral obesity individuals at 1 month, 2 months, and 3 months. The error bars in antibody titers indicate the 95%CI of the GMTs. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001. CI, confidential interval; anti-RBD-IgG, spike receptor-binding domain IgG antibody; NAbs, neutralizing antibodies.
Fig. 5
Taken together, individuals with central obesity may have worse humoral immune responses to inactivated COVID-19 vaccines than those without central obesity.
Discussion
The main finding of our study was that the inactivated COVID-19 vaccines were safe and well tolerated with relatively mild adverse events in Chinese individuals. However, full-course inactivated COVID-19 vaccination induced a poor humoral and cellular immune response in Chinese individuals with obesity/overweight.
In this study, the AE rate was 13.9% in the high BMI group and 14.6% in the normal BMI group within 7 days of receiving inactivated vaccines. All AEs were mild and moderate, and the most common local AE was injection-site pain, which was consistent with a clinical trial of inactivated COVID-19 vaccines.10 , 11 This finding demonstrates that vaccination with inactivated vaccines is safe and well tolerated in Chinese individuals with a high BMI.
Next, we evaluated the antibody responses among individuals with obesity/overweight. The overall NAb seropositivity rates in the obesity group and the overweight group were 57.5% and 54.4%, respectively, which were significantly lower than that in the normal BMI group (73.20%). A sharp reduction in NAb seropositivity over time was observed in the high BMI group at 3 months with a seropositivity rate of only 26.7%. This finding indicates that individuals with obesity/overweight may have worse antibody responses than individuals with normal BMI. Considering that NAbs are critical against live SARS-CoV-2, a long-term follow-up study for this population is needed. In addition, the GMTs of anti-RBD-IgG and NAb in all participants were reduced over time, but GMTs were significantly reduced in the obesity and overweight groups compared with the normal BMI group. Similar results were reported for mRNA vaccines and in other regions. Pellini12 recruited 252 healthcare workers in Italy and analyzed the antibody titers 21 days after the first dose of BNT162b2 vaccination. They found that the antibody levels detected in underweight and normal-weight individuals were increased, and normal body weight was positively correlated with antibody response. A Turkish study4 found that individuals of young age (P < 0.01), female sex (P < 0.01), and not overweight or obese (P = 0.020) exhibited a significant IgG response to the SARS-CoV-2 inactivated vaccine. The latest study found that in the vaccinated cohort, there were increased risks of severe COVID-19 outcomes for people with obesity compared with the vaccinated population with a healthy weight.13 In fact, antibody titers may play an important role in the protection from SARS-CoV-2 infection and severe COVID-19 outcomes. A recent study reported that the occurrence of breakthrough infections with SARS-CoV-2 was correlated with lower neutralizing antibody titers during the peri-infection period.14 In summary, inactivated COVID-19 vaccines induced a poor antibody response in Chinese individuals with obesity/overweight; thus, more concern should be given to this special population.
Furthermore, we analyzed the RBD-specific B-cell response and spike-specific T-cell response because many studies have identified protective epitopes on the S-RBD protein, and most human-neutralizing monoclonal antibodies target this domain.15 , 16 MBCs and subsets are associated with protection from the reinfection of antigens. The rMBCs can persist for a long time and rapidly respond upon rechallenge with the same pathogen. ActMBCs are potential plasma cell precursors that represent the earliest population to migrate from the peripheral blood from germinal centers.17 IntMBCs likely represent a transitional state between MBC subsets. Our study showed that participants with obesity/overweight had a lower frequency of RBD-specific MBCs, which suggests that the durable humoral immune response may be impaired. Individuals with high BMI had lower frequencies of rMBCs and actMBCs but higher frequencies of intMBCs. A recent study reported a decreased frequency of rMBCs but increased frequencies of intMBCs and actMBCs in mild COVID-19 patients, which likely represents ongoing immune activation after SARS-CoV-2 infection.18 In addition, individuals with high BMI had a lower frequency of CD38+ CD27+ MBCs, which are also known as antibody-secreting cells (ASCs). Thus, this discrepancy indicated that B-cell immune reactivation after full-course vaccination with inactivated COVID-19 vaccines may be impaired in our studied subjects. Indeed, a decreased B-cell response has been observed in obese individuals on other vaccines. A small human cohort study19 reported that after influenza vaccination, obesity was associated with reduced antibody titers and decreased frequency of class-switched MBCs as well as increased frequency of exhausted MBCs in both young and elderly patients. Beyond humoral responses, successful protection against SARS-CoV-2 infection can be accomplished by cellular immune responses, including CD4+ T cells, CD8+ T cells, and their corresponding memory subsets.20 A recent study showed that T-cell activation was reduced in morbidly obese patients following double vaccination with BNT162b2.21 Reduced IFN-γ and TNF-α production in stimulated polyclonal T cells from subjects with obesity demonstrated an impaired T-cell antiviral immune response to influenza.22 These functional defects in T cells were consistent with our findings on inactivated COVID-19 vaccines in participants with obesity. Recently published studies hypothesize that obesity can cause chronic, low-grade inflammation, which may lead to T and B-cell defects, thus interfering with the immunogenicity of the COVID-19 vaccine.23
Considering that the data about the immune response to inactivated COVID-19 vaccines under different body fat distribution patterns are deficient in Asian populations, we compared the humoral responses to inactivated COVID-19 vaccines in Chinese individuals with and without central obesity. Our results showed that compared with the noncentral obesity group, the central obesity group had a weaker antibody response. The frequencies of RBD-specific CD38+ actMBCs in people with central obesity were significantly lower. Watanabe's study24 also showed similar results, namely, a higher waist circumference was associated with a lower antibody titer. In summary, this result demonstrated that compared to individuals with noncentral obesity, Chinese individuals with central obesity had weaker humoral responses to inactivated COVID-19 vaccines.
One of the limitations of our study is that these are preliminary results of a cross-sectional study. However, the study is ongoing, and further follow-up data will be reported in future publications. Another limitation is that we do not test the effectiveness of vaccines on circulating virus variants due to insufficient serum.
In conclusion, our results indicate that inactivated COVID-19 vaccines are safe and well tolerated but induce poor humoral and cellular immune responses in Chinese individuals with obesity/overweight. A large population-based cohort is necessary to further investigate the association between BMI and the effectiveness of COVID-19 vaccination. It is worth noting that our study outcomes disseminate the evidence to support the vaccination of COVID-19 booster doses for populations with obesity/overweight.
Author contributions
Concept and design: Hong Ren, Dachuan Cai. Funding acquisition: Hong Ren, Mingli Peng, Min Chen. Participant recruitment: Qian Zhu, Yingzhi Zhang, Juan Kang, Mingli Peng, Min Chen, Ying Mei, Jie Yang, Xiaoya Qi. Experiment execution: Yingzhi Zhang, Dejuan Xiang, Shuang Xiao, Gaoli Zhang. Acquisition, analysis, or interpretation of data: Yingzhi Zhang, Qian Zhu, Juan Kang, Zhiwei Chen, Min Chen, Hu Li, Mingli Peng, Dachuan Cai, Hong Ren. Drafting and critical revision of the manuscript: Qian Zhu, Yingzhi Zhang, Juan Kang, Zhiwei Chen, Dachuan Cai, Hong Ren. All authors were involved in writing the paper and had final approval of the submitted and published versions.
Conflict of interests
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Funding
This work was supported by the 10.13039/501100018537 National Science and Technology Major Project of China (No. 2017ZX10202203-007, No. 2017ZX10202203-008, and No. 2018ZX10302206-003) and a pilot project of clinical cooperation between traditional Chinese and western medicine for significant and complicated diseases of National Administration of Traditional Chinese Medicine: hepatic fibrosis. We also acknowledge the support of the 10.13039/501100001809 National Natural Science Foundation of China (No. 81772198), and the 10.13039/501100005230 Natural Science Foundation of Chongqing, China (No. cstc2020jcyj-msxmX0389).
Appendix A Supplementary data
The following is the Supplementary data to this article:Multimedia component 1
Multimedia component 1
Acknowledgements
The authors acknowledge the health management center and department of clinical laboratory of the Second Affiliated Hospital, Chongqing Medical University for their support and also extend acknowledgement to BioRender.com for the creation of images.
Peer review under responsibility of Chongqing Medical University.
Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.gendis.2022.10.023.
==== Refs
References
1 Ko J.Y. Danielson M.L. Town M. Risk factors for coronavirus disease 2019 (COVID-19)-associated hospitalization: COVID-19-associated hospitalization surveillance network and behavioral risk factor surveillance system Clin Infect Dis 72 11 2021 e695 e703 32945846
2 Tartof S.Y. Qian L. Hong V. Obesity and mortality among patients diagnosed with COVID-19:results from an integrated health care organization Ann Intern Med 173 10 2020 773 781 32783686
3 Butsch W.S. Hajduk A. Cardel M.I. COVID-19 vaccines are effective in people with obesity: a position statement from the Obesity Society Obesity 29 10 2021 1575 1579 34212511
4 Şenol Akar Ş. Akçalı S. Özkaya Y. Factors affecting side effects, seroconversion rates and antibody response after inactivated SARS-CoV-2 vaccination in healthcare workers Mikrobiyol Bul 55 4 2021 519 538 34666653
5 Weber D.J. Rutala W.A. Samsa G.P. Obesity as a predictor of poor antibody response to hepatitis B plasma vaccine JAMA 254 22 1985 3187 3189 2933532
6 Eliakim A. Schwindt C. Zaldivar F. Reduced tetanus antibody titers in overweight children Autoimmunity 39 2 2006 137 141 16698670
7 Centers for Disease Control and Prevention (CDC) Intensive-care patients with severe novel influenza A (H1N1) virus infection - Michigan MMWR Morb Mortal Wkly Rep 58 27 2009 749 752 19609249
8 Banga N. Guss P. Banga A. Incidence and variables associated with inadequate antibody titers after pre-exposure rabies vaccination among veterinary medical students Vaccine 32 8 2014 979 983 24394442
9 Chinese Health Management Association Expert consensus on the procedure of body weight management among patients with overweight or obesity (2021) Chinese J Heath Manag 15 4 2021 317 322
10 Zhang Y. Zeng G. Pan H. Safety, tolerability, and immunogenicity of an inactivated SARS-CoV-2 vaccine in healthy adults aged 18-59 years: a randomised, double-blind, placebo-controlled, phase 1/2 clinical trial Lancet Infect Dis 21 2 2021 181 192 33217362
11 Xia S. Zhang Y. Wang Y. Safety and immunogenicity of an inactivated SARS-CoV-2 vaccine, BBIBP-CorV: a randomised, double-blind, placebo-controlled, phase 1/2 trial Lancet Infect Dis 21 1 2021 39 51 33069281
12 Pellini R. Venuti A. Pimpinelli F. Early onset of SARS-COV-2 antibodies after first dose of BNT162b2:correlation with age, gender and BMI Vaccines 9 7 2021 685 34206312
13 Piernas C. Patone M. Astbury N.M. Associations of BMI with COVID-19 vaccine uptake, vaccine effectiveness, and risk of severe COVID-19 outcomes after vaccination in England: a population-based cohort study Lancet Diabetes Endocrinol 10 8 2022 571 580 35780805
14 Bergwerk M. Gonen T. Lustig Y. Covid-19 breakthrough infections in vaccinated health care workers N Engl J Med 385 16 2021 1474 1484 34320281
15 Zost S.J. Gilchuk P. Case J.B. Potently neutralizing and protective human antibodies against SARS-CoV-2 Nature 584 7821 2020 443 449 32668443
16 Noy-Porat T. Makdasi E. Alcalay R. A panel of human neutralizing mAbs targeting SARS-CoV-2 spike at multiple epitopes Nat Commun 11 1 2020 4303 32855401
17 Lau D. Lan L.Y. Andrews S.F. Low CD21 expression defines a population of recent germinal center graduates primed for plasma cell differentiation Sci Immunol 2 7 2017 eaai8153
18 Ogega C.O. Skinner N.E. Blair P.W. Durable SARS-CoV-2 B cell immunity after mild or severe disease J Clin Invest 131 7 2021 e145516
19 Frasca D. Ferracci F. Diaz A. Obesity decreases B cell responses in young and elderly individuals Obesity 24 3 2016 615 625 26857091
20 Chen Y. Yin S. Tong X. Dynamic SARS-CoV-2-specific B-cell and T-cell responses following immunization with an inactivated COVID-19 vaccine Clin Microbiol Infect 28 3 2022 410 418 34715346
21 Palacka P. Pol'anová M. Svobodová A. Effectiveness, adverse events, and immune response following double vaccination with BNT162b2 in staff at the national comprehensive cancer center (NCCC) Vaccines 10 4 2022 558 35455308
22 Costanzo A.E. Taylor K.R. Dutt S. Obesity impairs γδ T cell homeostasis and antiviral function in humans PLoS One 10 3 2015 e0120918
23 Holder K. Reddy P.H. The COVID-19 effect on the immune system and mitochondrial dynamics in diabetes, obesity, and dementia Neuroscientist 27 4 2021 331 339 32981462
24 Watanabe M. Balena A. Tuccinardi D. Central obesity, smoking habit, and hypertension are associated with lower antibody titres in response to COVID-19 mRNA vaccine Diabetes Metab Res Rev 38 1 2022 e3465 33955644
| 36466314 | PMC9707021 | NO-CC CODE | 2022-12-15 23:15:54 | no | Genes Dis. 2022 Nov 29; doi: 10.1016/j.gendis.2022.10.023 | utf-8 | Genes Dis | 2,022 | 10.1016/j.gendis.2022.10.023 | oa_other |
==== Front
J Infect Public Health
J Infect Public Health
Journal of Infection and Public Health
1876-0341
1876-035X
The Authors. Published by Elsevier Ltd on behalf of King Saud Bin Abdulaziz University for Health Sciences.
S1876-0341(22)00322-7
10.1016/j.jiph.2022.11.029
Original Article
First case series and literature review of coronavirus disease 2019 (COVID-19) associated pulmonary tuberculosis in Southeast Asia: Challenges and Opportunities
Siranart Noppachai a
Sowalertrat Walit a⁎1
Sukonpatip Manichaya a
Suwanpimolkul Gompol b
Torvorapanit Pattama b
a Faculty of Medicine, Chulalongkorn University and King Chulalongkorn Memorial Hospital, Thai Red Cross Society, Bangkok, Thailand
b Department of Medicine, Faculty of Medicine, Chulalongkorn University and King Chulalongkorn Memorial Hospital, Bangkok Thailand
⁎ Correspondence to: Department of Medicine, Faculty of Medicine, Chulalongkorn University and King Chulalongkorn Memorial Hospital, Thai Red Cross Society 1873 Rama 4 Road, Pathumwan, Bangkok, Thailand 10330. Fax: +66-2-256-3804.
1 Co-first authors
29 11 2022
29 11 2022
19 9 2022
9 11 2022
24 11 2022
© 2022 The Authors. Published by Elsevier Ltd on behalf of King Saud Bin Abdulaziz University for Health Sciences.
2022
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Background
Subclinical tuberculosis (TB) is accidentally detected by radiologic and microbiologic findings. Transmission by those with subclinical TB could delay prevention effort. However, our study demonstrated positive aspect of COVID-19 outbreak as it could allow subclinical TB to be detected faster through a chest X-Ray (CXR).
Methods
This cross-sectional cohort study aimed to report demographics, comorbidities, and outcomes related to early detection of TB among COVID-19 patients, and to elaborate the association between SARS-CoV-2 and pulmonary TB. Data of patients with SARS-CoV-2 co-infection with Mycobacterium tuberculosis (MTB) diagnosed between March 2020 – March 2022 was collected.
Results
Out of 12,275 COVID-19 patients, 26 were definitively diagnosed with MTB infection (mean age 48.16±20.17 years). All cases that had suspicious CXR that were not typical for COVID-19, were tested for MTB. On average, pulmonary TB was diagnosed after admission 5(3-10) days, the treatment initiation period was 3(1-5) days from the TB diagnosis.
Conclusions
This suggests an early detection of tuberculosis among COVID-19 patients by quicker screening CXR and sputum comparing to previous symptom guided screening. Thereby reducing the chance of TB transmission demonstrated during COVID-19 pandemic. So, clinicians should be aware of pulmonary tuberculosis in COVID-19 patients with atypical radiologic findings.
Keywords
Mycobacterium tuberculosis
subclinical tuberculosis
SARS-CoV-2
COVID-19
==== Body
pmc1 INTRODUCTION
Since 2020, the coronavirus disease-2019 (COVID-19) has still been an ongoing pandemic with 430 million cases and over six million deaths reported until February 2022 [15]. Simultaneously, tuberculosis (TB) also continues to be a global burden with 1.3 million deaths, ranking as the second leading cause of death after COVID-19 pandemic. (World Health Organization, 2020a) However, there are limited case reports or case series describing TB and COVID-19 co-infection, mostly reported in TB high burden countries such as India, China, and Bangladesh. A meta-analysis by Song et al. [8] suggested that patients with TB and COVID-19 coinfection were 2.21 and 2.27 times at risk of death and developing severe disease, respectively, more than COVID-19 patients without TB. Owing to the worse prognosis and overlapping clinical symptoms of these two diseases, routine screening for Mycobacterium tuberculosis (MTB) among COVID-19 cases was recommended in high-TB burden countries such as India, China, Indonesia, and Philippines. Thailand is also ranked 19th in TB high burden countries, yet no study was conducted. Therefore, this is the first study on COVID-19 associated pulmonary tuberculosis in Thailand and Southeast Asia. In addition, there is still a knowledge gap regarding the association between these two diseases whether tuberculosis just developed after COVID-19 or formerly dormant in the asymptomatic patients. This study will present the epidemiology, clinical presentation, diagnostic test, and treatment data of patients unexpectedly diagnosed with pulmonary TB after COVID-19 diagnosis.
2 METHODS
2.1 Definitions
To define COVID-19 cases, all patients with suspected clinical symptoms and/or compatible imaging findings were confirmed by nasopharyngeal and throat swab using reverse transcription polymerase chain reaction (RT-PCR) for SARS-CoV-2. FDA emergency-use authorized PCR kits used in this study including Cobas ® SARS-CoV-2 Duo Test and Fosun ® COVID19 RT-PCR detection kit.
For TB diagnosis, patients with radiographic findings that were atypical for COVID-19 but suspicious for TB, were screened with acid fast bacilli (AFB) smear of respiratory specimens, and subsequently confirmed by mycobacterium culture, RT-PCR for Mycobacterium tuberculosis complex, or Xpert MTB/RIF assay. Suspicious chest x-ray findings can be pulmonary consolidation shown as homogeneous or patchy opacities mostly in the middle and lower lobes, with or without hilar lymphadenopathy. Other findings suggestive of active pulmonary TB may include miliary opacities, pleural effusion, and pulmonary edema [1].
2.2 Study design
This was a retrospective study of epidemiological and clinical data from patients with SARS-CoV-2 coinfection with MTB within 6 months after the diagnosis of COVID-19. Data were collected during March 2020 to March 2022 at King Chulalongkorn Memorial Hospital (KCMH), Bangkok, Thailand.
2.3 Study procedures
Patients' information was collected using electronic medical records and a case report form. The information included demographic data, underlying conditions, history of upper respiratory infection symptoms, diagnostic tests for SARS-CoV-2 infection, history of closed contact TB, history of previous TB diseases and current TB symptoms, along with clinical course and medication history. Participants underwent chest radiographs, which were interpreted by radiologists. Those testing positive on symptom screening or chest radiographs were proceeded for further sputum smear microscopy, culture for M. tuberculosis and/or GeneXpert MTB/Rifampin (RIF) (Xpert; Cepheid, Sunnyvale, CA). Participants diagnosed with SARS-CoV-2 infection were admitted at King Chulalongkorn Memorial Hospital or underwent home/community isolation during the infectious period. Participants also underwent blood screening of HIV antibodies using 4th generation enzyme-linked immunoassay (ELISA) method. In case of positive anti-HIV screening, one another confirmatory test with HIV viral load would be also tested during admission.
2.4 Data analysis
Medical data were transferred in electronic format. Data were compiled and analysed in Microsoft Excel. Mean, standard deviation (SD), median, interquartile range (IQR) and frequencies (%) were used to describe individuals’ characteristics in this study.
3 RESULTS
According to a cross-sectional data collected at King Chulalongkorn Memorial Hospital, Bangkok, Thailand during March 2020 – March 2022, there were 12,275 COVID-19 patients diagnosed at our centre. Out of these, twenty-six cases demonstrated SARS-CoV-2 with MTB coinfection, 17 (65.38%) were male and 9 (34.62%) were female patients. Mean age was 48.16 years (range 17 – 85;SD 20.17). Five (19.23%) patients had history of smoking while 14 (53.85%) ones had chronic medical conditions.
Among all patients, the symptoms that can be associated with COVID-19 were cough (73.7%), low grade fever (57.69%), rhinorrhoea (23.08%), sore throat (34.32%), anosmia (15.38%), and myalgia (7.69%). Mean duration of COVID-19 symptom onset was 4.15 days (SD 3.44) before seeking medical care. Nearly half of them (46.15%) did not report symptoms compatible with pulmonary tuberculosis. However, for symptomatic patients, symptoms that associated with MTB infection were chronic cough; persisted more than three weeks, unrecognized and did not affect their daily activities (38.46%), and unintentional significant weight loss; defined as weight loss more than 5% in 30 days (15.38%). All patients were tested negative for anti-HIV. History of household contact with a pulmonary tuberculosis patient, was found in two cases. And two other patients previously served sentence in a correctional facility in Bangkok, which also constituted a risk factor of MTB contact. No drug-resistant tuberculosis was found in any patient by both genotypic and phenotypic susceptibility testing. The demographic data, clinical symptoms, laboratory investigations, radiological findings, treatments, and outcomes of each patient were described in Table 1 . Table 1 Characteristics and outcomes of twenty-six patients with subclinical tuberculosis (TB) and coronavirus disease (COVID-19) co-infection.
Table 1Case No. Age
/Gender Underlying conditions COVID-19 onset (days) COVID-19 symptoms Imaging diagnosis COVID-19 treatment History of pulmonary TB TB symptoms History of TB contact Laboratory for TB confirmation Period of TB diagnosis after COVID-19 (days) TB treatment initiation after diagnosis (days) TB treatment regimen Outcome at
6 months
1 31/F - 3 cough, low grade fever cavitary lesion at left upper lung zone Favipiravir for 5 days with supportive treatment no asymptomatic household contact pulmonary TB 1 year PTA RT-PCR for MTBC 4 3 2HRZE/4HR loss F/U
2 22/F - 14 sore throat reticulonodular opacity with cavitary lesion at right upper lung zone Favipiravir for 5 days with supportive treatment no subacute cough for 2 weeks no history of contact MTB C/S 5 3 2HRZE/4HR cured
3 18/M - 2 cough, rhinorrhea, fever, sore throat reticulopatchy opacity at right upper lung and left apex supportive treatment no chronic cough for 6 months no history of contact RT-PCR for MTBC and MTB C/S 3 3 2HRZE/4HR cured
4 50/M HCV infection without cirrhosis, 7 fever, sore throat, rhinorrhea nodular opacities at apical regions of both lungs Favipiravir for 5 days with supportive treatment no asymptomatic no history of contact RT-PCR for MTBC 2 3 2HRZE/4HR cured
5 60/M - 7 anosmia, cough with few sputum reticulonodulopatchy infiltration at left upper lung zone Favipiravir for 5 days with supportive treatment no asymptomatic no history of contact MTB C/S 4 2 2HRZE/4HR cured
6 33/M - 14 cough, rhinorrhea, anosmia, fever cavitary lesion at right upper lung zone Favipiravir for 5 days with supportive treatment no chronic cough for 1 month no history of contact RT-PCR for MTBC 6 3 2HRZE/4HR cured
7 54/F DM, HT,
active smoker 30-pack-year 1 cough, rhinorrhea, anosmia mixed patchy and nodular opacity at right lung Favipiravir for 10 days with supportive treatment no asymptomatic no history of contact RT-PCR for MTBC 3 0 2HRZE/4HR loss F/U
8 26/M Aggressive behaviour with history of amphetamine usage 2 fever, cough with sputum ground glass opacity at right lung Favipiravir for 10 days with supportive treatment no asymptomatic no history of contact RT-PCR for MTBC and MTB C/S 4 1 2HRZE/4HR loss F/U
9 76/M - 4 cough, sore throat, fever, myalgia ground glass opacity at periphery of left upper lung zone Favipiravir for 10 days with supportive treatment no asymptomatic no history of contact RT-PCR for MTBC 5 1 2HRZE/4HR loss F/U
10 17/M - 3 cough with sputum, right chest pain due to cough reticulonodular infiltration at apical region of both lungs with cavitary lesion at right upper lung Favipiravir for 10 days with supportive treatment no asymptomatic no history of contact RT-PCR for MTBC and MTB C/S 1 3 2HRZE/4HR transferred to other center
11 44/M - 8 cough, sore throat, fever, myalgia ground glass opacity at the bilateral lower lung Favipiravir for 5 days with supportive treatment no chronic cough with no sputum for 1 month,
weight loss 1-2 kg/month for 2 months correctional facility 2 years PTA RT-PCR for MTBC and MTB C/S 6 2 2HRZE/4HR cured
12 59/F DM 4 asymptomatic patchy opacity in left upper lung zone Favipiravir for 5 days with
dexamethasone 6 mg/day for 5 days yes whitish sputum, chronic cough,
low grade fever at night, unintentional weight loss 4-5 kg/months for 3 months no history of contact RT-PCR for MTBC 12 3 2HRZE/4HR treatment completed
13 32/F - 6 cough, sore throat, rhinorrhea ground glass and patchy opacity in both lungs Favipiravir for 10 days with Prednisolone 20 mg/day for 3 days then 10 mg/day for 3 days yes asymptomatic no history of contact RT-PCR for MTBC 8 17 2HRZE/4HR loss F/U
14 79/M pulmonary HT, COPD, PV/ET, CKD stage 3 3 fever, rhinorrhea, cough with sputum massive pleural effusion of both lungs Favipiravir for 2 days then Remdesivir for 8 days, with
dexamethasone 6 mg/day for 3 days then dexamethasone 10 mg/day for 5 days then dexamethasone 6 mg/day for 4 days then prednisolone 20 mg/day for 1 days then dexamethasone 6 mg/day for 4 days then prednisolone 20 mg/day for 6 days no asymptomatic no history of contact RT-PCR for MTBC and MTB C/S 4 1 2HRZE/4HR died
15 33/F DM 2 anosmia, cough recticular infiltration at both upper lungs supportive treatment no unintentional weight loss 3 kg/months for 2 months no history of contact RT-PCR for MTBC and MTB C/S 8 6 2HRZE/4HR cured
16 66/M - 3 cough with diarrhea reticulonodular infiltration at both upper lungs Remdesivir for 10 days no chronic cough with no sputum for 1 months, weight loss no history of contact RT-PCR for MTBC and MTB C/S 1 1 2HRZE/4HR cured
17 85/M - 2 cough, sore throat, fever reticulopatchy opacity at both lungs Favipiravir for 10 days, then Remdesivir for 2 days with Dexamethasone 6 mg 3 days, 10 mg 1 days, methylprednisolone 250 mg 3 days, Dexamethasone 20 mg 4 days, 15 mg 1 day no none no history of contact MTB C/S 4 1 2HRZE/4HR treatment completed
18 51/F DM, HT, DLP, Hyperthyroidism, Chronic HF 3 fever, dyspnea recticular infiltration at both upper lungs Favipiravir for 10 days no chronic cough household contact pulmonary TB 10 years PTA RT-PCR for MTBC and MTB C/S 30 6 2HRZE/4HR treatment completed
19 44/M - 5 cough, sore throat, fever normal lungs supportive treatment no chronic cough, unintentional weight loss 2 kg/months for 3 months correctional facility 6 months PTA RT-PCR for MTBC and MTB C/S 35 12 2HRZE/4HR cured
20 79/M HT, COPD 2 fever, dyspnea recticular infiltration at both upper lungs Favipiravir for 10 days, then Remdesivir for 2 days
with Dexamethasone 6 mg 3 days, 10 mg 1 days, methylprednisolone 250 mg 3 days, Dexamethasone 20 mg 4 days, 15 mg 1 day no chronic cough no history of contact MTB C/S 60 5 2HRZE/4HR treatment completed
21 75/M COPD, BPH 3 cough, sore throat reticulonodular opacity in both upper lungs zone Favipiravir for 10 days no chronic cough household contact pulmonary TB 6 years PTA RT-PCR for MTBC 10 1 2HRZE/4HR cured
22 51/F epilepsy, hypothyroidism 1 fever, dyspnea patchy opacity in left lower lung zone, with pleural effusion Remdesivir for 5 days with dexamethasone for 3 days yes none no history of contact RT-PCR for MTBC 20 5 2HRZE/4HR cured
23 60/M DM, ESRD, HT, DLP, TVD 2 cough, fever, dyspnea perihilar infiltration Remdesivir 1 dose then Favipiravir for 5 days yes palpable supraclavicular lymph node, subacute fever for 2 weeks no history of contact RT-PCR for MTBC and MTB C/S 21 5 2HRZE/4HR cured
24 29/F - 3 fever reticulonodular opacity in both upper lungs zone supportive treatment yes chronic cough for 4 months no history of contact MTB C/S 0* 2 2HRZE/4HR cured
25 40/M - 2 cough, atypical chest pain patchy opacity with lucency in the left lung supportive treatment no unintentional weight loss 4 kg/months for 3 months no history of contact RT-PCR for MTBC and MTB C/S 8 3 2HRZE/4HR cured
26 38/M s/p CDKT 2 cough recticular infiltration at left upper lung Remdesivir for 6 days yes none no history of contact MTB C/S 1 8 2HRZE/4HR cured
CDKT; Cadaveric kidney transplantations
CKD; Chronic Kidney disease
COPD: Chronic obstructive pulmonary disease
DLP; Dyslipidemia
DM; Diabetes mellitus
ESRD; End stage renal disease
HF; Heart failure
HT; Hypertension
PV/ET; Polycythemia vera with essential thrombocythemia
TVD; Triple vessels disease
C/S; Microorganism culture and sensitivity
H; Isoniazid
R; Rifampicin
Z; Pyrazinamide
E; Ethambutol
F/U; follow-up
Treatment completed; microorganism negative on culture or RT-PCR, but lesion was still found in chest radiograph
Cured; treatment completed, and no lesion found on chest radiograph
⁎ Tested potitive for COVID during TB treatment
All cases had suspicious CXR that were not typical for solely SARS-CoV-2 infection, which made the attending physicians suspected other respiratory diseases, namely reticulonodular opacity (19.2%), ground glass appearance (15.4%), patchy infiltration (15.4%), reticular infiltration (15.4%), cavitary lesions (7.7%). All patients were confirmed of MTB infection by RT-PCR detection, or microorganism culture and sensitivity. Subsequently, appropriate tuberculosis treatment regimens were initiated with an average period of 3.50 days (SD 3.54) from COVID-19 diagnosis.
All patients received standard symptomatic care. Favipiravir, which was a standard antiviral for mild to moderate COVID-19 of Thai national guideline during that time, was distributed to 15 patients (57.69%, mean duration 7.13 days, SD 2.87). While 6 patients (23.08%) received remdesivir (mean duration days 4.00, SD 2.75) because of more severe symptom or having medical risk of COVID-19 severity progression.
There are 6 patients who received corticosteroid as adjunctive treatment. Dexamethasone, prednisolone, and methylprednisolone were used in varying dosage and duration as described in Table 1 .
On average, TB was diagnosed median 5 days (IQR 3-10 days) after admission. Anti-TB medications were initiated shortly after the diagnosis median 3 days (IQR 1-5 days). Twenty-five patients recovered without respiratory complication or hypoxia needing oxygen supplementation. Only one required ambulatory oxygen canula. All patients had no drug adverse reaction, nor drug interaction between favipiravir and anti-TB drugs including isoniazid (H), rifampicin(R), pyrazinamide (Z) and ethambutol (E).
The outcome of every patient was documented after TB treatment. Among 26 TB and COVID-19 coinfected patients, 18 patients (69%) had completed the HRZE treatment with clinically full recovery, sputum AFB conversion, and improved CXR. While 4 patients (15%) completed the treatment with minimal residual lung lesion such as minor focal atelectasis and minor consolidation at left upper lung. One patient with disseminated TB also completed treatment with full recovery, which makes 84% treatment completion rate. In addition, there was no drug resistance reported among these patients. However, one patient (3.8%) with TB pleuritis following massive pleural effusion both lungs was later dead with congestive heart failure. Unfortunately, 3 (12%) patients had lost follow-up from the study hospital, probably due to COVID-19 situations.
We hereby presented three index cases of COVID-19 patients that piqued our attending physician interests, one of them was MTB infection shortly after discharge from COVID-19 admission (case #13), the other was a newly diagnosed MTB infection (case#14), and the third case was disseminated TB (case #23)
Case #13: A 32-year-old female with a history of pulmonary tuberculosis for 5 years ago. She had no underlying conditions. Her COVID-19 presented with cough, sore throat, and rhinorrhoea. After RT-PCR for SARS-CoV-2 was confirmed, oral favipiravir was prescribed for 10 days. Owing to her hypoxia, prednisolone tapering dose 0.2-0.4 mg/kg/day prescribed for 6 days also. After a-week, she recovered from COVID-19 symptoms and was discharged without any complication. Seventeen days after COVID-19 diagnosis, chest radiograph shows ground glass opacity at both lungs. This makes the attending physician suspect pulmonary TB. Afterwards, RT-PCR for MTB complex was performed, and detected. Pulmonary TB treatment were administered consisting of 2-month induction of 2HRZE and then 4-month maintenance therapy with 4HR. She now resolved from pulmonary TB.( Fig. 1)Fig. 1 A Chest X-Ray of case #4 shows nodular opacity in both lungs. B Chest X-Ray of case #10 showed reticular reticulonodular infiltration at apical region of both lungs, and cavitary lesion at right upper lung.
Fig. 1
Case #14: 79-year-old Thai male, with pulmonary hypertension, chronic obstructive pulmonary disease, and chronic kidney disease stage 3, had fever, rhinorrhoea, and cough for 3 days as the initial symptoms of COVID-19. He had no history of previous MTB infection. Chest radiograph showed massive pleural effusion in both lungs, which made the attending physician suspect pulmonary TB infection. The imaging was shown in the Fig. 2. RT-PCR for MTB complex was subsequently done and respectively detected. After 2-day of oral favipiravir oxygen desaturation was progressed, so he was given intravenous remdesivir for 8 days later. High dose dexamethasone was also administered intravenously for 5 days. On fourth day of dexamethasone, his oxygenation became worse, with lowest at 81% saturation. He complained of shortness of breath, so he received step-up oxygen supplementation by oxygen cannular, and then high flow nasal cannula. He also developed sharp chest pain which cardiac rubbing sound. Chest computed tomography (CT) showed bilateral pleural thickening with pleural effusion. Needle thoracocentesis was proceeded which later showed positive for AFB staining of pleural effusion suggesting TB pleurisy. Dexamethasone was tapered to oral prednisolone with total 23-day duration. His TB was treated with 2HRZE and 4HR continuum. He was markedly improved since a couple week of TB treatment and now his CXR returned to previous background with resolved effusion. Unfortunately, he died 3 months after the treatment was completed due to congestive heart failure.Fig. 2 A Chest X-Ray of case #14 showed multiple lungs nodules in the right lung, along with reticular infiltration in the right lower lung field. B and 2 C Chest computed tomography of case #14 showed multiple hyperdensity lesions in both lungs (arrow), and bilateral pleural effusion (star).
Fig. 2
Case #23: Thai male, aged 60 years, with end stage renal disease, hypertension, triple vessels disease, and type II diabetes mellitus, was admitted to our hospital due to COVID-19 RT-PCR detection two days after symptoms of cough, low grade fever, and mild dyspnoea. His CXR revealed perihilar infiltration in both lungs. He had no previous history of MTB infection also. Oral Favipiravir was administrated. He did not need corticosteroid because of mild COVID-19. On the fourth day of admission, the physician noticed the palpable mass at right supraclavicular area. Fine needle aspiration was done at the lesion, then AFB and PCR for MTB complex both detected. The imaging was shown in the Fig. 3 . Sputum RT-PCR for MTB complex was detected also. His disseminated TB treatment consisted of 2HRZE and 4HR regimen. The patient was discharged after 14 days of admission with clinical improvement and no adverse drug reaction found.Fig. 3 A Neck computed tomography of case #23 showed enlarged right supraclavicular lymph node (arrow). B Chest computed tomography of case #23 showed multiple hyperdensity lesions (arrows) in the right lung.
Fig. 3
4 DISCUSSION
In the year 2021, Thailand is one of the thirty countries with highest burden for tuberculosis with an incidence of 150 per 100,000 people. According to our results, the incidence of newly diagnosed TB infection was 26 of the 12,275 COVID patients, equalled to 205 per 100,000 people, which was little higher than the incident rate among general population without COVID-19. The possible explanation could be that the included patients might be subclinical TB with revealed their abnormal CXR during COVID-19 diagnosis.
Twenty-one (21/26, 80.8%) patients were diagnosed with TB coinfection within 2 weeks after COVID-19 diagnosis. It could possibly be subclinical TB as we discussed before, or TB reactivation that just occur after COVID-19. Nevertheless, this implied a promising possibility of early detection of TB infection, which leads to early effective treatment. Consequently, early diagnosis helps cut down transmission within households and communities, decrease burden of TB, and leads to better outcomes, which include curation (culture conversion among bacteriologically confirmed pulmonary TB patients), and treatment completion (treatment completed without evidence of failure, but without sputum smear or culture results (World Health Organization, 2020b)).
Regarding other five TB cases diagnosed more than 2 weeks after COVID-19 diagnosis, these findings point out a possible association between TB and COVID-19 infection. SARS-CoV-2 virus as well as its treatment, potentially had an influence on the immune system through various mechanisms which finally resulted in the reactivation of tuberculosis. Firstly, there was direct viral damage to the cells of the respiratory system on which epithelium expresses angiotensin-converting enzyme 2 (ACE2), the main target receptor of SARS-CoV-2 [5]. Secondly, there is cell-mediated immune damage by excessive inflammatory cell infiltration, activation of complement cascade and release of biologically active substances in the lung parenchyma, aggravating diffuse alveolar damage [7]. Thirdly, in severe cases of COVID-19, most patients might develop lymphopenia. Numbers of B-cells, CD4+ & CD8+ T-cells, and NK T-cells would be decreased, especially T-cells, which were the key adaptive immunity to fight against tuberculosis. [16] To summarize, both long-lasting lung damage and impaired immune function following COVID-19 possibly predisposed patients to progression, reactivation, or acceleration of MTB infection.
Apart from viral infection and host immune responses, other medications prescribed during COVID-19 treatment, especially anti-inflammatory agents, also play an important role in immune disturbance, might further predispose to secondary infections. Generally, corticosteroids use could elevate risk of developing tuberculosis by regulation of numerous genes responsible for immune response that is required in tuberculosis control; for example, inhibition of lymphokine effect, monocyte chemotaxis, and Fc-receptor binding, circulated lymphocytes, especially T-cells. [3]
For COVID-19 patients, a prospective meta-analysis of 7 clinical trials reported reduction in all-cause mortality among those with severe and critically ill conditions, regardless of whether they required mechanical ventilation. As a result, corticosteroids were recommended as a part of standard care for critically ill COVID-19 patients (Group TWREAfC-TW, 2020). On the other hand, corticosteroids administration was not supported in mild or less severe patients. World Health Organization (WHO) guideline, launched in 2020, suggested systemic corticosteroids, administered either orally or intravenously, at a dose of 6 mg of dexamethasone, which is equivalent to 150 mg of hydrocortisone, or 40 mg of prednisone, or 32 mg of methylprednisolone, for up to 7-10 days in severe and critical COVID-19 patients. (World Health Organization, 2020c) However, the previous study from American Thoracic Society (ATS) and US Center for Disease Control and Prevention (CDC) (2000) revealed that the risk of TB reactivation increased with oral prednisolone for more than 15 mg per day, or equivalent, for 2-4 weeks. It is remarkable that the dose proposed for COVID-19 treatment substantially exceeds the dose at risk for enhance TB reactivation. So, TB reactivation should be monitored especially among severe to critical COVID-19 who received high and prolonged corticosteroids.
The TB/COVID-19 Global Study Group (2022) had reported 767 TB and COVID-19 co-infected patients from 34 countries. There were 74% of patients with history of previous TB, 16.5% diagnosed within the same week from suspicious imaging, and 9.5% initially diagnosed with COVID-19. Interestingly, among the last group, 35 out of 71 patients were diagnosed COVID-19 more than 30 days prior to TB, which is considered sufficient time for TB disease development, thus the study discussed that COVID-19 may not play a major role in TB progression. Another cohort study from South Africa by Western Cape Department of Health in collaboration with the National Institute for Communicable Diseases SA (2019), included 22,308 COVID-19 patients, of which 2,128 (27%) got TB and COVID-19 coinfection. Among this group, 22% had history of previous TB and 5% also known as current MTB infection. However, more evidence will be needed to fill gaps in knowledge about the role of COVID-19 in TB reactivation.
Favipiravir and remdesivir are the RNA-dependent RNA polymerase inhibitors which were not used widely until the COVID-19 pandemic. From previous meta-analysis, both drugs were associated with clinical and laboratory improvement in COVID-19 patients with no serious adverse drug reactions [6]. Favipiravir is mainly metabolised in the liver by aldehyde oxidase which has a short half-life of 2-5.5 hours and can cause asymptomatic liver injury with increased liver transaminases [2]. Remdesivir, half-life of 20 hours, is also metabolised by the liver, causing transient elevation of aminotransferase which is not usually linked to clinical injury in COVID-19 patients. Also, anti-TB drugs are well known for drug-induced liver toxicity ranging from asymptomatic to severe toxicity. However, nearly all patients in our study, had no significant elevation of liver function tests during treatment with anti-TB drugs shortly after favipiravir or remdesivir. The possible explanation of these findings could be rapid clearance and short half-life of the COVID-19 medications that might not significantly raise liver toxicity. The interactions between COVID-19 treatment and anti-TB drugs are still limited, thus further studies are needed, and safety monitoring should be observed.
The outcome of patient TB treatment shows that 69% had completed the HRZE treatment and fully recovered. While 15% completed the course of treatment with minimal residual lung lesions. The mortality rate was 0%
We also compared the outcome with COVID-19 patients without TB sub-grouped into clinical spectrum defined by US-CDC; mild/moderate, severe and critical COVID-19. In our center, the mortality rate of the critical group, who had respiratory failure or multiorgan system dysfunction, thus required intensive care unit admission, was 170 in 390 (43.6%). While the mortality rate of the severe patients, who were complicated with pneumonia, acute respiratory distress syndrome, sepsis, acute kidney injury, or secondary bacterial infections, was 40 in 274 (14.6%). In contrast, the mortality rate of mild/moderate groups who did not need hospitalization, was only 15 in 7654 (0.2%). The overall mortality rate was 2% among patients without MTB coinfection.
While data collected between 2018-2021 in our center showed success treatment 446 in 538 (83%)) of patient with isolated TB infection; 111 (25%) were cured, and 335 (75%) received full course of medication. The overall mortality rate among solely TB infection was 31 in 538 (6%) that more than half found in complicated medical underlying hosts
However, TB and COVID-19 coinfection in our study mostly revealed among mild/moderate COVID-19. We could not conclude that TB has no effect on COVID-19 mortality. This was probably due to small number of index cases, and under-detection.
Our study also has several strengths to guide novel information and policy making. Firstly, it is the first study in Southeast Asia that reports the incidence of COVID-19 and TB coinfection and outcomes after receiving medications. Secondly, characteristics, symptoms, and findings of each case were elaborately described. Every patient was diagnosed with COVID-19 and TB infection by using RT-PCR, which reported within 3 days, resulting in early detection, early treatment, and reduction of transmission. Thirdly, the literature regarding mechanisms and impacts of COVID-19 infection as well as corticosteroids therapy on tuberculosis infection were delicately reviewed. This study, by reporting cases with COVID-19 and TB coinfection, emphasized the importance of further research to obtain more evidence and clearer understanding on both topics. Also, health care providers should be more aware of tuberculosis infection when treating COVID-19 patients with atypical radiographic findings especially in high TB burden countries.
However, our study was a retrospective study and had a small number of index cases. We are aware that our index TB cases mainly triggered by using CXR imaging, which might result in under-reported cases. However, we collected data for two years and followed patients who received steroids whether they develop MTB infection after 6 months of treatment. Although about 5-10% of patients with latent TB could progress to active disease within the first two years in general, this number might be shorter among immune dysregulated hosts. So we suggest further research and authority could implement an active screening of tuberculosis among COVID-19 patients to obtain more cases, which might result in higher early detection and good treatment outcomes. Extended period of follow-up in post COVID-19 patients could also be implemented to obtain clearer association between COVID-19/COVID-19 treatment with corticosteroids and TB reactivation. Furthermore, additional laboratory investigation, involving the immune mechanisms such as certain proinflammatory cytokines and immune cells might be collected to analyse possible correlation, and to serve as a biomarker to predict disease severity and risk of coinfection in the future.
Moreover, our study reported an incidence of tuberculosis on COVID patients in high burden areas. These results could not be applicable for low burden countries, thus further studies will be needed.
5 CONCLUSION
Tuberculosis is an epidemic in the Asia-Pacific region. The current COVID-19 pandemic might help speed up the diagnosis process as it allows patients to be screened quicker than before. Clinicians should be aware of pulmonary tuberculosis in COVID-19 patients with atypical radiologic findings, especially in high TB burden countries. The association between SARS-CoV-2 and TB remains unclear, and needs further studies.
Ethical Approval statement
Informed consents were obtained from the patients for the publication of their information and radiographic images.
Funding Source
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Conflict of Interest
None
Uncited references
[10], [11], [12], [13], [14], [4], [9]
==== Refs
References
1 Burrill J. Williams C.J. Bain G. Conder G. Hine A.L. Misra RR. Tuberculosis: a radiologic review Radiographics 27 5 2007 1255 1273 10.1148/rg.275065176 17848689
2 Du Y.X. Chen X.P. Favipiravir: Pharmacokinetics and Concerns About Clinical Trials for 2019-nCoV Infection Clin Pharmacol Ther 108 2 2020 242 247 10.1002/cpt.1844 32246834
3 Fauci A.S. Dale D.C. Balow J.E. Glucocorticosteroid therapy: mechanisms of action and clinical considerations Ann Intern Med 84 3 1976 304 315 10.7326/0003-4819-84-3-304 769625
4 Sterne J.A.C. Murthy S. Diaz J.V. Slutsky A.S. Villar J. Angus D.C. Association Between Administration of Systemic Corticosteroids and Mortality Among Critically Ill Patients With COVID-19: A Meta-analysis Jama 324 13 2020 1330 1341 10.1001/jama.2020.17023 32876694
5 Lin L. Lu L. Cao W. Li T. Hypothesis for potential pathogenesis of SARS-CoV-2 infection-a review of immune changes in patients with viral pneumonia Emerg Microbes Infect 9 1 2020 727 732 10.1080/22221751.2020.1746199 32196410
6 Rezagholizadeh A. Khiali S. Sarbakhsh P. Entezari-Maleki T. Remdesivir for treatment of COVID-19; an updated systematic review and meta-analysis Eur J Pharmacol 897 2021 173926 10.1016/j.ejphar.2021.173926
7 Short K.R. Kroeze E. Fouchier R.A.M. Kuiken T. Pathogenesis of influenza-induced acute respiratory distress syndrome Lancet Infect Dis 14 1 2014 57 69 10.1016/S1473-3099(13)70286-X 24239327
8 Song W.M. Zhao J.Y. Zhang Q.Y. Liu S.Q. Zhu X.H. An Q.Q. COVID-19 and Tuberculosis Coinfection: An Overview of Case Reports/Case Series and Meta-Analysis Front Med (Lausanne) 8 2021 657006 10.3389/fmed.2021.657006
9 Targeted tuberculin testing and treatment of latent tuberculosis infection. American Thoracic Society MMWR Recomm Rep 49 Rr-6 2000 1 51
10 Tuberculosis and COVID-19 co-infection: description of the global cohort Eur Respir J 59 3 2022 10.1183/13993003.02538-2021
11 Risk Factors for Coronavirus Disease 2019 (COVID-19) Death in a Population Cohort Study from the Western Cape Province e2005-e15 South Africa. Clin Infect Dis 73 7 2021 10.1093/cid/ciaa1198
12 World Health Organization. Corticosteroids for COVID-19: living guidance, 2 September 2020; Geneva; 2020.
13 World Health Organization. Definitions and reporting framework for tuberculosis [2013 revision: updated December 2014 and January 2020]. https://www.who.int/publications/i/item/9789241505345, 2020 (accessed 22 February 2022)
14 World Health Organization. Global tuberculosis report 2020. Geneva; 2020.
15 World Health Organization. WHO COVID-19 Dashboard. https://covid19.who.int/, 2020 (accessed 22 February 2022)
16 Zheng H.Y. Zhang M. Yang C.X. Zhang N. Wang X.C. Yang X.P. Elevated exhaustion levels and reduced functional diversity of T cells in peripheral blood may predict severe progression in COVID-19 patients Cell Mol Immunol 17 5 2020 541 543 10.1038/s41423-020-0401-3 32203186
| 36495816 | PMC9707022 | NO-CC CODE | 2022-12-08 23:15:57 | no | J Infect Public Health. 2023 Jan 29; 16(1):80-89 | utf-8 | J Infect Public Health | 2,022 | 10.1016/j.jiph.2022.11.029 | oa_other |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.